MXPA99003771A - Molding machine by injection, controlled by adaptive process, auto-adjust - Google Patents

Molding machine by injection, controlled by adaptive process, auto-adjust

Info

Publication number
MXPA99003771A
MXPA99003771A MXPA/A/1999/003771A MX9903771A MXPA99003771A MX PA99003771 A MXPA99003771 A MX PA99003771A MX 9903771 A MX9903771 A MX 9903771A MX PA99003771 A MXPA99003771 A MX PA99003771A
Authority
MX
Mexico
Prior art keywords
ram
speed
signals
signal
machine
Prior art date
Application number
MXPA/A/1999/003771A
Other languages
Spanish (es)
Inventor
C Bulgrin Thomas
Original Assignee
Van Dorn Demag Corporation
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Van Dorn Demag Corporation filed Critical Van Dorn Demag Corporation
Publication of MXPA99003771A publication Critical patent/MXPA99003771A/en

Links

Abstract

The present invention relates to an injection molding machine which uses a law of control of multiple summed terms to control the water hammer speed during the injection stroke of a molding cycle to emulate a speed profile adjusted per user. An automatic calibration method adjusts ram speeds without load, to double ram speeds adjusted per user. Finite impulse response filters produce no-load, open-loop control signals at advanced positions in the velocity profile to account for the delay in response of the system. An adaptive error term indicative of load perturbation that is observed from a preceding cycle is added to the advanced travel position that is predicted by the finite impulse response filter to produce a control signal for the machine supply valve of supply of the machine

Description

MACHINE FOR MOLDING BY INJECTION, CONTROLLED BY ADAPTIVE PROCESS, SELF-ADJUSTED This invention relates to injection molding machines and more particularly to the control system of the molding machine. The invention is particularly applicable to and will be described with specific reference to a system for controlling the speed of the piston or ram of a hydraulically powered injection molding machine during the injection stroke. However, those skilled in the art will recognize that the invention has broader application and can be applied to all electric injection molding machines, blow molding machines and other devices where the translation speed of a reciprocating primary motor it must be controlled precisely. INCORPORATION BY REFERENCE My previous US patents No. 5,456,870 dated October 10, 1995, with the title "Barrel Temperature State Controller for Injection Molding Machine" (Barrel Temperature State Controller for Injection Molding Machine) and No. 5,493,503 dated February 20, 1996 with the title "Clamp Control for Injection Molding Machine "(Clamping Control for Injection Molding Machine), here incorporated by reference.
My prior patents do not form part of the present invention but rather explain details of the controls of the injection molding machines that are useful for understanding the present invention, such that these details do not necessarily need to be described in detail here. BACKGROUND The molding cycle performed by an injection molding machine typically comprises the phases of gripping, injecting, packing, recovering and expelling. This invention deals mainly with the injection phase or injection stroke of the molding cycle. The study of how plastic circulates through the mold is called mold flow analysis. Plastic flow is critical for a number of factors in the final part including surface faults and structural integrity. Mold flow analysis predicts the shape of the part at all times, as it is formed and can predict areas of the part where a minimum injection speed is required to fill the part before the gates freeze or probably a maximum injection speed over which the unfolding can cause superficial faults. The experience gained by the use of any particular mold will show when certain speeds are required in the filling of mold to produce acceptable molded parts. In a production base, the typical practice is to minimize the injection time to achieve maximum performance of the machine. All this is achieved by controlling the injection speed of the ram during mold filling and, to a lesser extent, by the packing phase of the cycle which ensures that the mold remains filled with molding material under a desired pressure until the molding material it solidifies. The first molding machines were operated manually. To inject plastic into the mold, a crank wheel was used. The faster the wheel rotated the more. fast would be the injection speed. To increase the pressure, the crank would turn harder. When the crank wheels were replaced with hydraulic systems, the injection was controlled by the oil flow to the injection cylinder. Various techniques have emerged over the years to control oil flow and pressure. As noted above, controlling the flow of the oil controls the speed of the ram and controlling the speed of the ram is regulated as the molding material already in the mold. As the mold flow analysis matured into a science, the art of injection molding was replaced with decisions dictated by the mold flow analysis that requires aggregation. controls to the hydraulic system to control the flow and pressure of the oil. This invention is directed to said system (although the invention is also applicable to controlling the speed and torque of motors employed in "all-electric" molding machines). The control systems were initially simply synchronizers that caused opening of various valves at certain times during the injection stroke. The synchronizers gave rise to the microswitches that were triggered by water hammer to cause various adjustments of the valve. The microswitches were replaced by position detectors with feedback in current use developing signals used by programmable controllers to regulate the speed of the ram. The typical injection molding machine most likely observed today has an operator station with a display and keyboard display that sends signals to a programmable logic controller (PLC = Programmable Logic Controller). The operator station typically includes a screen where the operator can adjust desired ram speeds to fixed ram travel increments. Typically, the ram travel of the machine is divided into ten zones, lengths or equal sections. The operator adjusts the speed in each zone so that a series of bar graphs are assembled. The ram follows the bar graphs. Recent improvements in control have replaced the bar graphs with points in each zone boundary, so that the ram is not programmed to travel at a constant speed within each zone but a speed that constantly varies from a fixed point in a boundary of zone to another fixed point in an adjacent zone border. When the operator makes a number of desired speed adjustments to fixed ram travel positions that the ram will follow, it establishes a "speed profile". The objective of the control system is to actually cause the ram to travel at the user's adjustment speeds in the positions set by the user, that is, to emulate the velocity profile. As will be explained in the detailed description of the invention below, this invention causes better water hammer control to be possible than has been possible to date. The control within the PLC that causes the ram to follow the velocity profile is typically a PID controller (ratio / integration / derivative controller). A PID controller receives a speed feedback signal from a 'detector on the machine and compares it with the control signal of speed adjusted by user, to generate a signal of control compensated by error by which the speed of the machine is controlled. The control signal is then converted to an analog pulse signal which controls a solenoid valve which regulates a hydraulic supply valve, which in turn controls the flow of a pump to a primary motor causing movement of the ram. The PID controller is the typical mechanism for achieving closed-loop control. It is generally used because the machine control modules are typically purchased by injection molding machine manufacturers from the control suppliers who assemble control systems for special applications, such as injection molding machines, of any amount of common control modules that have response times, sensitivities, desired robustness, etc. The systems use a common control such as a PID controller. The set points supplied by the operator are converted into speed control signals after the machine has been configured and calibrated. Functionally, the set points provided by the user in the machine console are converted into setpoint speed signals by the PLC and output in the water course positions as analog pulse signals. PLC uses the controller PID to provide closed loop control (front-facing water hammer detector signals) to ensure that the appropriate setpoint speed signal, corrected by error, is sent out as the impulse signal. To prevent the PID controller from generating large error corrections, the normal practice is to manually calibrate the supply valve for setpoint speed signals. The solenoid voltage is adjusted manually until the movement of the ram is visually detected and the analog signal causes this initial movement to be stored in digital form as a valve displacement. The solenoid voltage is then manually adjusted to a valve where the maximum rated speed of the ram is observed to occur. For example, if the maximum ram speed for the machine is 2.54 cm (1 inch) per second and the climb path was 12.7 cm (five inches), a technician, using your watch, will manually adjust the solenoid voltage to which is capable of causing the ram to travel 12.7 cm (five inches) in about 5 seconds. At that point, the expansion or extension voltage in digitized form will be provided. When considering a straight line relationship between displacement and extension, a digital signal that corresponds to a user's adjustment speed at any point between zero and 2.54 cm (1 inch) per second is calculated by the machine control. If the machine has an energy saving mode, a second calibration must be performed. This technique is not precise. However, conventional thinking is that any errors that occur can be addressed in the PID control and the calibration can only be performed without load. In this way, adjustments are made to simply ensure the capacity of the machine. The PID control, however has to be tuned. While of course pre-defined factory settings are provided, the adjustment is made through a trial and error procedure by the molding machine manufacturer's technician during machine configuration. Basically, the procedure followed after adjusting the extension and effect is to reinforce the ram during the injection stroke and observe the speed response. Various techniques of "form in the specialty", are used to regulate factory settings if the response is considered slow. Many current machines allow the operator to choose open loop or closed loop control to achieve speed profiling. Open loop mode is achieved by simply using the manually calibrated valve settings that are established during the machine configuration, as described. The closed loop is achieved through the PID controller set as described or by using the predefined factory settings for the PID controller. In practice, it often happens that the PID loop is out of sync with wear and age or if the moulder simply changed the molded part. The machine user does not have the sophistication to re-tune the PID loops and the closed-loop control does not follow the velocity profile. In fact, in many cases the machine with the open loop control will more closely follow the velocity profile even when, considering that perfect calibration of the supply valve with the open loop control can not take into account the load and specifically the disturbances or resistances imposed by the fusion in the ram during the injection stroke. The foregoing summarizes in large part what the assignee of this invention has observed in the market with respect to the current control systems (apart from the control system of the transferee which is the subject of this invention). The literature has however described a number of control techniques applied for injection molding control systems. The patents of the U.S.A. No. 5,645,775 issued to Spahr et al; No. 5,258,918 granted to Giancola; No. 5,182,716"issued to Stroud, III et al. And No. 5,062,785 issued to Stroud, III et al., Describe control systems that have specific characteristics to control the speed of the ram, in these systems, the velocity profile is broken down into zones. As discussed above, in order to move uniformly from one zone to the next zone, the controls move from open loop to closed loop within a zone.In addition, positive power is discussed, but in the sense of feeding a signal present in advance Additionally, adaptive learning concepts are described The latest techniques are well known in control theory and the references simply show that they have been applied to control systems for injection molding machines. No. 5,482,662 issued to Nakamura et al., Describes a somewhat more sophisticated approach to control of ali positive feedback for water hammer speed, since the pressure detected by the valve is used to develop a positive feed signal to account for valve latency response and another control term is added when the feedback signal of the water hammer position reaches a given differential ratio at the determined speed limit. The valve control is said to eliminate excessive correction or overcorrection tendencies of the valve resulting from changing signals and speed feedback contributes to precision during steady state. Nakamura's system seems to be more advanced than the previously discussed concepts, but still uses an additional term commuted in or out of control during injection, depending on the feedback of a current event. The patent of the U.S.A. No. 5,578,256 issued to Austin, utilizes the relationship between ram velocity and mold flow to develop a plastic flow characteristic, ie, pressure, detected in the mold during a stroke that is then fed to the control as a term of adaptive error in the next successive cycle. The patents of the U.S.A. Nos. 4,753 ', 588 granted to Kiya and 5,552,690 granted to Hiraoka, are related to injection molding machines with electric displacement, which control water hammer speed. Kiya describes using a search table of excessive power correction to modify the adjustment speeds. Hiraoka uses separate control terms to control motor speed and motor torque. The feedback control in the Zone boundaries are used to regulate the torque and speed settings. None of the systems discussed seems to use the concept of positive feeding in the predictive sense as described in my U.S. patent. Do not. ,493,503 and this invention can be seen as an extension of and an improvement to said patent. In addition, systems in general are articulated to detect events in zones and make changes during zone advances when switching control modes or simply by adding learned, adaptive error signals. None of the cited systems discuss the set point signals. They simply generate the set point signals that correspond to the user set points and then use feedback techniques to produce the desired ram control. SUMMARY OF THE INVENTION According to this, a main objective of the present invention is to provide a control system for an injection molding machine, wherein the speed of the ram continues without seams or joints, without delay of the speed profile established in a learned form. This feature together with other aspects or features of the present invention is achieved in a system (method and apparatus) for variably controlling the ram velocity of an injection molding machine during the injection stroke of a molding cycle and wherein at the operating station, a plurality of user adjustment points that specify fixed ram speeds in fixed ram travel positions, are supplied by the machine operator. A mechanism within the machine control automatically establishes a second, larger plurality of machine set point signals automatically from user set points, with each set point signal indicative of water hammer speeds set in travel positions. of ram established and cumulatively defines a travel site and speed points that include the operator set points and which in turn define a speed profile that the ram emulates during the injection stroke. A control structure is provided to determine a load compensated speed control signal for each signal of machine set points during the injection cart. The control structure that establishes the control signal includes a mechanism to detect the real-time position of the ram during the injection stroke and a mechanism to apply a calibrated factor automatically set to each machine set point signal which causes the ram to move at approximately the speed of a set point signal when the machine is not under load. The calibrated factor, by ensuring repeatable duplication of the speed profile with the machine that is not under load, provides a basis or foundation upon which any control law can be constructed, to take into account load disturbances when the machine fills the mold with a load of molding material. According to an important feature of the invention, the structure for determining the load-compensated velocity control signal includes a predictive mechanism for selecting setpoint signals in advanced ram travel positions that have not been traversed at the time that the set point signal is chosen to process as a speed control signal, such that the speed control signal takes into account the response latency of the machine. Significantly, the set point signal that occurs in the future when it is chosen to process, is a signal of predictive set points and not a common positive power signal because it is that speed signal that occurs in future time at a specific ram travel distance in the speed profile and not a positive feed signal present in time for processing. According to another important feature of the invention, the structure for determining the speed control signal (which in turn are transformed into a continuous analog pulse signal) also includes a disturbance mechanism for modifying each pre-selected control signal by a error signal. The disturbance mechanism includes a feedback structure for comparing a real-time ram speed signal with a real-time set point signal in the velocity profile at the real-time ram position, to produce a signal of difference that corresponds to- the error signal. A disturbance storage mechanism stores each error signal produced in fixed ram travel positions during any given injection stroke. The disturbance mechanism chooses an error signal stored in the advanced ram travel position corresponding to the advanced travel position in any given predicted control signal. The predictive control signal is then added with the advanced disturbance error signal. The stored error signals are overwritten by the error signals (preferably a percentage of them) generated during a cycle of error. current or current molding, which are then employed in the next successive molding cycle to provide open loop control compensated for disturbance of predictive load of the ram speed during the injection stroke of the molding machine. According to another important feature of the invention, the structure for determining the speed control signals includes the mechanism for developing real-time water speed signals from the real-time water position signals detected in the machine and a controller for developing error signals that modifies the signals real-time set points processed as speed control signals. The controlling mechanism (which in the preferred embodiment is a PID controller) includes a feedback structure for comparing real-time ram velocity signals with specific setpoint signals in the velocity profile at the ram travel positions in time. real, to produce a difference signal corresponding to an error signal whereby any setpoint signal selected to be processed as a speed control signal is modified by the error signal, in a closed loop structure, to take into account differences between a specified ram speed in real time and a water hammer speed detected in real time resulting from load disturbances that are detected in the current injection stroke. According to yet another important aspect of the invention, the structure for determining the charge-compensated speed control signal selectively summs the various velocity control signals defined above to produce the driving signal at the option of the operator. In this way, the operator has the option of i) operating the machine only in a closed loop mode, such that only speed control signals produced by an error signal comparing real-time speed profile speeds. with real-time ram velocities, they are used in the above-mentioned way to control ram velocity or to detect load disturbances in real time or ii) only one open-loop predictive velocity control signal is used to control velocity of the ram or iii) only, an open-loop predictive ram velocity signal as defined above, modified by the learned charge disturbance error signals or iv) an open-loop predictive velocity control signal with or without control adaptive is added with the closed-loop real-time speed control signal. In all the In some cases, the resulting control signals produce the ram velocity for impulse signal control. According to a specific important characteristic of the invention, the machine is hydraulically displaced and the method by which the calibrated factor for modifying the set point signals is adjusted by machine and determines when performing the following step: i) initialize the machine wherein the translational axis of the ram is adjusted and the water hammer detectors of the machine are adjusted to record the voltage and water hammer speed signals which are fed to the hydraulic which supplies the control ram speed as digitized signals; ii) increasing the voltage output to the valve until the water movement detected by the machine occurs and the machine registers the output signal as a valve displacement signal; iii) increase the voltage output beyond the displacement and record the water hammer speed detected by each increase; iv) Subsequently, register in machine the voltage output level where the nominal maximum ram speed is detected as a valve extension signal. By providing a program for machine calibration on the supply valve, a precise correlation, without load between set points and water hammer speed, is obtained for each specific value, not However, the variations in values, etc., between identical model machines. Additionally, an automated calibration procedure allows non-skilled operators to easily recalibrate the machine as it ages, or to provide multiple calibration adjustments if the machine is equipped with an energy saving feature. According to another specific feature of the invention, the calibration method also includes the additional steps of recording the time it takes for the ram to reach at least a certain percentage of the maximum ram velocity and use that time for the machine to automatically adjust the PID controller. An object of the present invention is to provide an injection molding machine with improved speed profiling characteristics, by using an automatic valve calibration process that establishes valve adjustments including extension and displacement, such that the ram will consistently travel to speeds equal to the speed points established by the user with the machine that is not under load and in this way produce a basis for developing a control law to model the machine speed under load.
A more specific object of the invention in conjunction with the immediately preceding object is also to provide in the valve calibration process, an automatic tuning of a PID controller incorporated in the machine control system. Still another object of the invention is to provide an injection molding machine with improved speed profiling characteristics, which is capable of performing the entire injection stroke under closed loop control or under open loop control or under open loop control and closed loop combined. An important objective of the invention is to provide an injection molding machine that performs speed profiling by any one or more or any combination of the following characteristics: a) closed-loop control by real-time feedback control through the use of a PID controller or otherwise; b) open loop predictive position based on speed control through FIR filters or otherwise; c) open loop predictive position based on speed control adjusted by learned perturbation response; Y d) calibrated adjustment for speed set points to produce known responses under no load conditions. Still another important objective of the invention is to provide an improved injection molding control for speed profiling which is self-learning and is adapted over several cycles to consistently reproduce any speed profile that is established by the operator. An important objective of the invention is to provide a speed profiling system for an injection molding machine that is capable of better emulating the water velocity position signals set by the user than other systems and / or that can process a range wider profiles than other systems. An important objective of the invention is a speed profiling system for an injection molding machine, which generates a large number of signals at discrete distance distances, almost infinitesimally spaced, to accurately model a desired velocity profile. Still another important objective of the invention, in combination with the immediately preceding objective, is the use of finite impulse response filters, which are used to generate predictive signals based on variable speed user adjustment points and real-time information automatically detected from machine calculations, thus improving the system's ability to process a wide range of speed profiles established or adjusted by the user. Still another object of the invention is the provision of a method for automatically tuning the PID controller used in conventional injection molding machines, thus avoiding trial-and-error approaches conventionally employed and providing more responsive control, no matter what the law of the invention. Control is actually used by the machine to control the speed profiling. Another important object of the invention is to provide an improved system for achieving speed profiling in an injection molding machine by means of a control law that adds the control signals developed specifically to address different factors affecting the water hammer speed, thus avoiding inconsistent results that are produced by other systems that already use complicated methods or modulation techniques that can easily fail in practice for any number of reasons.
Still another object of the invention is to provide an improved system for achieving speed profiling in an injection molding machine that utilizes existing physical machine equipment and existing physical control equipment, such that an injection molding machine having capabilities improvements may be offered without price increases normally associated with said improved ones. Another important object of the invention is to provide a control law for speed profiling for an injection molding machine that is superior to that used in other conventional systems. Yet another object of the present invention is to provide in a hydraulically displaced injection molding machine, a method to automatically calibrate the displacement and extension settings of the proportional valve to provide improved speed profiling characteristics of the machine, no matter what they are. the control techniques currently employed to provoke speed profiling. These and other objects, features and advantages of the present invention will be apparent to those skilled in the art of reading and comprehension of the detailed description of the invention. is set forth below taken in conjunction with the drawings. BRIEF DESCRIPTION OF THE DRAWINGS The invention may take form in certain parts and in an arrangement of certain parts as a whole and with the drawings forming part thereof and in which: Figure 1 is a partial schematic view of the ram of a machine injection molding; Figure 2 is a general schematic of the control system employed for the injection molding machine shown in Figure 1; Figure 3 is a graph indicative of a speed profile screen illustrating prior art concepts; Figure 3a is a graph indicative of a speed profile screen similar to Figure 3 illustrating concepts employed in the present invention; Figure 4 is a flow diagram showing steps that are carried out to calibrate the machine according to the present invention; Figure 5 is a block-shaped schematic illustrating the general processing of water hammer signals in the invention; Figure 6 is a schematic, block diagram showing internal control loops employed in the present invention; Figure 7 is a schematic diagram similar to Figure 6, but showing the external control loops; Figure 8 is a distance versus time plot similar to Figure 6 of my U.S. patent. No. 5,493,503 showing the water hammer distance traveled over a period of time; and Figure 9 is a general schematic in the form of blocks of portions of the sequencing and analog cards illustrated in the control system of Figure 2. DETAILED DESCRIPTION OF THE INVENTION Now with reference to the drawings, wherein the illustrations are with the purpose of showing a preferred embodiment of the invention only and not for the purpose of limiting the same, illustrated in Figure 1, is a schematic representation of an injection mechanism 10 employed in an injection molding machine. A fragmentary portion of the machine holding mechanism indicated by the reference numeral 12 is also illustrated.
The injection mechanism 10 includes a spindle 14 which moves and rotates disposed within a tubular barrel 15. In the preferred embodiment, the translation of the spindle 14 within the barrel 15 is achieved by a hydraulic coupling or hydraulic actuator shown including a piston of sealing 16 mobile within a cylinder 17. Spindle rotation occurs by rotation of the displacement arrow 18 subject to a mechanical coupling 20, which in the schematic diagram illustrated, is a gear box 20. In the structure illustrated in FIG. Figure 2, the gearbox 20 is connected to and displaced by an electric motor 21. Alternately, rotation of the spindle 14 can occur by a conventional hydraulic piston motor. For schematic illustration purposes, the displacement arrow 18 is shown slotted to the piston 16, such that piston 16 can slide inside the cylinder 17, to cause translation of the spindle 14 while the rotation of the pulse arrow 18 causes the piston 16 turn spindle 14 for spindle recovery purposes, etc. When the spindle 14 is moved by translation to the mold mechanism 12 for injecting the molding material into a mold cavity 23, it is commonly referred to as a ram. Because the present invention is related to controlling the translation movement of the spindle 14 during the injection stroke or injection phase of the molding cycle, the spindle 14 hereinafter will be referred to as a ram 14. Although other terminology such as "piston" can be used to describe the spindle 14 during injection, as used herein, "ram" means the spindle 14 and covers the movement of the spindle, whether or not it rotates while moving in the barrel 15 during injection of the molding material into the mold. Also for definition purposes, the injection stroke of the molding cycle begins when the molding material pushed by the ram 14 leaves an open end 24 of the barrel 15 (shut-off valve not shown) to initially enter the mold cavity 23, and continues until the mold cavity 23 is initially filled with the molding material, where the injection stroke ends. Additional movement of the ram 14 to supply additional molding material to the mold cavity 23, to take into account volumetric shrinkage as the molding material solidifies, is regulated by the packaging phase of the molding cycle through a different control structure that the employee to control the movement of ram during injection.
The movement of the piston 16 which in turn directly controls the translation movement of the ram 14 is controlled, in the preferred embodiment, by a directional supply valve 25. Again, Figure 1 is a schematic. In current operation, an injection manifold that includes, but is not limited to, valves that control the output of the injection pump (s) is typically provided. Similarly, a separate manifold can be provided for valves that control the holding mechanism 12. However, in each of these manifolds a separate supply valve is employed in the preferred embodiment. Specifically, in the preferred embodiment, the supply valve 25 is controlled in flow. It is usual for the supply valves to regulate both the pressure and the flow separately. Due to the size of the pump, the high torque output of the motor displacing the pump and other considerations, the pressure does not have to be controlled separately (apart from a conventional safety relief valve). Accordingly, in the schematic of Figure 1, a constant supply pump 26 is moved by a motor 27 with the pump 26 having a capacity and the motor a power, such that the flow control by the valve supply 25 will control the speed of the ram 14. The pump Constant supply 26 is provided with a conventional safety relief valve 28 connected to the manifold 29. A machine controller 30 controls the operation of the injection molding machine. The machine controller 30 essentially comprises a station or operator console 32 in which the operator supplies data or set points that define how the machine will be operated and a programmable logic controller, PLC 34, which receives operator instructions and data from the detectors in the machine, process the data and generate control signals to the actuators in the machine. As far as the present invention is concerned, the machine has a water path position detector 35 which generates a transducer signal in a water position detector line 35a. In this way, the exact position of the ram 14 indicated schematically by the arrow 36 is detected in real time and an internal clock within PLC 34 allows the PLC 34 to track the ram travel positions at synchronized increments to generate signals of observed speed. Alternatively, physical equipment differentiation can be used to obtain an analog feed speed signal. In response to speed points set by the operator and travel positions of observed ram, PLC 34 will generate an analogue shift signal in the impulse signal line or controller 37 to a solenoid valve that controls the operation of the supply valve 25. As noted above, by controlling the amount in which the supply valve 25 is opened or closed, the oil flow rate from the pump 26 is regulated to control the speed of the piston 16 and ram 14. The injection molding machine is also equipped with a detector 38 to detect the pressure exerted by the ram 14 in the fusion and generates a pressure signal on the line 38a to the PLC 34. Additionally, the injection machine is optionally coupled with a pressure sensor (not shown) that detects the pressure of the melt in the cavity of mold 23 and generates a pressure signal in the pressure signal line in cavity 39 to PLC 34. Conceptually, pressure detectors can be used as a substitute for the water path position detector or as an additional signal for controlling the speed of the ram 14. The components so far described as components are conventional. Now with reference to Figure 3, a velocity profile graph 40 is generally illustrated as it appears on the speed screen that the operator of the machine may require at the operator station 32.
The screen or graph displays traces of the travel distance of the ram on the x-axis and the speed of the ram on the y-axis. Typically, the operator adjusts desired ram speeds in desired ram travel positions up to a maximum of approximately 10 or 11 set points. The operator sets points as illustrated by the reference letters A, B, C, D, E, and F in Figure 3, and if the points are connected by straight line segments as illustrated, the velocity profile curve indicated by the reference number 40 can be drawn. Conventional controls observed in use currently divide the ram travel into equally distant zones, typically 10. The first controls adjusted the constant speed for the ram in each zone. For example, an average velocity for each zone represents that the midpoint of the velocity profile curve in each zone can be used to adjust the ram velocity for that zone. A profile curve as illustrated by the graph 42 passing through the circles will result. As long as the speed profile does not change significantly in speed, a reasonable result would theoretically occur. The constant speed zone approach was subsequently improved by adjusting speed points at the zone boundaries and ramping the ram speed between the fixed points. border. This would theoretically produce a current velocity profile curve designated by the reference number 43 that passes through triangles in dashed lines shown. Whenever the given velocity profile was gradual, or the fixed points changed in the border areas, a reasonably closed velocity profile could theoretically result. However, if there were an abrupt change in speed such as at points "A" and "C", the ram would not emulate the fixed points per user. The discussion is theoretical up to now and considers that the battering ram can actually reach the speeds of the predetermined points. In practice, this has not been observed on a consistent basis. In fact, several of the patents discussed in the background use techniques where the water course in a portion of a given area would be conducted on an open loop basis and then would switch to a closed loop control at some point during travel of a determined area. Additionally, it should be noted that during operation, the operator display will show a trace of the current speed of the ram superimposed on the velocity profile established in Figure 3. Graphs 42, 43 theoretically show what the ram velocity profiles would be like if the machine in fact complied with the points established by the control systems. Graphics 42, 43 simply show that current control systems can not, depending on the speed profile established, theoretically comply with the established speed profile curve 40. In current practice, the ram velocities do not follow the fixed zone points supplied by the control system. machine. To some degree, there is simply a current physical limitation on what the team might be. For example, the point C in the velocity profile curve 40 is a transition marked from a rather steep deceleration to a gradually increased acceleration which is established at the point D. If the increase in the curve between the points C and D is escaping, the machine can not emulate the velocity profile at point C. Apart from this, the problem is that the control simply does not cause the ram 14 to meet the predetermined points. The injection molding machine is an item of complicated machinery composed of many systems and subsystems. It has been observed that identical machines manufactured with the same components (all of which are supplied under strict quality control procedures) having the same capacity and model designation, will not produce the same speed profiles for identical parts manufactured in the same machines. due to variations between valves, pumps, motors, etc.
One of the foundations of the invention then is to recognize that the variations between identical machines must first be addressed before an acceptable law or control model can be applied. That is, speed profiling is a two-stage problem. This invention solves both problems but it should be recognized that solving the first problem in a certain proportion is independent of the second problem, ie the control law. That is, any control law can be developed to represent an improvement over existing systems as long as the first mentioned problem is addressed. The invention solves the first problem by machine adjustment for the control system taking into account any variations attributed to a specific machine, in such a way that the machine can precisely reproduce the speed profile when it is not under load. By adjusting the machine to precisely reproduce the speed profile set by the users when the machine is not under load, a reference base is established on which a control law can be modeled to take into account the disturbances and resistances encountered by the ram. when a melt charge is injected into the mold. Clearly, if the reference condition can not be established, the accuracy of any control law used by the machine when it is under load. All PLC control systems in injection molding machines develop digital control signals that are converted into a variable voltage analog signal in a D / A converter. When developing a digital speed signal, some feedback information regarding the position of the ram and / or the speed of the ram has to be fed to the PLC. As noted in the background, this feedback information takes the form of an error signal that modifies the setpoint signal set to produce the digital speed control signal which in turn generates the analogue shift signal. Typically, a PID controller is used within the PLC to produce the speed control signal. Conceptually, then the configuration of the machine requires: a) that the voltage signal controlling the supply valve 25 be coordinated with the digital speed adjustment points provided by the end user in the machine and b) the PID controller is adjusted to produce accurate responses to the set point signals. As previously indicated, adjusting the speed and response of the machine to changes in speed was done manually by the technician during machine configuration. (When the machine ages, either the manufacturer's technician or the machine operator had to recalibrate manually). The solenoid voltage operating the supply valve 25 is manually adjusted until the ram is first observed to move and the voltage is recorded as a digital valve displacement signal within the PLC 34. The voltage was then manually changed until the Ram travel at maximum rated speed will be observed and that the voltage level will likewise be recorded in PLC 34 as the extension voltage. The straight line between the two points then establishes a digital adjustment for any speed of the machine. The technician will then reinforce the speed of the ram during injection, to see how quickly the PID controller responds. Various manual adjustments to the PID loops are made until the technician feels that the PID controller causes the ram to exhibit a rapid change of speed, ie "increase". The present invention, automatically or by machine, calibrates the control system and follows the flow chart illustrated in Figure 4. Essentially, the automatic valve calibration procedure involves applying a test voltage sequence to the machine and observing the response. The nature of the test voltages and their sequence is specific to the valve that is calibrated, however the following stages General are taken for each valve. First, the ram axis of the machine is adjusted or aligned and other machine conditions are set in such a way that either the machine is operated in a normal mode or in an energy saving mode in block 50. This is followed by a safety check in block 51. In block 52, the machine (factory calibrated at zero voltage) zeroes the solenoid valve voltage to the supply valve 25 and digital speed values in the PLC adjust to zero. While the remaining steps of the calibration process can be performed manually with verification from the data observed on the screen at the operator station 32, a calibration program using data acquired from the water path position detector 35 in the line 35a. The controller 30 causes the voltage to increase slowly in the voltage increase block 54. When the water movement is detected by the water position detector 35 in the block 55, the voltage increase in the block 54 is stopped and the digital adjustment recorded in the machine's memory as a valve displacement in block 57. The digital values are at some value in the valve displacement voltage and those values are increased more and more in ten percent increments (in the preferred) and steady state ram speeds detected by the ram travel position detector 35 stored in memory at each voltage increase. The digital voltage signal is increased until the maximum rated ram speed of the machine is detected in block 60 and the value at maximum ram speed is stored in memory as the reel displacement value in block 61. new, the detected speeds are steady state speeds and adjustments are made automatically by the machine, although a similar procedure can be followed manually, but without the precision achieved by the machine settings. A signal adjustment for digital speed control correlated with the water hammer speed can now be established in the control memory by constructing a graph that passes through the incremental voltage points from displacement to valve extension. In practice, the relationship is linear and a straight line passes through max (reel displacement) and min (valve displacement) is constructed (and verified by the incremental voltages recorded in block 58 that can displace the line). The machine can now produce a steady state speed equal to any speed that is set by the user in a discharged condition.
With the machine now capable of producing, with no constant speed loading at any fixed point of determined speed, the PID controller is adjusted. Specifically, the time it takes for the machine to reach a speed, the augmentation time, is again recorded by the water position feedback detector 35. That is, with the ram at rest, the maximum rated speed is adjusted and the time on the display screen as the speed increases until a certain speed is reached, it is recorded. In the preferred embodiment, time measurements for the ram to reach twenty percent of the maximum nominal travel and time measurement for the ram reaching eighty percent maximum travel is recorded in block 63. The time values are then used to adjust the PID in block 65. In general, digital routines are used to adjust PID analog gain values for proportional, derived, and integral loops. Digital routines follow known adjustment methods such as CDH or Ziegler-Nichols-based methods. Essentially, two of the loops have their gains reduced to zero while the gain of the third loop is increased until a fixed or determined rise time is observed. The other loops have their earnings adjusted to a ratio of Loop adjustment and the time of climb of the machine is finally verified. Alternately, after the first loop is incrementally adjusted, the second loop adjusts incrementally in gain until a second time of ascent is observed. Then, the final loop adjusts incrementally until the time of ascent recorded at 20% and 80% meets certain time periods. Again, the process is carried out automatically through a programmed routine. If the machine is equipped with an energy saving mode, the auto-tuning procedure is repeated with the machine setting for the energy saving mode. As the machine ages, periodic recalibrations are carried out. Now with reference to Figure 5, the general process steps by which the displacement signal is generated in the impulse signal line 37 are illustrated. As already discussed, the operator in the operator station 32 enters a plurality of what are referred to as "user set points" in the user set point block 65. The user set points are indicated by the AF reference letters shown in Figure 3A and indicate desired ram speeds at distances of specific ram travel. After the set points are supplied in block 65, PLC 34 calculates approximately one thousand signal points of adjustment in the signal block of set points 66. If the distance axis shown in Figure 3a is divided into one thousand equal increments, PLC 34 for each increment will calculate a velocity such that the site of the loop points equally spaced over the The distance axis will have a determined speed for each point to define the speed profile graph 40. In practice, line segments between the reference letters are connected by straight lines. However, given the large number of water hammer / ram velocity adjustment points calculated by PLC 34 (which in the preferred embodiment are approximately one hundred points per 2.54 cm (1 inch)), shape adjustment curves are possible. the signals of set points. When the mold flow analysis demands said profiles, the control system of the present invention will be able to duplicate these profiles. Currently, and in the preferred embodiment, the speed profile adjustment signals are simply generated according to the straight line equation y = mx + b, for the signals set points that are between the reference letters set points adjacent. In fact, previous versions of the control system simply map a set of setpoint signals that are then accessed by a lookup table to develop an inline velocity profile curve straight of Figure 3A. The present invention contemplates, as an alternative embodiment, accessing a plurality of water / speed distance signals from a data map of said face-to-face signals to the search tables to generate a velocity profile curve 40. In any mode, the plurality of set point signals is significantly larger than the plurality of user set points. Again, because the preferred embodiment calculates each set point signal in the velocity profile, higher order velocity profile curves can be generated as well as straight line profiles. Still further, it is possible to take the velocity profile adjustment points in block 66 and modify them by calibration factors set forth in Figure 4. In the preferred embodiment, the calibration factors are applied to the signal points of adjustment as the control signals are developed in accordance with the control law of the present invention. The setpoint signals are then converted into control signals in the control signal block 68. The control signal block 68 is illustrated in greater detail in Figure 6, with its inner loops and with its outer loops in the Figure 7. The developed control signals are then converted into analogue displacement signals through the signal block displacement D / A 69. The analog shift signal as explained is fed in the line of displacement signals 37 to the supply valve 25 in the preferred embodiment. Alternatively, the control signals left by the control signal block 68 can be fed as the displacement signal to an electric motor if the injection molding machine has an electric driver. For example, if conventional ball screws are used as the displacement in place of piston 16, the speed of the electric motor moving the ball screw will be controlled by the control signals from the control signal block 68. Even more specifically, if the motor that displaces the ball speed is an AC (AC) induction motsr, then the reference speed signal supplied to the vector control for this AC induction motor would be the control signal emanating from block 68. It can be made reference to the US patent No. 5,362,222 (incorporated herein by reference) which describes a vector control for an AC induction motor that displaces a ball screw, causing translational movement of the ram. The vector control requires a power signal to move the motor. The power signal is the control signal developed by this invention.
Within the control signal block 68 are three control signals, i.e. a closed loop, a real-time control signal developed by a PID controller 70, an open-loop predictive control signal developed by a predictive controller 72 and a adaptive control signal developed by an adaptive controller 75. While the machine has the ability to run in closed loop or open loop or open loop with adaptive control, in the preferred embodiment, the control signals for all three controllers simply add up at 76 to produce the final control signal. The control law of the invention can be established as: Control Signal = PID Control Signal + Predictive Control Signal + Adaptive Control Signal which can be rewritten as: u (t) = PID (t) + 0L (t)) + L (f (t)) Equation 1 where: u (t) = Control Signal @ timet Equation 2 PiD. { t) = P. { and i) - m + j. jf r) _ r x (t)} dt + D,. { - () } Equation 3 OL (f (t)) = Span - r (x (t + h)) + Offset Equation 4 L (f (t)) = L (x (t + h)) Equation 5 Now with reference to Figure 6, A schematic that best illustrates the control law described above is shown in greater detail. As discussed above, the ram position feedback detector 35 generates a continuous analog signal on the line 35a indicative of the current ram position in real time. The real-time analog ram position signal is digitized on an analog-to-digital converter 78. The control system is equipped with a conventional synchronization circuit and the digitized real-time position signal x (t) on line 79 is converts to a real-time speed signal in the speed conversion block 80 that includes a finite impulse response filter. The finite impulse response filter is used to generate the real-time speed signals based on current and past time ram position events. The real-time speed signal is fed to the PID controller 70 on line 82. As already explained with reference to Figure 5, the setpoints fed from the operator station 34 are sent through the setpoint line of feed 83 to the set point signal calculator 66 which takes the current ram position x (t) from the ram position line 79 and develops a signal of real-time velocity set points v @ R (x (t )) on the setpoint signal line 84. The speed setpoint signal is taken from the speed profile curve 40 at the real time position x (t) as indicated schematically by the reference letter "x" shown in graph 40, illustrated in the setpoint signal calculator 66. The speed setpoint signal of line 84 is summed with the real-time speed signal in line 82 to produce a error signal that is processed by the proportional, integral and derived loops of the PID controller 70 in a well-known manner to produce a closed-loop speed control signal on the line 86. In the preferred embodiment, the displacement and extension settings calibrated I do not know they apply to the speed signals of real-time set points since the error signal is processed by the PID controller 70. However, as noted in the previous discussion, the calibrated valve settings can be applied to the Real-time speed adjustment points signals. The real time position signal x (t) on the real time line 79 also passes through a filter of finite impulse response 88, whereby a predictive position signal x (t + h) is converted in the predictive position line 89 which is passed through the set point calculator 66 to produce a point velocity signal of predictive adjustment v @ R (x) (tth). The signal of predictive adjustment points is then passed through the predictive controller 72 that modifies the signal from advanced set points by the calibration values discussed above to produce a predictive speed control signal in the line 90. The finite impulse response filter 88 uses current real-time data with respect to past observed data to determine a distance that the batter travels attributed to the latency of response or delay of the whole system. This includes "instability" control as well as response latency of the supply valve, system moment, etc., all of which are included in the term "response latency". The distance traveled by the ram 14 attributed to the response latency is calculated by the finite response filter 88 and added to the real time position of the ram 14. This advanced travel position illustrated by the reference letter x 'in the Figure 5 (and also in Figure 3a) is the position employed in the setpoint signal block calculator 66, to generate the Advanced setpoint speed signal in the advanced position of speed profile 40. The mathematics by which the finite impulse response filter 88 calculates the predictive position is direct. The distance is equal to the speed by time. With reference to Figure 8, the velocity is the slope of the velocity profile curve 40 at any given time and a portion of the velocity profile curve between the reference letters B & C, is illustrated as a solid line ending in a real time position t? with future positions of the velocity profile curve 40B-C, which are shown as a dotted line extending from there. The velocity is the slope of the solid line 40B-C, that is, the velocity and can be expressed as: x = Xi ~ "tj. - t0 Equation 6 where Xi is the current position of the ram 14, x0 is a previous position of the ram 14 at time t0, tx is the time at which the current position of ram 14 is detected, t0 is usually 50 milliseconds before the current time ti- The predicted position xat at the advanced time,? T corresponds to the delay time (such as response latency) can be described as: xat = Xi + x '? T Equation 7 Equation (6) can be rewritten as follows: xat = (? -t + 1). x - (? -t) .x0 Equation 8 which in turn is the classical finite impulse response filter form of: xat = Si-Xi + a2.x0 Equation 9 where: t-l ~ t0 a, =? t ti t0 Equation 9 predicts time. present the position in which the ram 14 will be after the response latency has been reached. If the > graph of a plurality of xat's taken over a period of time, the line of points and dashes designated by 40'BC will be generated and 40'BC would be parallel to 40 BC (considering constant speed and precise sampling) but the displaced or advanced from there a time delay equal to ? t. The line 40'B-C will intersect any point in the water hammer position in time, Δt, before the ram 14 would in fact reach the water ram position. The finite impulse response filter takes and harnesses the feed signal to a predicted position that is compared to the current position and the position is calculated by analyzing the past and present positions of the ram over time to determine how far the ram will travel. during the response latency of the control or other system variables. It should be noted that the calculation is not complex and can be handled quickly by the processor (less than 0.75 millisecond in the preferred mode). This makes the control system responsive and precise. In addition, it will be appreciated that the forward position will vary because the water hammer speed is not constant. That is, not only the velocity profile 40 is adjusted to be at varying speeds but within each straight section of the velocity profile 40, the ram velocity will follow an undefined heading and will be subject to random variations (interference), as it travels beyond various points of position. To produce a straight profile and allow precise use of a finite impulse response filter, there are numerous speed setpoint signals for numerous machine calculator, ie 1000. At each calculation, several points in x: and x0 are taken and averaged (in the preferred modality, four adjacent points are averaged). By taking an average number of set points spaced close to very fast intervals, the predictive ram positions xat will remain very close and precisely the velocity profile curve 40. This is shown in an exaggerated position in Figure 3A. In general, a large number of set points allow rapid FIR sampling to produce predictive position setpoint signals, to carefully follow the velocity profile curve 40 from which the desired setpoints speed signal can be adjusted. Again, unlike other positive power techniques, the x signal simply will not be output directly but takes into account the events that will arrive, to generate a new signal from the velocity profile. The speed signal at x 'based on a predictive distance calculation is used to establish the control signal. This control signal when modified by valve calibration settings in the predictive control block 72 will now produce an open loop signal that can carefully and precisely control the speed of the no-load ram. As noted earlier, many injection molding machines use constant supply pumps and the Water hammer speed is controlled by regulating only the flow from the pump because the pump pressure is more than sufficient to provide necessary torque. In practice, there are many small load type molding applications where the machine in normal operation mode (and not in energy-saving mode that limits the output capacity of the pump) can process the molded part with only the loop Predictive open That is, load disturbance does not significantly affect the ability of the machine to travel at calibrated "no-load" speeds. The error signal generated from the PID controller 70 as a closed-loop control signal is saved for each real-time ram position where it is generated in a look-up table. In this way, at the end of each injection stroke, a search table of error correction factors for the velocity profile 40 of the immediately preceding injection stroke is generated. This search table is combined with the real-time error correction signals that are generated during a current injection stroke. In practice, only a percentage of the error signal we say 50% is added to the search table. Also, the table modification continues only for a limited number of molding cycles. In the modality preferred, the table modification stops after eight cycles. The error correction search table is illustrated diagrammatically in Figure 7 as a disturbance search table 95. The disturbance search table in this manner is a memory of the disturbance or load for each point on a profile curve. speed 40. Physically, the load during the injection stroke is due to the resistance of the plastic to circulate and changes considerably during the injection stroke as the plastic freezes over the guides and as the fusion front passes through obstacles and turns in the geometry of the mold. The effect of the adaptive control on the currently produced traced velocity profile is schematically represented by the trace 45 which is the first molding cycle, trace 46 which is the second molding cycle trace "apprehended" and the tracing 47 which is the third speed profile trace "apprehended again". Now with reference to Figure 5, the predictive position signal, x (t + h) in the predictive position line 89 and the set point in line 83 is sent to the calculated set point controller 66 to develop a disturbance , speed setpoint signals, v @ L (x) (t + h), on line 96 that are sent to the adaptive controller 75. The adaptive controller 75 access the disturbance search table 95 to generate a disturbance control signal on line 97. It should be remembered that the predictive controller 72 adjusts the speed control signals taking into account the response latency of the system, with the system that It is not under load. The adaptive controller 75 adjusts the speed control signals that take the load into account. The only way in which the load can be taken into account is to observe that resistance imposes the load on the ram 14 during the injection stroke. This observation is made conveniently but precisely by the error signal generated in the PID controller 70. In this way, a predictive open loop speed control signal is generated, which is adjusted to the measurement by each individual machine by the Control signals without load, calibrated and for each particular molded part by the disturbance control signals. In the preferred embodiment, all speed control signals are summed at the summing junction 76, to produce the composite control signal that is sent to the digital-to-analog converter 69 to develop the control signal that is sent to the control valve. Supply 25. In summary, of the three terms of the control law of the invention, first the predictive open loop signal is sent out to generate a signal of speed that absent of all the loads and disturbances, will produce a speed trace that corresponds to the desired speed profile of the users, as closely as possible as a result of the automatic calibration. Next, the term load memory or disturbance is added to compensate for changes in machine load and geometry of molding that naturally occur during the injection stroke. This now produces an output that compensates for any consistent load changes. Finally, the traditional PID loop then runs on any remaining errors (that is, the inconsistent perturbations that result from variations in charge load) using terms that were adjusted during the automatic calibration. The control system with all the terms implemented as described has produced remarkable improvement in speed control over all other known systems in use in injection molding machines including significant improvements in tracking error, rise time, over load and other measures of closed loop performance with injection reinforcement. In the preferred embodiment, the control system can be operated with all three control signal terms of the control law implemented as just described. Alternatively, the control system can be operated simply with closed loop, or open loop or open loop with adaptive control. This is easily achieved by simply adding to the desired speed control signal factors that constitute the control law defined above. It is illustrated schematically in Figure 5 by designated switches 98a, 98b and 98c that connect the control signals to the summing junction 76. A dotted line is illustrated to indicate that the switch 98c can optionally be activated only when the switch 98b is activated. Closing all the switches activates all the terms of the control law in such a way that the control operates as described. Closed-loop control occurs when closing switch 98a and opening switches 98b and 98c. Open loop predictive control without adaptive control occurs upon closing switch 98b and leaving switches 98a and 98c open while predictive open loop predictive control occurs with closed switches 98b and 98c and switch 98a open. Now with reference to Figure 7, the external control loops for the control system of the invention are illustrated which are not illustrated in Figure 6 for convenience of drawing. The calibration settings / routine described with respect to calibration steps in Figure 4 are made and stored in the calibration routine block 99. As illustrated, the position signal of Real-time ram on line 79 accesses the calibration routine block 99, so that the PID gain terms for those real-time signals can be adjusted by the tuned settings for the PID loops established in the calibration. Similarly, the real-time ram position signals are also fed to the disturbance search table 95, so that the error signal developed by the PID controller 70 can be stored in the proper ram position. Referring now to Figure 2, a general schematic of the machine controller 30 is illustrated. In the preferred embodiment, the controller 30 is the Pathfinder controller of the transferee, specifically the Pathfinder 3000 or 5000 series. The operator station 32 includes a smart operator station (UPC (CPU)) 100 board, with memory communicating with a display unit 101, a keyboard 102, where the operator feeds mold cycle instructions and a PCMCIA slot 103 which is typically used to extract mold cycle path data for SPC, storage and other purposes. As far as the invention is concerned, the data is fed either through the keyboard 102 or the PCMCIA slot 103 and the UPC in the operator station board 100 converts this data into machine instructions that are sent to PLC 34 for processing. In addition, the operator station board 100 receives PLC machine data 34 and outputs it to the PCMCIA slot 103 and / or to the display 101 through a video signal display card under control of the station. UPC operator. The video display 101 also exhibits power data as well as output data. As far as the injection stroke is concerned, the set points are typically supplied by the operator on the keypad 102 and converted into machine signals which are sent to PLC 34 and are also sent to the signal display board of video within the operator station board 100 where the data provided is itrated in the display 101. Of routines stored in the operator station memory, the operator station 32 can construct the visual operator of the speed profile 40 shown in FIG. Figures 3 and 3A as well as other information displayed. Alternatively, the velocity profile 40 can be constructed from PLC 34 and fed back as machine signals to be processed by the video signal display card. However, the current detected speed shown as a superimposed stroke in the speed profile 40, itrated in the display unit 101, is generates from machine signals that are fed to the video card from PLC 34. The machine cycle signals are processed through PLC 34 which contains a number of boards, of which each is intelligent and transports your own memory and UPC (s). In general, PLC 34 sends digital output signals on line 105 and analog output command signals on line 106 (such as the control signal on line 37 to supply valve 25), to a number of devices of output such as valves, motors, pumps, solenoids, etc., which are shown by the block of machine elements 107. Machine elements such as the piston 16 and the process detectors (such as the water path position detector 35) also function as feedback detectors shown by block 108, which can develop either digital feedback signals on line 109 or analog feedback signals on line 110 (i.e. the line 35a in Figure 1) to the PLC 34. The main boards within the PLC 74 comprise an output / input board 112 communicating with the operator station board 100 through a serial link and with other boards PLC 74 system through a VME 13 duct that in turn transports or accesses an overall memory storage 114. Other boards include an analog processor board 115 and a temperature processing board 116, both of which interconnect with an analog terminal board 117 that receives analog feedback feeds and generates analog outputs. PLC 34 also includes the communication board 118 that carries a processor allowing communication via SPI protocol with auxiliary devices connected to the injection molding machine such as robotic handling mechanisms. PLC 34 also includes the logic sequencer board 120 which not only communicates with the VME duct but also with a digital output board 121 and a digital power board 122 through an input / output duct 124. Also, the logic sequencer 120 has a high speed link (not shown) connected to the analog board 115 for fast transmission of analog data processed simultaneously by the analog board 115. Reference should be made to US patents previous 5,493,503 and 5,456,870 which are incorporated by reference herein for a more detailed description and specification of the operation of PLC 34 than that with which it will be described or set forth herein. For purposes of this invention, the sequencing card 120 is the programmable controller that contains the graphic controls or programmable routines that They control the injection molding site of the injection molding machine. Basically, the sequencer card 120 uses a quantity of data generated by the other PLC boards including the user instructions face-to-face the operator station 32, and performs a series of logical instructions that: i) determine the value of certain detector feeds, ii) perform logic and numerical calculations based on detector information, which may be time or account dependent and iii) determine certain output signals based on the detector feeds that control the molding cycle. This user-defined program executed by the sequencer board 120 is reviewed or digitized periodically, so that changes can be made to the output command signals and the molding cycle can be sequenced through its normal event progression. As illustrated in Figure 9, the sequencer card 120 includes a CPU 130, which in the preferred embodiment is a Motorola 68000 16/32 operator operating at a frequency of 12 MHz. The sequencer card 120 also includes the RAM memory 131 which comprises memory storage, power and output tables, etc., of data or signals that are received from all other cards and overall memory 114 through the VME 113 pipeline or the link high-speed 135 (not shown in Figure 2) face-to-face a programmable peripheral interface 133 and ROM 132 containing the execute program language stored in an execute logical program, which in the preferred embodiment is written to the Motorola 68000 Assembly language. While some steps of the automatic valve calibration procedure described with reference to Figure 4 are described in the sequence language "statement list" stored in the logical program of ROM 132, for purposes of describing the present invention, it can be considered that the sequencing card 120 simply institutes the start and end of the injection stroke while performing its scanning of the program that controls the molding cycle performed by the injection molding machine. Control of ram position and ram velocity can be considered to be performed by the analog card 115 (although those skilled in the art will readily recognize that some functions can be performed by other cards within the PLC 34). The development of the current control signals is of course caused by program routines or software (software), which apart from the general formulation of the control law defined above, no, per se, they are part of the invention, although the invention must process the program to work. That is, programs per se are well known or easily apparent to those with skill in the specialty and can be generated by any right-handed programmer, given the functional parameters to be performed by the program as set forth herein. Accordingly, the programs themselves will not be described in detail here. The analog card 115, like the sequencer card 120, has a UPC (CPU) 140, random access memory RAM 139 and read-only memory ROM 140. The UPC (CPU) 140 in the preferred embodiment is a 16 / processor. 32-bit Motorola 68000 operating at 12 MHz frequency. The ROM 140 contains the program routines, the instructions to execute and the search, power and output tables. The RAM 139 contains the data calculated from routines that are stored in the ROM which in turn can be stored in appropriate tables. Data from the analog to digital converter 78 or from the VME 113 duct, or the high-speed link 135, is accessed through the programmable peripheral interface of the analog card 141. The data is output via the digital-to-analog converter 67 or to the VME 113 duct or the high speed link 135 through the interface programmable peripheral 141. The high speed link 135 is not used for speed profiling signals. Illustrated as separate blocks in the analog card 115, the calibration block 144, the adjustment signal block 145 and the control law block 146. Each block 144, 145 and 146 contains R / AM memory and ROM with processed routines by the UPC 138. The calibration block 144 stores established data from blocks 57, 58 and 61, draws a straight line speed / gain through the data and then calculates the gain signal for any desired speed in the profile of speed 40 which is employed by the control law block 146 to develop the control signal that is outputted to the supply valve 25. Similarly, the stored values of block 63 in Figure 3 are used to adjust gain values for the proportional, derivative and integral loops of the PID controller 70, depending on the actual speed signals detected by the control system. The gain terms for the PID controller 70 are sent similarly to the control law block 146 as well as to the PID controller 70 via the D / A converter 77. The adjustment signal block 145 accesses through the peripheral interface programmable 141, the overall memory 114 having in the form of machine language, the user set points which are fed into the operator station 32 processed or sent by the output / input card 112 to the global memory 114. A routine in the setting signal block 145 generates the one thousand set point signals approximately from the user's settings points. As already discussed, adjacent user adjustment points are interconnected as straight lines and the slope of the straight line is used to assign setpoint signals to discrete path increments. The signals set points are then stored in an output table to be used by the control law block as required. Alternatively, the adjustment signal block 145 may have a map of speed signals stored in the form of a look-up table that are accessed to form the output table dependent on the selected user set points. Still further, the calibration block 145 can be accessed to modify the output table by the valve gain settings, which are developed for each set point signal in the output table. The control law block 146 contains routines to perform the calculations for the three terms of the control law as defined above. The terms that are processed are triggered or activated by the user selecting open loop, closed loop or loop open / closed as defined above. The user settings for the control law block 146 are similarly stored in the global memory 114 and accessed by the control law block 146. The use of finite impulse response filters is particularly important for implementation of the control law . As noted in the discussion with respect to Figure 5, FIR 80 is used to develop the speed signal in real time and FIR 88 is used to develop the speed and predictive position signals from the position detector 35. The signal of ram travel position on line 35a, after passing through a smoothing filter (not shown), and after passing through a multiplexer (along with all other analog detector signals not shown or described herein) , because they are related to other operations performed by the machine) a digital signal is passed in the A / D converter 78 and the signal is stored in the buffer 150. The buffer 150 has a number of channels (eight in the preferred embodiment), with each channel storing information at synchronized intervals, face-to-face the control circuit 151 for one or more specified sensors. The signal for water path position detector 35 is stored in channel 150a at each time interval of 0.25. millisecond. Each channel stores 256 samples before sequentially writing the stored data. That is, sample 257 will write about the first sample. At every 0.75 millisecond, channel 150a is sampled. The current position signal and the three immediately preceding position signals (each of which precedes its "adjacent" signal in the time by 0.25 millisecond) are sampled, summed and averaged to obtain the real-time signal Xi. The channel 150a is accessed simultaneously to obtain adjacent signals stored 50 milliseconds prior, which are sampled, summed and averaged to obtain the past time signal x0. The "spacing" of time of 50 milliseconds, is collected as a compensation between the precision of the estimation of x 'and the frequency of updating the calculation of'. With the terms xx and x 'of the FIR equation explained, the advanced time or? T is now applied depending on the use of the finite impulse response filter. That is, there is a time delay to process the signal through PLC 34. This time delay is applied to correct the real-time speed signals sensed by the water path position detector 35. However, for the predictive detector signal, there is a delay time in the response of the supply valve 25 and thus, a set is made additional calculation for the valve delay. In addition, there is a delay time that is attributed to the moment of the ram and a third set of FIR equations are performed to take into account the delay attributed to the moment of the system. In this way, an FIR filter (PLC delay) is used to determine the speed signal in real time. Three FIR filters (PLC delay, valve delay and system delay) are used to determine the predictive position signal. In practice, the three filters are combined in a filter that performs all the mathematical functions in an operation. The FIR routines are performed (as a number of routines for other analog base machine functions) by the analog digitization control logic 154, which is conventional (like all the illustrated components) and includes standard UPC (CPU) instructions such how to move / add / multiply to select and send appropriate detector signals from the multiplexer (not shown) to the converter 78 for processing. The data is retrieved from the channel 150a by ASCL 154 which includes or has an associated offset and memory address generator, multiplexer channel counter, a digitizer status controller and synchronization circuit 151. The UPC (CPU) 138 performs the calculations of finite impulse response and in conjunction with the selection circuit device 155 controls the selection and return of data by ASCL 154 from buffer 150. In general summary of some of the main concepts employed in the present invention, first there is an automated calibration method that results in stored data, which are interpolated and calculate, to produce control signals, that can consistently and reliably produce open-loop control of water hammer speed. Next, the system uses a large plurality of machine-generated setpoint signals, from a limited number of user adjustment points to provide a large number of control signals to replicate almost any reachable velocity profile that is you want for the end user. Significantly, the large number of setpoint signals generated by the machine allow the use of predictive filters (and similar concepts) that in an important way select the speed signal required by the speed profile at the advanced time. This provides an open-loop signal that takes system delay into account in a manner far superior to that which would otherwise be achieved. For this predictive open control loop, a learned correction of adaptive load disturbance is applied. In this way, while the initial open loop signal is set based on an unloaded condition, the learned response attributed to the load results in a corrected open loop signal within a few cycles. Therefore, the machine can control, in open loop, any molded part given to velocity profiles that more closely follow the adjustment profile than the closed loop systems of the prior art. Finally, current load disturbances are detected and taken into account by the traditional and familiar PID controller. While the use of a PID controller to take charge disturbance into account is chosen in the preferred embodiment, due to its ability to be tuned by the automatic calibration method described and due to its familiarity with users of injection molding machines, other controls can be used instead of the PID control to take into account observed current load disturbances and produce the error signal used in the adaptive control. Specifically, the pressure detected by the ram can be used to take into account load disturbances. The pressure detected by the ram is a direct measure of the load resistance and can easily be compared to a reference pressure, the differential of which will provide the charge disturbance term to the control law instead of the PID controller or even as a additional term to join in the control law. In addition, water pressure readings can be compared to the pressure readings in the mold for further verification and control. The invention has been described with reference to a preferred embodiment. Modifications and alterations will occur to those with skill in the specialty when reading and understanding the detailed description of the invention herein established. For example, the invention has been described as using a finite impulse response filter that is sometimes otherwise known as an FIR filter, a transverse filter, a derivative delay line filter, or a moving average filter and At least one control author has identified all these filters as a non-recursive filter. In contrast to the filter described, those skilled in the art will recognize that other types of filters such as an infinite impulse response filter or other recursive type filters can be employed. It is intended to include all these modifications and alterations that fall within the scope of the present invention.

Claims (35)

  1. CLAIMS 1. A system for variably controlling the speed of the ram of an injection molding machine during the injection stroke of a molding cycle, characterized in that it comprises: means for selecting a plurality of user adjustment points specifying speeds of water hammer and adjusted ram travel positions, which will be achieved by the ram during the injection stroke; means for automatically establishing a second greater plurality of machine set point signal from the adjusted user points, indicative of adjusted ram speeds in adjusted ram travel positions and defining a site of adjusted speed and travel points that a In turn, they define a speed profile that the ram will emulate during the injection stroke, means for determining a load-compensated speed control signal for each signal of machine set points at a given time during the injection stroke, the means for determination include means for detecting the real-time position of the ram during the injection stroke and means for applying a calibrated factor set automatically to each set point signal, which causes the ram to move at approximately the speed of the signal Points of adjustment when the machine is not under load; means for applying the speed control signals with variable control signals to the machine, to cause the ram to move variably during the injection stroke at speeds corresponding to the velocity profile. The system according to claim 1, characterized in that the means for determining the speed control signals further include predictive means for choosing set point signals to advanced ram travel positions that have not been traversed at the time when the set point signal is chosen to process as the speed control signal, whereby the speed control signal takes into account the response latency of the machine, the selected set point signal occurring in the future it is a signal of predictive adjustment points. The system according to claim 1, characterized in that the means for determining the speed signal further include means for developing real-time water-speed signals from the real-time water-position signals and controlling means for develop error signals that modify the signals of real-time set points processed as speed control signals, the controlling means include means of feedback to compare the real-time ram speed signals with signals specific setpoints in the velocity profile at the real-time ram position, to produce a difference signal corresponding to an error signal thereby any setpoint signal selected for processing in a speed control signal is modified by the error signal to take into account the differences between the specified real-time water hammer speed and the real-time water hammer speed. . The system according to claim 2, characterized in that, the means for determining the speed control signals, further include disturbing means for modifying a predictive control signal by an error signal, the disturbing means include feedback means for comparing a real-time ram speed signal with a real time set point signal in the velocity profile to the real-time ram travel position, to produce a difference signal corresponding to an error signal and means of disturbance storage, to store each error signal produced in the adjusted ram travel positions during any given injection stroke, the disturbance means select a error signal stored in the position of ram travel for any given predictive control signal and modifying the predictive control signal by the selected error signal whereby the speed control signal takes into account perturbations attributed to the resisting fusion the movement of the ram. 5. The system according to claim 4, characterized in that the means for determining the speed signals, further include means for developing real-time water-speed signal from the real-time water-position signals and controlling means. to develop error signals that modify real-time setpoint signals processed as speed control signals, the controlling means includes feedback means for comparing real-time water velocity signals with signals specific setpoints in the velocity profile. speed in the real-time ram position, to produce a difference signal corresponding to an error signal, whereby a set point signal selected to be processed in a speed control signal is modified by the signal of error to take into account differences between the specified ram speed in real time and a water hammer speed detected in real time. 6. The system according to claim 1, characterized in that the machine includes valve means that control the water hammer speed and the means for applying the calibration factor include means for machine determination of control signals corresponding to displacement and extension adjustments. of the valve means. The system according to claim 1, characterized in that the plurality of speed setpoint signals number approximately one hundred signals separated by 2.54 cm (inch) of ram travel. The system according to claim 3, characterized in that the controlling means includes a PID controller and the means for applying the calibrated factor include machine adjustment control loops for the PID controller. The system according to claim 4, characterized in that the predictive means includes a finite impulse response cycle to determine the predictive position, in the future where the predictive signal is generated. 10. The system according to claim 3, characterized in that the means for generate real-time speed signals include a finite impulse response filter. The system according to claim 1, characterized in that the control signal that regulates the operation and closing of a hydraulic valve that controls the output of a hydraulic pump. The system according to claim 1, characterized in that the control signal comprises the speed command signal and a vector-controlled AC (AC) induction motor, which controls the speed and position of the ram. 13. System for variably controlling the speed of the ram of an injection molding machine during the injection stroke of a molding cycle, characterized in that it comprises: means for selecting a plurality of user adjustment points that specify, speeds of desired ram a in specified ram travel positions, means for automatically establishing the user in points, a second plurality of higher setpoint signals, indicative of a user adjustment velocity profile having a plurality of adjusted ram velocities in adjusted ram travel positions that the ram will emulate during the injection stroke; means for modifying the signals of selected set points of the speed profile at specified ram positions to produce speed control signals; the means for modifying include predictive means for selecting setpoint signals at advanced ram travel positions that have not been traversed at the time when the setpoint signal is chosen to process as the speed control signal, the signals of selected set points are predictive control signals; and means for applying the speed control signal as a control signal to the machine, to cause the ram to move in a variable manner during the injection stroke at speeds corresponding to the speed profile. The system according to claim 13, characterized in that the means for modifying further include means for generating water position signals indicative of the current water hammer position in real time during the injection stroke and means for generating speed signals in Real time of water from the water hammer position signals. The system according to claim 14, characterized in that the means for modifying further include control means for developing error signals that modify the set point signals, the controlling means include means of feedback to compare the real-time ram velocity signals with specific setpoint signals in the velocity profile at the ram travel position in real time to produce a difference signal corresponding to an error signal, with what any signal of set points selected to process in a control signal is modified by the error signal to take into account the differences between a specified water hammer speed and a water hammer speed detected in real time. The system according to claim 15, characterized in that the means for modifying include disturbing means for modifying a predictive control signal by the error signal, the disturbing means include the feedback means and means for stopping disturbance for storing each error signal that occurs at adjusted ram travel positions during an injection stroke, the disturbance means selects an error signal stored at the ram travel position for any given predictive control signal and modifies the control signal predictive by the selected error signal so that the speed control signal takes into account disturbances attributed to the fusion that resists the movement of the ram. 17. . The system according to claim 16, characterized in that it also includes means for automatically adjusting the set point signals by a calibrated factor such that any set point signal set at a certain speed will cause the ram to move approximately at the same speed when the machine is not under load. The system according to claim 17, characterized in that the machine includes valve means controlling the ram velocity, and the means for applying the calibrated factor include means for machine determination of control signals corresponding to displacement and adjustments. of extension of the valve means. The system according to claim 17, characterized in that the controlling means includes a PID controller and the means for applying the calibrated factor include machine adjustment control loops for the PID controller. The system according to claim 19, characterized in that the predictive means includes a finite impulse response filter to determine the predictive position in the future where the predictive signal is generated. 21. The system according to claim 20, characterized in that the plurality of speed setpoint signals number approximately one hundred signals separated by 2.54 cm (inch) of ram travel. 22. Method for controlling the speed of a ram in an injection molding machine during the injection step of the molding cycle after the operator has adjusted a plurality of speed adjustment points, the method is characterized in that it comprises the stages of: a) generating in machine a second larger plurality of set point signals from the operator set points, each set point signal, is indicative of a set ram speed in a water ram position set in a sequence that defines a speed profile, the ram is to emulate during the injection stage; b) Automatically adjust the signals of selected set points to process by a machine-generated calibrated factor, such that any signal of selected set points of a given speed will cause the ram to move at approximately the same speed when the machine it is not under load; c) modify the signals set-up points to take into account the load imposed on the machine by the molding material injected to produce speed control signals at adjusted ram positions corresponding to the velocity profile; and d) applying the control signals as a control signal to the machine to regulate the speed of the ram. 23. The method according to claim 22, characterized in that the machine is hydraulically moved and the calibrated factor is determined by performing the following steps: i) initializing the machine in such a way that a) the translational axis of the ram is adjusted, b) the water hammer detector of the machine is adjusted to record the ram speed and c) voltage signals that are fed to the hydraulic supply valve that controls the ram speed are adjusted to correspond to digitized signals; ii) increasing the voltage output to the valve until the water movement detected by the machine occurs and the machine records the output as a digitized valve displacement signal; iii) increase the voltage output beyond the displacement and record the water rates detected by each increment; iv) and then register in machine the voltage output level where the nominal maximum ram speed is detected as a digitized valve extension signal. 24. The method according to claim 22, characterized in that the modification step further includes the step of generating water position signals indicative of the current position of the ram in real time during the injection stroke and selecting signals setpoints in the profile of speed in advanced ram travel positions that have not been traversed at the time when the set point signal is chosen for application as the control signal, the set point signal is selected from the velocity profile in the advanced position is an open-loop predictive control signal. The method according to claim 22, characterized in that the modification step further includes the step of generating water position signals indicative of the current position of the ram in real time during the injection stroke and generating the speed signals of Real-time ram from the ram position signals, select setpoint signals in the velocity profile in the ram position in which the ram position signals are detected and compare the ram velocity signals in real time with selected setpoint signals in the detected water position to produce an error signal and modify the signals of selected set points to process in the stage (b) as the control signal, with the error signal in step (c) to produce the control signal in the stage (d) as a closed loop control signal. 26. The method according to claim 24, characterized in that the modification step further includes the steps of generating real-time ram velocity signals that correspond to the ram positions in real time; compare the ram speed signals in real time with the set point signals in the velocity profile at the detected water hammer positions, to generate an error signal; storing the error signals in the detected real-time ram positions and modifying the predictive control signals by the error signal stored in the advanced path positions of the predictive control signal to produce the control signal as a balanced load signal open loop. 27. The method according to claim 26, characterized in that the modification step further includes the steps of modifying select setpoint signals in the velocity profile at the real-time water hammer positions detected, wherein the velocity signal in real time is generated by the error signal and add the control signal compensated with error in the real-time ram position, with the predictive control signal modified by the stored error signal to produce the control signal as a signal compensated by open loop and combined closed loop load. 28. The method according to claim 22, characterized in that the set point signals in the velocity profile number approximately 100 equally spaced signals per 2.54 cm (inch) of ram travel. 29. The method according to claim 25, characterized in that the step of comparing is achieved through a PID controller. 30. The method according to claim 24, characterized in that the advanced ram travel position is determined by a finite impulse response filter. 31. The method according to claim 25, characterized in that the real-time ram velocity signal is determined by a finite impulse response filter. 32. The method according to claim 29, characterized in that the calibration step further includes the additional steps of v) recording the time it takes for the ram to reach at least a set or determined percentage of maximum ram velocity and use the time to reach the adjusted percentage to tune the PID controller to the machine. 33. The method according to claim 26, characterized in that only an adjusted percentage of the error is stored, each current molding cycle adjusts the predictive control signal by the error signal stored in the advanced position of the immediately preceding molding cycle, while writing over the stored errors of the preceding cycle with adjusted percentages of real time real-time error signals. 34. The method according to claim 31, characterized in that the overwriting or writing step proceeds only for a limited number of molding cycles. 35. The method according to claim 34, characterized in that the overwriting step factors the error signals generated during the current cycle with stored errors.
MXPA/A/1999/003771A 1998-04-23 1999-04-23 Molding machine by injection, controlled by adaptive process, auto-adjust MXPA99003771A (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US09065399 1998-04-23

Publications (1)

Publication Number Publication Date
MXPA99003771A true MXPA99003771A (en) 2000-08-01

Family

ID=

Similar Documents

Publication Publication Date Title
CA2262431C (en) Auto-tuned, adaptive process controlled, injection molding machine
EP1297941B1 (en) Model predictive controll apparatus and methods for motion and/or pressure control of injection molding machines
Chen et al. Quick testing and modeling of thermally-induced errors of CNC machine tools
US4094940A (en) Injection molding machine controls
US6192299B1 (en) Method of measuring operation characteristic of proportional electromagnetic control valve, method of controlling operation of hydraulic cylinder, and method of modifying operation characteristic of proportional electromagnetic control valve
CN100515725C (en) Controller for injection molding machine
US5062785A (en) Injection molding controller with process variable learning
CN105843167B (en) The thermal displacement correction device of lathe
US5256345A (en) Injection control method for injection molder
US5482662A (en) Control method for injection molding machines
EP0204277B1 (en) Power transmission
Koch et al. Observer-based sliding mode control of hydraulic cylinders in the presence of unknown load forces.
CN110161973A (en) Model predictive control apparatus, control method, message handling program and recording medium
US5260008A (en) Resin property detection method and injection control method for injection molder
MXPA99003771A (en) Molding machine by injection, controlled by adaptive process, auto-adjust
JP2628266B2 (en) Speed control method and apparatus for injection molding machine
EP0274046A1 (en) Power servo system
JP3761429B2 (en) Resin evaluation method and apparatus using injection molding machine
CN106773698A (en) A kind of beverage filling control method based on on-line parameter identification
JP2001314954A (en) Injection control method for die casting machine
JP3280792B2 (en) Reservoir internal pressure adjusting device in injection molding machine
JP3466772B2 (en) Injection pressure control method for injection molding machine
JPH0124055B2 (en)
JPS604017A (en) Setting of operating condition of injection molding machine
Elser et al. Implementation of a model based numerical control for the gravity die casting process