CN115254529B - Dispensing compensation method and system based on time and pressure - Google Patents
Dispensing compensation method and system based on time and pressure Download PDFInfo
- Publication number
- CN115254529B CN115254529B CN202210774244.XA CN202210774244A CN115254529B CN 115254529 B CN115254529 B CN 115254529B CN 202210774244 A CN202210774244 A CN 202210774244A CN 115254529 B CN115254529 B CN 115254529B
- Authority
- CN
- China
- Prior art keywords
- dispensing
- time
- neural network
- air pressure
- value
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000004044 response Effects 0.000 claims abstract description 57
- 239000003292 glue Substances 0.000 claims description 80
- 238000013528 artificial neural network Methods 0.000 claims description 78
- 230000006870 function Effects 0.000 claims description 31
- 230000004913 activation Effects 0.000 claims description 11
- 239000002131 composite material Substances 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 4
- 230000001276 controlling effect Effects 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000002474 experimental method Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000000084 colloidal system Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05C—APPARATUS FOR APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05C5/00—Apparatus in which liquid or other fluent material is projected, poured or allowed to flow on to the surface of the work
- B05C5/02—Apparatus in which liquid or other fluent material is projected, poured or allowed to flow on to the surface of the work the liquid or other fluent material being discharged through an outlet orifice by pressure, e.g. from an outlet device in contact or almost in contact, with the work
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05C—APPARATUS FOR APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
- B05C11/00—Component parts, details or accessories not specifically provided for in groups B05C1/00 - B05C9/00
- B05C11/10—Storage, supply or control of liquid or other fluent material; Recovery of excess liquid or other fluent material
- B05C11/1002—Means for controlling supply, i.e. flow or pressure, of liquid or other fluent material to the applying apparatus, e.g. valves
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention belongs to the technical field of dispensing quantity control, and particularly relates to a dispensing compensation method and system based on time and pressure, wherein the method comprises the following steps: collecting dispensing time, a response air pressure value at an outlet of the electromagnetic valve and an air pressure value of an air source supplied to the electromagnetic valve; substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value; summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time; and controlling the dispensing device to dispense according to the output actual dispensing time. The invention can calculate the dispensing time compensation value under the current condition in real time, ensures the stability of the dispensing quantity, has simple steps and is easy to realize, and is convenient to apply in engineering.
Description
Technical Field
The invention belongs to the technical field of dispensing quantity control, and particularly relates to a dispensing compensation method and system based on time and pressure.
Background
The whole structure of the time-pressure type dispensing system mainly comprises an air source, an electromagnetic valve, an air pipe, a rubber storage pipe and a needle head. Glue in the glue storage tube is extruded out through compressed gas and coated on the surface of a workpiece, and the quantity of glue is regulated by regulating the opening and closing time of the electromagnetic valve, so that the glue storage tube has the characteristics of simple structure and convenience in equipment maintenance. However, the method has the defects that the high precision and the high stability of the glue outlet amount are difficult to ensure, and the glue outlet amount is gradually reduced under the same glue dispensing air pressure and time conditions as the glue remaining amount in the glue storage tube is reduced in the glue dispensing process and is influenced by the change of the compressibility of gas in the tube. In order to solve the problem, a plurality of related researches have been carried out for a long time, and the main research direction is to model the glue dispensing process, and compensate the actual glue dispensing time each time by matching with an automatic control algorithm, so as to ensure that the glue dispensing amount is stable in the initial state when the glue storage tube is filled with glue.
In the current modeling mode, the most common is the traditional mechanism modeling aiming at a pneumatic system, and a great deal of researches are carried out on analyzing and modeling the fluid characteristics of electromagnetic valves, pneumatic systems, gas flow, newtonian fluid and non-Newtonian fluid. Meanwhile, the conditions affecting the dispensing quantity are analyzed, and the method mainly comprises the following steps: time and pressure, residual glue height, needle head shape, viscosity and temperature, matching of dispensing height and needle inner diameter, matching of dispensing height and glue distributing speed, magnitude and corresponding speed of negative pressure source air pressure, whether air bubbles are contained, wherein the influence on the glue dispensing amount is the greatest or the glue amount in the glue storage tube is reduced. However, because the whole pneumatic system is complex, conditions are often set during research, and the model is simplified, the model established at present is difficult to apply to the actual production process, and even only stays in the simulation stage. In the aspect of compensation control, the method is mainly improved and innovated based on the traditional automatic control principle, and the main method comprises the steps of establishing a control model based on empirical data; establishing a control model based on offline batch control; and predicting the weight of the glue point through the established model, and performing PID control by taking the predicted value as a feedback quantity. These methods all have the problem of limited model fitting capability, and are difficult to be practically applied.
Disclosure of Invention
The technical problems to be solved by the invention are as follows: in the prior art, when the glue dispenser is actually applied, the glue storage tube is affected by the change of the compressibility of the gas in the tube along with the reduction of the glue remaining amount in the glue storage tube in the glue dispensing process, and the glue dispensing amount under the same glue dispensing air pressure and glue dispensing time conditions can be gradually reduced, so that the glue dispensing fruits are uneven.
The technical scheme adopted for solving the technical problems is as follows: a dispensing compensation method based on time and pressure comprises the following steps:
s1, collecting dispensing time, a response air pressure value at an outlet of an electromagnetic valve and an air pressure value of an air source supplied to the electromagnetic valve;
s2, substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value;
s3, summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time;
s4, controlling the dispensing device to dispense according to the actual dispensing time output in the S3.
Further, in S1, the response air pressure value at the outlet of the electromagnetic valve is obtained by sampling an air pressure sensor placed at the outlet of the electromagnetic valve at preset time intervals within a preset time period.
Further, the preset time period is 45ms before the dispensing start time, and the preset time interval is 3ms.
Further, in S2, the preset dispensing compensation model is obtained by performing nonlinear system fitting on a composite neural network; the composite neural network comprises a circulating neural network module and a fully-connected neural network module;
the circulating neural network module is used for taking a response voltage value and corresponding time at the outlet of the electromagnetic valve as response air pressure sequence data, extracting time sequence characteristics of the response air pressure sequence data through a four-layer neural network, and obtaining an output vector; the output vector of the circulating neural network module, the air pressure value of the air source and the dispensing time are spliced into a new vector which is used as the input of the fully connected neural network module;
and the full-connection neural network module predicts the input new vector through a seven-layer full-connection neural network and outputs a dispensing time compensation value.
Further, the activation function in the neural network in the recurrent neural network module adopts a tanh function.
Further, the activation function of the fully-connected neural network in the fully-connected neural network module adopts a Relu function.
A time and pressure based dispensing compensation system comprising:
the collecting module is used for collecting the dispensing time, the response air pressure value at the outlet of the electromagnetic valve and the air pressure value of an air source for supplying the electromagnetic valve;
the modeling module is used for establishing a preset dispensing compensation model;
the storage module is used for storing the dispensing compensation model established in the modeling module;
the processing module is used for substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value;
the calculation module is used for summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time;
the control module is used for controlling the dispensing device to dispense according to the actual dispensing time obtained by the calculation module.
A network side server, comprising:
at least one processor; the method comprises the steps of,
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the time and pressure based dispensing compensation method of any of the above.
A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the time and pressure based dispensing compensation method of any of the above.
The beneficial effects of the invention are as follows: according to the time and pressure-based dispensing compensation method and system, the dispensing time, the response air pressure value at the outlet of the electromagnetic valve and the air pressure value of an air source supplied to the electromagnetic valve are collected; substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value; summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time; and controlling the dispensing device to dispense according to the output actual dispensing time. The invention can calculate the dispensing time compensation value under the current condition in real time, ensures the stability of the dispensing quantity, has simple steps and is easy to realize, and is convenient to apply in engineering.
Drawings
The invention is further described below with reference to the drawings and examples.
FIG. 1 is a flow chart of a time and pressure based dispensing compensation method according to a first embodiment of the present invention;
FIG. 2 is a graph showing experimental performance of the dispensing compensation method based on time and pressure according to the first embodiment of the present invention;
FIG. 3 is a block diagram of a time and pressure based dispensing compensation system provided in accordance with a second embodiment of the present invention;
FIG. 4 is a schematic structural diagram of a dispensing compensation system based on time and pressure according to a second embodiment of the present invention;
fig. 5 is a schematic structural diagram of a network-side server according to a third embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The first embodiment of the invention relates to a dispensing compensation method based on time and pressure, which comprises the following steps of S1, collecting dispensing time, a response air pressure value at an outlet of an electromagnetic valve and an air pressure value of an air source for supplying the electromagnetic valve; s2, substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value; s3, summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time; s4, controlling the dispensing device to dispense according to the actual dispensing time output in the S3. The method can calculate the dispensing time compensation value under the current condition in real time, ensures the stability of the dispensing quantity, has simple steps and is easy to realize, and is convenient to apply in engineering.
The implementation details of the time and pressure based dispensing compensation method of the present embodiment are specifically described below, and the following is only implementation details provided for easy understanding, but not essential to implementation of the present embodiment, and a specific flow of the present embodiment is shown in fig. 1, where the present embodiment is applied to a server side on a network side.
Step S1, collecting dispensing time, a response air pressure value at an outlet of an electromagnetic valve and an air pressure value of an air source supplied to the electromagnetic valve;
specifically, the response air pressure value at the outlet of the electromagnetic valve is obtained by sampling an air pressure sensor arranged at the outlet of the electromagnetic valve at preset time intervals within a preset time period. The preset time period is 45ms before the dispensing starting time, and the preset time interval is 15 sampling points of 3ms. Meanwhile, as a supplement, the known air pressure value P of an air source supplied to the electromagnetic valve and the dispensing time T are designed as collected data, and the collected data are in the format of: [ d0, d1 … … d14, P, T ], wherein d0 is the response air pressure value at the first 45ms of the dispensing start time, and d1 is the response air pressure value at the first 42ms of the dispensing start time.
And S2, substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value.
Specifically, the preset dispensing compensation model is obtained by performing nonlinear system fitting on a composite neural network; the composite neural network comprises a circulating neural network module and a fully-connected neural network module;
the circulating neural network module is used for taking a response voltage value and corresponding time at the outlet of the electromagnetic valve as response air pressure sequence data, extracting time sequence characteristics of the response air pressure sequence data through the four-layer neural network, and obtaining an output vector; the output vector of the circulating neural network module, the air pressure value of the air source and the dispensing time are spliced into a new vector which is used as the input of the fully connected neural network module;
and the full-connection neural network module predicts the input new vector through a seven-layer full-connection neural network and outputs a dispensing time compensation value.
Specifically, the training method of the dispensing compensation model comprises the following steps:
in step S21, a test data set is established, the sampling time of the response air pressure value is the first 45ms of the dispensing start time, the sampling interval is 3ms, and the number of sampling points is 15. Meanwhile, as a supplement, the known air source pressure P and dispensing time T are also taken as input values. The format of the input data is: input= [ d0, d1 … … d14, P, T ]. The output value of the dispensing compensation model is a dispensing time compensation value deltat, and the unit is milliseconds. The output data format is: [ d0, d1, d2 … … d14, P, T, |Deltat ]. Wherein d0 is the response air pressure value at the first 45ms of the dispensing start time, and d1 is the response air pressure value at the first 42ms of the dispensing start time.
The glue dispensing device comprises an air source, an electromagnetic valve, an air pressure sensor, an air pipe, a glue storage pipe and a control board, and under the condition that the volume of the air pipe, the volume of the glue storage pipe and the ambient temperature are kept unchanged, the glue outlet quantity is mainly related to an air source air pressure value, glue allowance and glue dispensing time, wherein the air source air pressure value and the glue allowance in the glue storage pipe can influence a response curve of response air pressure at an outlet of the electromagnetic valve when dispensing, and therefore, input and output values under different air source air pressures and initial glue starting times are collected as sample data when training set data.
Specifically, in the range of the maximum and minimum dispensing air source air pressure, selecting a plurality of groups of air source air pressure conditions at certain intervals, and respectively collecting experimental data under the selected dispensing air source air pressure conditions; respectively carrying out 9 groups of glue dispensing experiments with the glue allowance of 10% -90% under the condition of air pressure of each air source; each group of dispensing experiments need to be used for testing the dispensing conditions under different dispensing time within the range of 50ms to 200 ms; each situation is repeated 15 times, and the response air pressure value, the air source air pressure value, the dispensing time and the weight of the dispensed glue, which are 45ms before the outlet of each experimental solenoid valve, are recorded, wherein the weight of each dispensing is too small to be weighed, so that the total weight of 15 dispensing glue is weighed, and the average value of each dispensing is calculated.
Taking the dispensing quantity when the residual quantity of the glue in the acquired data is 90% as the standard dispensing quantity under the conditions of the dispensing air pressure and the time to which the glue belongs. For the collected data, finding out the data which has the same dispensing air pressure and the same glue allowance but different glue allowance, and calculating the dispensing time difference value of the two pieces of data to be used as the compensation time delta t of the group of dispensing data.
Step S22, the data is normalized, so that the convergence rate of the network can be improved, and the adverse effect of the difference between the data can be eliminated. Respectively normalizing the air pressure sequence data, the air source air pressure data and the dispensing time in all sample data, wherein the normalization formula is as follows:
wherein x is max And x min Respectively, the maximum and minimum values of all sample data, x is the sample data that needs to be normalized, and this formula scales all sample data to within the range of 0-1.
In step S23, the dispensing compensation model is a deep neural network model combining a Recurrent Neural Network (RNN) and a Fully Connected Neural Network (FCNN). Since two types of data are contained in the input data: a time-series response air pressure value, a non-time-series air source air pressure value and an initial glue starting time. Inputting the acquired 15 response air pressure values into a four-layer cyclic neural network (RNN) for processing, extracting 15 response air pressure value time sequence characteristics, wherein each RNN layer comprises 15 nodes, and an activation function in the neural network in the cyclic neural network module adopts a tanh function, and a tanh (x) function formula is as follows;
the 15 response barometric pressure values are substituted into the activation function to obtain a first vector.
The first vector of the RNN layer, the air pressure value P of the air source and the initial glue time T are combined into a second vector which is 2 longer than the first vector output by the RNN layer.
Inputting the second vector into a seven-layer fully-connected neural network for prediction, wherein the number of nodes of the seven-layer fully-connected neural network is 16, 32, 64, 32, 16 and 1 in sequence, the output of the fully-connected neural network is a predicted current dispensing time compensation value, an activation function of the fully-connected neural network in the fully-connected neural network module adopts a Relu function, and the Relu (x) function formula is as follows:
Relu(x)＝max(0,x)；
in the training process, an average absolute error (MAE) is selected as an objective function, and an estimated value is compared with an observed time delay to calculate loss, so that counter propagation and weight updating are realized, and an average absolute error formula is as follows:
wherein O is t Is the actual value of the residual glue, P t Is the dispensing amount compensation value, and N is the sample number.
After the glue dispensing compensation model is trained, air source air pressure and initial glue dispensing time are set when the glue storage tube is full, so that the glue outlet amount reaches an expected value, when glue dispensing is carried out once, the response air pressure value of the first 45ms is sampled through the air pressure sensor arranged at the outlet of the electromagnetic valve, the response air pressure value, the air source air pressure value and the glue dispensing time are substituted into the trained glue dispensing compensation model, and the glue dispensing time compensation value is calculated.
In addition, since the influence of each dispensing action on the colloid allowance is small, the average value of the plurality of dispensing data can be combined into one calculation for improving the efficiency and reducing the calculated amount.
And S3, summing the dispensing time and the dispensing time compensation value to obtain the actual dispensing time.
Specifically, the actual dispensing time is equal to the sum of the dispensing time and the dispensing time compensation value, and the calculation formula is as follows:
T 0 ＝T+ΔT；
wherein T is 0 Is the actual dispensing time, T is the dispensing time, and DeltaT is the dispensing time compensation value.
S4, controlling the dispensing device to dispense according to the actual dispensing time output by the S3, and ensuring the stability of the dispensing amount every time.
The above steps of the methods are divided, for clarity of description, and may be combined into one step or split into multiple steps when implemented, so long as they include the same logic relationship, and they are all within the protection scope of this patent; it is within the scope of this patent to add insignificant modifications to the process or introduce insignificant designs, but not to alter its algorithm and the core design of the process.
The dispensing device comprises an air source, an electromagnetic valve, an air pressure sensor, an air pipe, a rubber storage pipe and a control board, wherein the volume of the air pipe is 28ml, the volume of the rubber storage pipe is 40ml and is kept unchanged, the ambient temperature is kept at about 20 ℃, and the glue used in the dispensing experiment is UV shadowless glue.
A group of training set data is acquired through experiments to carry out neural network model training, experimental data acquisition under 9 air pressure conditions with air pressure values P of 60kPa-300kPa are respectively carried out, nine groups of experimental data with glue allowance S of 10% -90% are respectively acquired under each air pressure condition, and each group of processed data comprises 81 data with initial glue time T of 50 ms-130 ms, wherein the specific conditions are as follows:
air source pressure value P: 60. 90, 120, 150, 180, 210, 240, 270, 300kPa;
initial glue time T: 50. 51, 52, … …, 128, 129, 130ms;
glue residual value S:10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%.
A total of 6561 sets of data were acquired. After the data is normalized, the data is divided into a training set and a testing set according to the proportion of eight to two. After training, the model predictive performance is verified with a test set.
The evaluation index of all models used Root Mean Square Error (RMSE) in addition to loss function MAE, the formula is as follows:
wherein O is t Is the actual value, P t Is the predicted value and N is the number of samples.
Test set performance is compared to table 1 below.
TABLE 1
The experimental results are shown in table 1 and fig. 2, comparing three experimental performances of the five methods, for verifying the performance advantages of the proposed rnn+fcnn structure over other neural network structures. Compared with an RNN model and an FCNN model of a single network structure, the RNN+FCNN model has obvious advantages in that the composite network structure can independently process input data with different properties, and characteristic information in the data can be fully extracted and utilized; whereas RNN networks are more suitable for processing data in the present invention than other commonly used variant networks, compared to the comparison surfaces of rnn+fcnn and lstm+fcnn models.
The dispensing compensation model adopts a composite neural network combining a circulating neural network module and a fully-connected neural network module, and has the characteristics of simple steps, simple system device, obvious effect, easy application and popularization in engineering and the like.
As shown in fig. 3, a second embodiment of the present invention relates to a dispensing compensation system based on time and pressure, comprising: the system comprises an acquisition module 201, a modeling module 202, a storage module 203, a processing module 204 and a control module 206.
Specifically, the collection module 201 is configured to collect the dispensing time, the response air pressure value at the outlet of the electromagnetic valve, and the air pressure value of the air source supplied to the electromagnetic valve; the modeling module 202 is configured to establish a preset dispensing compensation model; the storage module 203 is configured to store the dispensing compensation model established in the modeling module 202; the processing module 204 is configured to substitute the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculate to obtain a dispensing time compensation value; the calculating module 205 is configured to sum the dispensing time and the dispensing time compensation value to obtain an actual dispensing time; and the control module 206 is configured to control the dispensing device to perform dispensing according to the actual dispensing time obtained by the calculation module 205.
As shown in fig. 4, the collection module 201 collects the dispensing time, the response air pressure value at the outlet of the electromagnetic valve 2 at the first 45ms of the beginning time of the dispensing, and the air pressure value of the air source supplied to the electromagnetic valve 2 at the air inlet pipe 1; through the modeling module 202, a preset dispensing compensation model is established, and the preset dispensing compensation model is stored in the storage module 203; substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model by adopting a calculation module 205, and calculating to obtain a dispensing time compensation value; summing the dispensing time and the dispensing time compensation value through a calculation module 205 to obtain actual dispensing time; finally, the glue stored in the glue storage tube 3 is dispensed by adopting the control module 206 to control the glue outlet time of the glue outlet 4 to be the actual glue dispensing time, so that the dispensed glue 5 is placed on the glue plate 6, and the glue outlet amount of each glue 5 on the glue plate 6 is ensured to be stable.
It is to be noted that this embodiment is a system example corresponding to the first embodiment, and can be implemented in cooperation with the first embodiment. The related technical details mentioned in the first embodiment are still valid in this embodiment, and in order to reduce repetition, a detailed description is omitted here. Accordingly, the related art details mentioned in the present embodiment can also be applied to the first embodiment.
It should be noted that each module in this embodiment is a logic module, and in practical application, one logic unit may be one physical unit, or may be a part of one physical unit, or may be implemented by a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, units that are not so close to solving the technical problem presented by the present invention are not introduced in the present embodiment, but this does not indicate that other units are not present in the present embodiment.
The third embodiment of the present invention relates to a network server, as shown in fig. 5, including at least one processor 301; and a memory 302 communicatively coupled to the at least one processor 301; the memory 302 stores instructions executable by the at least one processor 301, the instructions being executable by the at least one processor 301 to enable the at least one processor 301 to perform the data processing method described above.
Where the memory 302 and the processor 301 are connected by a bus, the bus may comprise any number of interconnected buses and bridges, the buses connecting the various circuits of the one or more processors 301 and the memory 302 together. The bus may also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art, and therefore, will not be described any further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver may be one element or may be a plurality of elements, such as a plurality of receivers and transmitters, providing a means for communicating with various other apparatus over a transmission medium. The data processed by the processor 301 is transmitted over a wireless medium via an antenna, which further receives the data and transmits the data to the processor 301.
The processor 301 is responsible for managing the bus and general processing and may also provide various functions including timing, peripheral interfaces, voltage regulation, power management, and other control functions. And memory 302 may be used to store data used by processor 301 in performing operations.
A fourth embodiment of the present invention relates to a computer-readable storage medium storing a computer program. The computer program, when executed by the processor, implements the time and pressure based dispensing compensation method in the first embodiment.
That is, it will be understood by those skilled in the art that all or part of the steps in implementing the methods of the embodiments described above may be implemented by a program stored in a storage medium, where the program includes several instructions for causing a device (which may be a single-chip microcomputer, a chip or the like) or a processor (processor) to perform all or part of the steps in the methods of the embodiments described herein. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
While the foregoing is directed to the preferred embodiment of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow. The technical scope of the present invention is not limited to the description, but must be determined according to the scope of claims.
Claims (6)
1. The dispensing compensation method based on time and pressure is characterized by comprising the following steps of:
s1, collecting dispensing time, a response air pressure value at an outlet of an electromagnetic valve and an air pressure value of an air source supplied to the electromagnetic valve;
s2, substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value;
s3, summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time;
s4, controlling the dispensing device to dispense according to the actual dispensing time output in the S3;
specifically, the preset dispensing compensation model is obtained by performing nonlinear system fitting on a composite neural network; the composite neural network comprises a circulating neural network module and a fully-connected neural network module;
the circulating neural network module is used for taking a response voltage value and corresponding time at the outlet of the electromagnetic valve as response air pressure sequence data, extracting time sequence characteristics of the response air pressure sequence data through a four-layer neural network, and obtaining an output vector; the output vector of the circulating neural network module, the air pressure value of the air source and the dispensing time are spliced into a new vector which is used as the input of the fully connected neural network module;
the full-connection neural network module predicts the input new vector through a seven-layer full-connection neural network and outputs a dispensing time compensation value;
the dispensing compensation model is a deep neural network model combining a circulating neural network (RNN) and a fully-connected neural network (FCNN); since two types of data are contained in the input data: a time-series response air pressure value, a non-time-series air source air pressure value and an initial glue starting time; inputting the acquired 15 response air pressure values into a four-layer cyclic neural network (RNN) for processing, extracting 15 response air pressure value time sequence characteristics, wherein each RNN layer comprises 15 nodes, and an activation function in the neural network in the cyclic neural network module adopts a tanh function, and a tanh (x) function formula is as follows;
substituting 15 response air pressure values into the activation function to obtain a first vector;
forming a second vector by the first vector of the RNN layer, the air pressure value P of the air source and the initial glue starting time T, wherein the second vector is 2 longer than the first vector output by the RNN layer;
inputting the second vector into a seven-layer fully-connected neural network for prediction, wherein the number of nodes of the seven-layer fully-connected neural network is 16, 32, 64, 32, 16 and 1 in sequence, the output of the fully-connected neural network is a predicted current dispensing time compensation value, an activation function of the fully-connected neural network in the fully-connected neural network module adopts a Relu function, and the Relu (x) function formula is as follows:
Relu(x)＝max(0,x)；
in the training process, an average absolute error (MAE) is selected as an objective function, and an estimated value is compared with an observed time delay to calculate loss, so that counter propagation and weight updating are realized, and an average absolute error formula is as follows:
wherein O is t Is the actual value of the residual glue, P t Is the dispensing amount compensation value, and N is the sample number.
2. The method according to claim 1, wherein in S1, the response air pressure value at the outlet of the solenoid valve is obtained by sampling an air pressure sensor disposed at the outlet of the solenoid valve at preset time intervals within a preset time period.
3. The method for dispensing compensation based on time and pressure according to claim 2, wherein the preset time period is the first 45ms of the dispensing start time, and the preset time interval is 3ms.
4. Dispensing compensation system based on time and pressure, characterized by comprising:
the collecting module is used for collecting the dispensing time, the response air pressure value at the outlet of the electromagnetic valve and the air pressure value of an air source for supplying the electromagnetic valve;
the modeling module is used for establishing a preset dispensing compensation model, wherein the preset dispensing compensation model is obtained by performing nonlinear system fitting on a composite neural network; the composite neural network comprises a circulating neural network module and a fully-connected neural network module;
the circulating neural network module is used for taking a response voltage value and corresponding time at the outlet of the electromagnetic valve as response air pressure sequence data, extracting time sequence characteristics of the response air pressure sequence data through a four-layer neural network, and obtaining an output vector; the output vector of the circulating neural network module, the air pressure value of the air source and the dispensing time are spliced into a new vector which is used as the input of the fully connected neural network module;
the full-connection neural network module predicts the input new vector through a seven-layer full-connection neural network and outputs a dispensing time compensation value;
the dispensing compensation model is a deep neural network model combining a circulating neural network (RNN) and a fully-connected neural network (FCNN); since two types of data are contained in the input data: a time-series response air pressure value, a non-time-series air source air pressure value and an initial glue starting time; inputting the acquired 15 response air pressure values into a four-layer cyclic neural network (RNN) for processing, extracting 15 response air pressure value time sequence characteristics, wherein each RNN layer comprises 15 nodes, and an activation function in the neural network in the cyclic neural network module adopts a tanh function, and a tanh (x) function formula is as follows;
substituting 15 response air pressure values into the activation function to obtain a first vector;
forming a second vector by the first vector of the RNN layer, the air pressure value P of the air source and the initial glue starting time T, wherein the second vector is 2 longer than the first vector output by the RNN layer;
inputting the second vector into a seven-layer fully-connected neural network for prediction, wherein the number of nodes of the seven-layer fully-connected neural network is 16, 32, 64, 32, 16 and 1 in sequence, the output of the fully-connected neural network is a predicted current dispensing time compensation value, an activation function of the fully-connected neural network in the fully-connected neural network module adopts a Relu function, and the Relu (x) function formula is as follows:
Relu(x)＝max(0,x)；
in the training process, an average absolute error (MAE) is selected as an objective function, and an estimated value is compared with an observed time delay to calculate loss, so that counter propagation and weight updating are realized, and an average absolute error formula is as follows:
wherein O is t Is the actual value of the residual glue, P t Is the dispensing quantity compensation value, N is the sample quantity
The storage module is used for storing the dispensing compensation model established in the modeling module;
the processing module is used for substituting the response air pressure value, the air source air pressure value and the dispensing time into a preset dispensing compensation model, and calculating to obtain a dispensing time compensation value;
the calculation module is used for summing the dispensing time and the dispensing time compensation value to obtain actual dispensing time;
the control module is used for controlling the dispensing device to dispense according to the actual dispensing time obtained by the calculation module.
5. A network side server, comprising:
at least one processor; the method comprises the steps of,
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the time and pressure based dispensing compensation method of any one of claims 1 to 3.
6. A computer readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the time and pressure based dispensing compensation method of any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210774244.XA CN115254529B (en) | 2022-07-01 | 2022-07-01 | Dispensing compensation method and system based on time and pressure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210774244.XA CN115254529B (en) | 2022-07-01 | 2022-07-01 | Dispensing compensation method and system based on time and pressure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115254529A CN115254529A (en) | 2022-11-01 |
| CN115254529B true CN115254529B (en) | 2024-03-15 |
Family
ID=83763153
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210774244.XA Active CN115254529B (en) | 2022-07-01 | 2022-07-01 | Dispensing compensation method and system based on time and pressure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115254529B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4989756A (en) * | 1986-10-14 | 1991-02-05 | Kabushiki Kaisha Shinkawa | Dispensing apparatus |
| JPH07240427A (en) * | 1994-09-05 | 1995-09-12 | Toshiba Corp | Method and equipment for die bonding |
| JP2000005690A (en) * | 1998-06-23 | 2000-01-11 | Tdk Corp | Method for coating of adhesive |
| CN106423754A (en) * | 2016-10-20 | 2017-02-22 | 常州铭赛机器人科技股份有限公司 | Glue dispensing controller and glue outlet air pressure closed-loop control method thereof |
-
2022
- 2022-07-01 CN CN202210774244.XA patent/CN115254529B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4989756A (en) * | 1986-10-14 | 1991-02-05 | Kabushiki Kaisha Shinkawa | Dispensing apparatus |
| JPH07240427A (en) * | 1994-09-05 | 1995-09-12 | Toshiba Corp | Method and equipment for die bonding |
| JP2000005690A (en) * | 1998-06-23 | 2000-01-11 | Tdk Corp | Method for coating of adhesive |
| CN106423754A (en) * | 2016-10-20 | 2017-02-22 | 常州铭赛机器人科技股份有限公司 | Glue dispensing controller and glue outlet air pressure closed-loop control method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115254529A (en) | 2022-11-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103745273B (en) | Semiconductor fabrication process multi-performance prediction method | |
| CN108520155B (en) | Vehicle behavior simulation method based on neural network | |
| CN107274067B (en) | Distribution transformer overload risk assessment method | |
| CN113130014B (en) | Rare earth extraction simulation method and system based on multi-branch neural network | |
| CN112113146B (en) | Synchronous self-adaptive check method for roughness coefficient and node water demand of water supply pipe network pipeline | |
| CN109948863A (en) | Drainage pipeline networks inspection shaft liquid level prediction technique based on shot and long term memory models LSTM | |
| CN105867138A (en) | Stable platform control method and device based on PID controller | |
| CN114415520A (en) | Full-canal-system self-adaptive prediction control method for water-based network irrigation area | |
| CN115254529B (en) | Dispensing compensation method and system based on time and pressure | |
| CN116306221A (en) | Air pipeline system control method based on LSTM prediction model | |
| CN104881512A (en) | Particle swarm optimization-based automatic design method of ripple-free deadbeat controller | |
| CN114936742A (en) | Water supply system scheduling agent decision method | |
| CN109143853A (en) | Fractionating column liquid level self-adaptation control method during a kind of oil refining | |
| CN117313029A (en) | Multi-sensor data fusion method based on Kalman filtering parameter extraction and state updating | |
| CN106570786B (en) | Reservoir adaptive scheduling method based on scheduling rule time-varying characteristics | |
| CN104483832A (en) | Pneumatic proportional valve fuzzy sliding mode self-adaptive control method based on T-S model | |
| CN116415394A (en) | Two-loop equipment model simulation method integrating machine learning and mechanism model | |
| CN101598754B (en) | System and method for automatic virtual metrology | |
| CN106094524A (en) | The rapid model prediction control method compensated based on input trend | |
| CN116602286B (en) | Plant protection unmanned aerial vehicle spraying operation control system and method | |
| CN115329823A (en) | Dynamic depth migration prediction method and device for furnace tube temperature of delayed coking heating furnace | |
| CN108446506A (en) | A kind of uncertain system modeling method based on section Feedback Neural Network | |
| CN105259761A (en) | Continuous sterilization and repetition learning self-adaptive controller on the basis of characteristic model | |
| Lin et al. | Developing a novel Gaussian process model predictive controller to improve the energy efficiency and tracking accuracy of the pressure servo control system | |
| CN114139442A (en) | Method for optimizing steam turbine valve flow characteristic function based on K-means-LSTM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |