CN102637011B - Robust control method and system for directly driving numerical control platform based on coordinate transformation and parameter adjustment - Google Patents

Abstract

A robust control method for directly driving a numerical control platform based on coordinate transformation and parameter adjustment includes the steps: firstly, determining the initial phase of a motor rotor; secondly, performing coordinate transformation and parameter adjustment for contour errors to obtain position errors, namely calculating positional deviation and judging whether to adjust positions or not; and finally, executing robust control algorithm, outputting control quantity and driving the numerical control platform. A control system used in the method comprises a voltage regulation circuit, a rectifier filter unit, an IPM (intelligent power module) inverter unit, a DSP (digital signal processor), a Hall sensor, a grating bar, a current sampling circuit, a positioning sampling circuit and an IPM isolation driving protection circuit. A robust contour controller is used for directly driving the numerical control platform. Based on a complete coordinate transformation and parameter adjustment function, the robust contour controller can be applied to any smooth contour curves, device modeling errors and interference are avoided, and stability of the robust control system is improved.

Description

Based on coordinate transform and parameter adjustment, directly drive digital control platform robust control method and control system thereof

Technical field

The invention belongs to fields of numeric control technique, relate to and a kind ofly based on coordinate transform and parameter adjustment, directly drive digital control platform robust control method.

Background technology

Numerically-controlled machine to accurate, at a high speed, the future development of compound, intelligent, environmental protection.Accurate and High-speed machining is had higher requirement to transmission and control thereof, higher dynamic perfromance and control accuracy, higher speed of feed and acceleration, lower vibration noise and less wearing and tearing.Direct-drive feed system utilizes linear electric motors can directly produce the feature of linear thrust, and system no longer needs the intermediate transmission links such as screw-nut body in forming.The application on machine tool feed drives of linear electric motors and Drive Control Technique thereof, makes the drive mechanism of lathe occur great change, and makes machine tool capability have new leap.

The linear servo system of numerically-controlled machine adopts direct drive mode, has eliminated a series of harmful effects that mechanical motion mapping device brings, and therefore, has very significantly advantage in micro-feed servo system of high precision, fast-response.But this has also increased the difficulty of controlling.

Machine tool control system generally reduces tracking error with design single shaft feeding driving shaft.Yet, about processing, for single shaft feeding driving shaft quadrature error component, give the contour curve comparison-tracking error of design more important.Up to the present proposed various for reducing the control method of profile errors.Because each feeding driving shaft profile and tracking error are used for calculating control inputs in most methods, this may have the decline of profile tracking performance.Meanwhile, consider that adjustment brings difficulty to controller parameter in each feeding driving shaft profile and tracking error.

The contour outline control method of the twin shaft feed drive system based on coordinate conversion is converted into quadrature error component the tracking error of each driving shaft and for the tangent line error component of desirable contour curve.Because this method has obtained two separated single-input single-output systems about quadrature and tangential direction, the controller of each direction can be by independent design.Yet this simplifies controller parameter adjustment, stability of control system is guarantee when the speed Dynamic Matching of two feed shafts and desirable slope track only.A kind of controller, for desirable profile traces, decomposing profile errors is forward and tangential error component, and operation obtains the process that a kind of non-mistake is cut for machine tool, work coordinate system is defined within the assigned address of feed drive system accordingly, and control system power is rewritten according to this coordinate system.The work coordinate system axle of the quadrature of corresponding ideal contour curve and tangential direction and quadrature error component are construed to the approximate of Ditermination of cam profile.Although this method provides stability of control system to any contour curve, the difference between real profile error and quadrature error component may cause a great profile errors.

Summary of the invention

For the practical problems existing in existing control technology, the invention provides and a kind ofly based on coordinate transform and parameter adjustment, directly drive digital control platform robust control method, coordinate system conversion based on complete and the contouring control system of parameter adjustment design, and the stability of robust control system is provided for equipment modeling error and interference.

For the profile errors that the difference compensating between real profile error and quadrature error component causes, in a profile indexing means, quadrature error component is replaced by the error signal of a new definition, has proposed a robust contour method of controlling based on pusher sliding formwork.Yet, in design robust controller, required the second derivative of new definition error signal and the upper limit of interference derivative, comprise friction force.

The inventive method comprises based on coordinate system conversion with reference to method of adjustment and robust Controller Design, the adjusting of ride gain reduces more to the quadrature error component of ideal curve than tangent line error component, with reference to method of adjustment, the error convergence speed of new coordinate system is individually adjusted, reference signal adjustment as shown in Figure 2.Robust controller is considered to disturb and equipment modeling error, only has the nominal value of device parameter and the higher limit of interference magnitude to be required.

The control system that the inventive method adopts comprises voltage-regulating circuit, rectification filtering unit, IPM inversion unit, DSP, Hall element, grating scale, current sampling circuit, position sampling circuit, IPM isolation drive holding circuit.

Alternating voltage exports rectification filtering unit input end to, rectification filtering unit output terminal access IPM inversion unit, IPM is connected with motor, motor fuselage is equipped with grating scale, grating scale link position sample circuit input end, Hall element gathers motor current signal, export current sampling circuit to, current sampling circuit output terminal and position sampling circuit output end all access DSP, DSP outputs signal to voltage-regulating circuit input end and IPM isolation drive holding circuit, voltage-regulating circuit is adjusted alternating voltage, IPM isolation drive holding circuit access IPM inversion unit.The incremental optical-electricity encoder that is 400 lines by resolution when speed and position signalling detects, and it produces pulse signal A and B, delivers to the event capturing mouth of DSP, utilizes the tally function of catching mouthful unit to obtain the rotating speed of rotor, and position is by Z signal acquisition.

Of the present inventionly based on coordinate transform and parameter adjustment, directly drive digital control platform robust control method, concrete steps are as follows:

Step 1: the initial phase of determining electric mover;

By rotor position sample circuit and current sampling circuit, gather respectively electric mover position, speed and current information.

Step 2: profile errors is carried out to coordinate transform, and the line parameter adjustment of going forward side by side, obtains site error, carries out position deviation calculating, judges whether to carry out position adjustments, is carry out step 3, otherwise carries out electric current adjusting;

Step 2.1: profile errors is carried out to coordinate transform

Suppose:

(a1) ideal trajectory r _iwith its derivative

with

can obtain;

(a2) local coordinate system ∑ _linclination angle [theta] and its derivative

with

can obtain;

(a3) feed drive system position x and its derivative are measurable;

(a4), for 2 device parameter values, be all available;

According to above-mentioned hypothesis, determine following control, comprise input voltage v, matrix H, E and the I of driving shaft motor _e:

$H=\mathrm{diag}\left\{\frac{m_i+n_i{\left(\frac{2\mathrm{\π}}{p_i}\right)}^2}{\frac{2\mathrm{\π}}{p_i}}\right\},$

$E=\mathrm{diag}\left\{\frac{c_i+d_i{\left(\frac{2\mathrm{\π}}{p_i}\right)}^2}{\frac{2\mathrm{\π}{k}_i}{p_i}}\right\},$

$I_e=\left[\begin{array}{cc}0& 1\\ -1& 0\end{array}\right]---\left(1\right)$

M wherein _i(> 0), c _i(>=0) and f _irespectively driving shaft i table quality, viscous friction coefficient and driving force, wherein K _vand K _pspeed and position feedback gain matrix.They are only had the diagonal matrix of positive element by hypothesis.Equation is below set up:

Because matrix H and R are nonsingular, lower relation of plane need to be satisfied for realizing equation (2):

From equation above, by feedback gain matrix K in suitable distribution equations (3) _vand K _preach a conclusion, when t levels off to when infinite, e _llevel off to 0 is available.In addition controller ∑ in Fig. 1, _leach axle energy alignment error speed of convergence.If for e _l2feedback gain is set up and is greater than e _l1, for ideal curve, profile errors can be reduced sooner than tangent line tracking error.N _ifor driving the motor inertia of feed drive system; k _ifor torque-voltage transitions rate; Pi is the joint of ball-screw; Ki is torque-voltage transitions rate.

Step 2.2: parameter adjustment, specific as follows:

For reducing tangential error component e _l1with increase stability of control system, the controller little feedback gain that is supposed to, a large tracking error e _l1at quadrature error component e _l2with real profile error e _cbetween may produce large difference.In this case, even e _l2can be converged zero profile errors e _calso no longer suppressed.

This parameter regulation means, slightly mobile ideal position and local coordinate system, reduce the difference of quadrature error and real profile error. because a tracking error is along l ₁exist, in error e _l2and e _cbetween may there is a unacceptable difference.For reducing a scheme of this difference, be to physical location x, to estimate nearest position, r at desirable contour curve _arepresent.Yet, if desirable contour curve is complicated, be difficult to the nearest position of real-time accurate Calculation.And if desirable contour curve is recessed for x, several nearest positions to physical location x may exist.Even if nearest position estimated, the method proposing about proximal most position local coordinate system can not be applied in error, because along the tracking error of desirable contour curve approximate zero always, no longer may along the tracking of desirable contour curve.Like this, a local coordinate system is produced, and its direction is similar to the coordinate system of proximal most position, and its initial point approaches the ideal position of desirable contour curve, r _nand ∑ _n.With new definition, it is ∑ _nreplace rotation matrix R and tracking error e in equation (1) _wand e _l, produce e _l2and e _cdifference.The following describes generation is ∑ _nmethod.

Suppose along l ₁tracking error have a negative value (e for example _l1bear) because for reducing e _l1the controller gain distributing is normal than being e _l2that distributes is smaller.Also suppose ideal position r and along desirable contour curve closest approach r _abetween distance and tracking error e _l1big or small approximately equivalent.In addition, the ideal velocity along this section approaches constant.So, require by the time t of this section _dcan be estimated as follows:

Fig. 2 coordinate system ∑ _amiddle initial point (r _a) and inclination angle (θ _a), be the proximal most position to x at desirable contour curve, can be estimated as follows:

r _a＝r(t-t _d)，θ _a＝θ(t-t _d) (5)

Wherein r () and θ () show the function of time.The ideal position r revising _nin Fig. 2, be expressed as:

r _n＝r+R _ad _r，d _r＝[0，-d _ra2] ^T (6)

R wherein _ait is ∑ in Fig. 2 _ato ∑ _wrotation matrix, be θ _aa function.Control inputs equation (1) is replaced by following formula:

E wherein _wn=x-r _n.So, obtain control power below, e _n1and e _n2speed of convergence can be individually adjusted:

Step 3: carry out Robust Control Algorithm, output controlled quentity controlled variable v, drives digital control platform.

In actual processing, comprise that the interference of non-linear friction and cutting force and equipment modeling mistake exists really, therefore, controller equation should be expanded and consider to disturb and equipment modeling error.The feeding that contains BOUNDED DISTURBANCES vector w drives dynamics.

Controller is below proposed:

$\hat{H}=\mathrm{diag}\left\{\frac{{\hat{m}}_i+{\hat{n}}_i{\left(\frac{2\mathrm{\π}}{{\hat{p}}_i}\right)}^2}{\frac{2\mathrm{\π}{\hat{k}}_i}{{\hat{p}}_i}}\right\}$

$\hat{E}=\mathrm{diag}\left\{\frac{{\hat{c}}_i+{\hat{d}}_i{\left(\frac{2\mathrm{\π}}{{\hat{p}}_i}\right)}^2}{\frac{2\mathrm{\π}{\hat{k}}_i}{{\hat{p}}_i}}\right\}---\left(10\right)$

Wherein

the nominal value that represents z, signal

(wherein Λ is a diagonal matrix and the constant that only has positive element) is used the replacement as follow-up stability analysis

rate signal.Symbol δ v is an input vector realizing robust control.By approximate distribution equations feedback gain matrix K _v, K _pand Λ, when leveling off to when infinite, t can obtain e _nlevel off to zero.In addition, along ∑ in Fig. 2 _nthe error convergence speed of each axle can independently be adjusted.

Robust Control Algorithm Guarantee control system stability, even in the time of equipment modeling error and interference existence.Robust control obtains following equation:

$\mathrm{\δv}=-\frac{\mathrm{\η}{R}_a{e}_v}{\mathrm{\ρ}},$

Wherein || a|| is the Euclid norm of a, and ε is little normal number.

In order to show the robust stability of the system that proposes, Leah Pu Nuofu function candidate be employed.

E reaches a conclusion _vand e _nconsistent final bounded.So, from

also be consistent final bounded.Therefore,, although excessive value can reduce control performance, the controller of proposition can improve robust stability.

Beneficial effect: for direct driving digital control platform, proposed a robust profile controller.This method, based on complete coordinate system conversion and a parameter adjustment function, can be applied to any smooth contour curve, for equipment modeling error and interference, provides robust control system stability.In the inertia that the validity of robust contouring control system is promoted at feed drive system, considered large modeling error, the method for following application development can realize the accuracy of sub-micron, and expands to the cutting of 3-5 axle.

Accompanying drawing explanation

Fig. 1 is embodiment of the present invention control system block diagram;

Fig. 2 is embodiment of the present invention parameter adjustment schematic diagram;

Fig. 3 is for realizing embodiment of the present invention control system hardware block diagram;

Fig. 4 is embodiment of the present invention electric machine control system main circuit schematic diagram;

Fig. 5 is A of the present invention, B phase current sampling circuit schematic diagram;

Fig. 6 is embodiment of the present invention grating scale signal sample circuit schematic diagram;

Fig. 7 is embodiment of the present invention IPM hardware driving circuit theory diagrams;

Fig. 8 is vector control system program flow diagram in embodiment of the present invention control method;

Fig. 9 is that embodiment of the present invention control method position adjustments interrupts processing sub-control program process flow diagram.

Embodiment

Below in conjunction with accompanying drawing, the present invention will be further described.

The present invention is applied to the X-Y digital control platform of a linear drive motor, and the position of this platform is connected to the linear encoder of each driving shaft, and the sensor resolution of linear encoder is 0.1 micron.The speed of each driving shaft is by reverse poor the calculating of position measurement, and this sampling period is 2 milliseconds.

(1) system hardware structure

Based on coordinate transform and parameter adjustment, directly drive digital control platform robust control system to comprise voltage-regulating circuit, rectification filtering unit, IPM inversion unit, DSP, Hall element, grating scale, current sampling circuit, position sampling circuit, IPM isolation drive holding circuit.As shown in Figure 3, voltage-regulating circuit adopts reverse voltage regulating module EUV-25A-II to the structure of system, can realize 0～220V isolation pressure regulation.Rectification filtering unit adopts the uncontrollable rectification of bridge-type, and large capacitor filtering, coordinates suitable resistance capaciting absorpting circuit, can obtain the IPM required constant DC voltage of working.IPM adopts the 6MBP50RA060 of company of Fuji Intelligent Power Module, withstand voltage 600V, maximum current 50A, maximum operation frequency 20kHz.Four groups of independently 15V driving power power supplies for IPM.Primary power input terminal (P, N), lead-out terminal (U, V, W), main terminal is fixed with the screw carrying, and can realize current delivery.P, N are the primary power input terminal after the rectifying conversion smothing filtering of frequency converter, and P is anode, and N is negative terminal, and the three-phase alternating current of inverter output is connected to motor by lead-out terminal U, V, W.Control system main circuit schematic diagram as shown in Figure 4.

DSP selects TMS320F2812 processor, and its supporting development board comprises the outer 256K*16 position RAM of traget ROM, analog interface, eCAN interface, serial boot ROM, user lamp, reset circuit, the asynchronous serial port that can be configured to RS232/RS422/RS485, SPI synchronous serial interface and sheet.

In control system, current sample adopts the Hall current sensor LT58-S7 of LEM company.By two Hall current sensors, detect A, B phase current, obtain current signal, through current sampling circuit, convert the voltage signal of 0～3.3V to, finally by the A/D modular converter of TMS320LF2812, convert the binary number of 12 precision to, and be kept in numerical value register.The current sampling circuit of A, B phase as shown in Figure 5.Adjustable resistance VR2 conditioning signal amplitude, adjustable resistance VR1 conditioning signal side-play amount, by the adjusting to these two resistance, can adjust to 0～3.3V by signal, then is sent into AD0, the AD1 pin of DSP.Stabilivolt in figure is in order to prevent that the signal of sending into DSP from surpassing 3.3V, causes DSP to be damaged by high pressure.Operational amplifier adopts OP27, and power supply connects positive and negative 15V voltage, at the indirect decoupling capacitor of voltage and ground.Circuit input end connects capacitor filtering, to remove high-frequency signal, disturbs, and improves sampling precision.

A phase and the B phase pulse signal of grating scale output will be isolated signal by rapid light coupling 6N137, then through bleeder circuit, signal level are converted to 3.3V by 5V, are finally connected to two-way quadrature coding pulse interface QEP1 and the QEP2 of DSP.Circuit theory as shown in Figure 6.Fig. 7 provides the schematic diagram of Liao Liu road isolated drive circuit.It is to be noted IPM emergency protection signal pin right be non-repetition transient fault, in native system, be achieved by the following measures: the fault output signal of IPM is received DSP's by optocoupler

pin, DSP sets high resistance state by all task manager output pins in time when guaranteeing that IPM breaks down.

(2) of the present inventionly based on coordinate transform and parameter adjustment, directly drive digital control platform robust control method, concrete steps are as follows:

Step 1: the initial phase of determining electric mover;

By rotor position sample circuit and current sampling circuit, gather respectively electric mover position, speed and current information.

Step 2: profile errors is carried out to coordinate transform, and the line parameter adjustment of going forward side by side, obtains site error, carries out position deviation calculating, judges whether to carry out position adjustments, is carry out step 3, otherwise carries out electric current adjusting;

Step 2.1: profile errors is carried out to coordinate transform

Suppose:

(a1) ideal trajectory r _iwith its derivative

with

can obtain;

(a2) local coordinate system ∑ _linclination angle [theta] and its derivative

with

can obtain;

(a3) feed drive system position x and its derivative are measurable;

(a4), for 2 device parameter values, be all available;

According to above-mentioned hypothesis, determine following control, comprise input voltage v, matrix H, E and the I of driving shaft motor _e:

$H=\mathrm{diag}\left\{\frac{m_i+n_i{\left(\frac{2\mathrm{\π}}{p_i}\right)}^2}{\frac{2\mathrm{\π}}{p_i}}\right\},$

$E=\mathrm{diag}\left\{\frac{c_i+d_i{\left(\frac{2\mathrm{\π}}{p_i}\right)}^2}{\frac{2\mathrm{\π}{k}_i}{p_i}}\right\},$

$I_e=\left[\begin{array}{cc}0& 1\\ -1& 0\end{array}\right]---\left(1\right)$

M wherein _i(> 0), c _i(>=0) and f _irespectively driving shaft i table quality, viscous friction coefficient and driving force, wherein K _vand K _pspeed and position feedback gain matrix.They are only had the diagonal matrix of positive element by hypothesis.Equation is below set up:

Because matrix H and R are nonsingular, lower relation of plane need to be satisfied for realizing equation (2):

From equation above, by feedback gain matrix K in suitable distribution equations (3) _vand K _preach a conclusion, when t levels off to when infinite, e _llevel off to 0 is available.In addition controller ∑ in Fig. 1, _leach axle energy alignment error speed of convergence.If for e _l2feedback gain is set up and is greater than e _l1, for ideal curve, profile errors can be reduced sooner than tangent line tracking error.N _ifor driving the motor inertia of feed drive system; k _ifor torque-voltage transitions rate; Pi is the joint of ball-screw; Ki is torque-voltage transitions rate.

Step 2.2: parameter adjustment, specific as follows:

For reducing tangential error component e _l1with increase stability of control system, the controller little feedback gain that is supposed to, a large tracking error e _l1at quadrature error component e _l2with real profile error e _cbetween may produce large difference.In this case, even e _l2can be converged zero profile errors e _calso no longer suppressed.

This parameter regulation means, slightly mobile ideal position and local coordinate system, reduce the difference of quadrature error and real profile error. because a tracking error is along l ₁exist, in error e _l2and e _cbetween may there is a unacceptable difference.For reducing a scheme of this difference, be to physical location x, to estimate nearest position, r at desirable contour curve _arepresent.Yet, if desirable contour curve is complicated, be difficult to the nearest position of real-time accurate Calculation.And if desirable contour curve is recessed for x, several nearest positions to physical location x may exist.Even if nearest position estimated, the method proposing about proximal most position local coordinate system can not be applied in error, because along the tracking error of desirable contour curve approximate zero always, no longer may along the tracking of desirable contour curve.Like this, a local coordinate system is produced, and its direction is similar to the coordinate system of proximal most position, and its initial point approaches the ideal position of desirable contour curve, r _nand ∑ _n.With new definition, it is ∑ _nreplace rotation matrix R and tracking error e in equation (1) _wand e _l, produce e _l2and e _cdifference.The following describes generation is ∑ _nmethod.

Suppose along l ₁tracking error have a negative value (e for example _l1bear) because for reducing e _l1the controller gain distributing is normal than being e _l2that distributes is smaller.Also suppose ideal position r and along desirable contour curve closest approach r _abetween distance and tracking error e _l1big or small approximately equivalent.In addition, the ideal velocity along this section approaches constant.So, require by the time t of this section _dcan be estimated as follows:

Fig. 2 coordinate system ∑ _amiddle initial point (r _a) and inclination angle (θ _a), be the proximal most position to x at desirable contour curve, can be estimated as follows:

r _a＝r(t-t _d)，θ _a＝θ(t-t _d) (5)

Wherein r () and θ () show the function of time.The ideal position r revising _nin Fig. 2, be expressed as:

r _n＝r+R _ad _r，d _r＝[0，-d _ra2] ^T (6)

R wherein _ait is ∑ in Fig. 2 _ato ∑ _wrotation matrix, be θ _aa function.Control inputs equation (1) is replaced by following formula:

E wherein _wn=x-r _n.So, obtain control power below, e _n1and e _n2speed of convergence can be individually adjusted:

Step 3: carry out Robust Control Algorithm, output controlled quentity controlled variable v, drives digital control platform.

In actual processing, comprise that the interference of non-linear friction and cutting force and equipment modeling mistake exists really, therefore, controller equation should be expanded and consider to disturb and equipment modeling error.The feeding that contains BOUNDED DISTURBANCES vector w drives dynamics.

Controller is below proposed:

$\hat{H}=\mathrm{diag}\left\{\frac{{\hat{m}}_i+{\hat{n}}_i{\left(\frac{2\mathrm{\π}}{{\hat{p}}_i}\right)}^2}{\frac{2\mathrm{\π}{\hat{k}}_i}{{\hat{p}}_i}}\right\}$

$\hat{E}=\mathrm{diag}\left\{\frac{{\hat{c}}_i+{\hat{d}}_i{\left(\frac{2\mathrm{\π}}{{\hat{p}}_i}\right)}^2}{\frac{2\mathrm{\π}{\hat{k}}_i}{{\hat{p}}_i}}\right\}---\left(10\right)$

Wherein

the nominal value that represents z, signal

(wherein Λ is a diagonal matrix and the constant that only has positive element) is used the replacement as follow-up stability analysis

rate signal.Symbol δ v is an input vector realizing robust control.By approximate distribution equations feedback gain matrix K _v, K _pand Λ, when leveling off to when infinite, t can obtain e _nlevel off to zero.In addition, along ∑ in Fig. 2 _nthe error convergence speed of each axle can independently be adjusted.

Robust Control Algorithm Guarantee control system stability, even in the time of equipment modeling error and interference existence.Robust control obtains following equation:

$\mathrm{\δv}=-\frac{\mathrm{\η}{R}_a{e}_v}{\mathrm{\ρ}},$

Wherein || a|| is the Euclid norm of a, and ε is little normal number.

In order to show the robust stability of the system that proposes, Leah Pu Nuofu function candidate be employed.

E reaches a conclusion _vand e _nconsistent final bounded.So, from

also be consistent final bounded.Therefore,, although excessive value can reduce control performance, the controller of proposition can improve robust stability.

(3) system software is realized

In the inventive method, vector control system program flow diagram as shown in Figure 4.The master routine of software comprises system initialization; Opening INT1, INT2 interrupts; Permission timer interrupts; Timer Interrupt Subroutine.Wherein initialize routine comprises and closes all interruptions, dsp system initialization, initialization of variable, task manager initialization, AD initialization and quadrature coding pulse QEP initialization.Interrupt service subroutine comprises protection interruption subroutine and T1 underflow interrupt service subroutine.Other parts are as mover initialization location, and PID regulates, and vector etc. are all carried out in timer TI underflow Interrupt Subroutine.

The protection interrupt response that IPM guard signal produces belongs to external interrupt, and INT1 interrupt priority level is than the height of timer T1.IPM can send guard signal automatically in abnormal conditions such as overcurrent, overvoltages, and this signal is connected to the power drive protection pin of DSP through conversion

once have abnormal conditions to occur, DSP can enter protection interruption subroutine, first forbids all interruptions, then block PWM output and make motor stall at once, play the effect of protection motor and IPM.

The smooth startup of vector control system, need to know the initial position of mover, utilize software can lead to the direct current of a constant amplitude to the mover of motor, make stator produce a constant magnetic field, the stationary magnetic field of this magnetic field and rotor interacts, and makes electric mover move to the position that two magnetic linkages overlap.And the reading of mover initial alignment, AD sampled value, the calculating of electric mover position, coordinate transform, PID regulate, the generation of SVPWM waveform fiducial value all completes in T1 underflow interrupt service subroutine.

In T1 underflow interruption subroutine, complete all vector control algorithm.Flow process as shown in Figure 9.In entering, have no progeny, first judge whether mover has completed initial alignment, if initial alignment completes, first program starts AD conversion, and the current value of being sent back to by hardware is collected in DSP.First the data that gather are the result register (RESUTLx being stored in separately, x=0,1) in, from result register RESULT0 and RESULT1, read after A phase and B phase current values are transformed to Q15 form and try to achieve C phase current, then three-phase current is carried out to coordinate transform obtain biphase current under rest frame.

Timer T2 adopts and to increase continuously counting mode, and the step-by-step counting to photoelectric encoder by QEP1 and QEP2 two-phase quadrature coding pulse circuit can obtain the physical location of motor as calculated.The direction of motion that can judge that by reading counting direction position mover is current.Input using the actual position recording and given position as position control, position control output is as the specified rate of q shaft current.Encoder resolution is 1 micron, and because linear motor stator electric pole span is 32mm, therefore timer T2 period register being set is 32000.When mover move distance reaches a pole span, the set of timer T2 interrupt flag bit, T2 restarts counting from 0 simultaneously.After reading new count value, need to remove timer T2 interrupt flag bit.The maximal rate that reaches of experiment is less, when calculating this cycle count value, is greater than 5000 with the difference of last cycle count value, just illustrates that counter has passed through overflow or underflow, correspondingly carries out motor actual calculation of location.It is high like that to the requirement of response speed that position ring is far from electric current loop to the requirement of response speed, therefore stipulate that T1 underflow interrupts all will carrying out the adjusting of ac-dc axis electric current each time, and a position PI adjusting just carried out in every 10 interruptions.

Vector controlled directly completes the adjusting to mover q shaft current by coordinate transform, changes mover thrust, and flow process as shown in Figure 8.By coordinate transform and space vector, calculate, and two vector zero vectors in six basic voltage vectors determining according to switching voltage vector table, correspondingly calculate three comparand register CMPR1, the value of CMPR2 and CMPR3.Its assignment just can be obtained to the SVPWM ripple of expectation to corresponding register, drive IPM to control linear electric motors operation.

The inventive method is finally realized by the control program embedding in dsp processor, and its control procedure is carried out according to the following steps:

Step 1 system initialization;

Step 2 allows TN1, TN2 to interrupt;

Step 3 starts T1 underflow and interrupts;

The initialization of step 4 routine data;

Step 5 is opened total interruption;

Step 6 interrupt latency;

Step 7TN1 interrupts processing sub-control program;

Step 8 finishes;

Wherein in step 6, position adjustments interrupts processing sub-control program according to the following steps:

Step 6-1 position adjustments is interrupted sub-control program;

Step 6-2 reads encoder values;

Step 6-3 judges angle;

Step 6-4 calculates and has walked distance;

Step 6-5 executing location controller;

Step 6-6 carries out robust controller compensation external disturbance;

Step 6-7 calculates current order output;

Step 6-8 interrupts returning;

Wherein in step 7, T1 interrupts processing sub-control program according to the following steps:

Step 7-1T1 interrupts sub-control program;

Step 7-2 keeps the scene intact;

Step 7-3 judges whether initial alignment; Be to enter step 4, otherwise enter step 10;

Step 7-4 current sample, CLARK conversion, PARK conversion;

Step 7-5 judges whether to need position adjustments; Otherwise enter step 9;

Step 7-6 position adjustments interrupts processing sub-control program;

Step 7-7 position control deviation is calculated

The output of step 7-8 controlled quentity controlled variable regulates

Step 7-9dq shaft current regulates;

Step 7-10PARK inverse transformation;

Step 7-11 calculates CMPPx and PWM output;

Step 7-12 position sampling;

Step 7-13 initial alignment program;

Step 7-14 restoring scene;

Step 7-15 interrupts returning.

Claims (2)

1. based on coordinate transform and parameter adjustment, directly drive a digital control platform robust control method, it is characterized in that: concrete steps are as follows:

Step 1: the initial phase of determining electric mover;

By rotor position sample circuit and current sampling circuit, gather respectively electric mover position, speed and current information;

Step 2: profile errors is carried out to coordinate transform, and the line parameter adjustment of going forward side by side, obtains site error, carries out position deviation calculating, judges whether to carry out position adjustments, is carry out step 3, otherwise carries out electric current adjusting;

Step 2.1: profile errors is carried out to coordinate transform

Suppose:

(a1) ideal trajectory r _iwith its derivative

with

can obtain;

(a2) local coordinate system ∑ _linclination angle [theta] and its derivative

with

can obtain;

(a3) feed drive system position x and its derivative are measurable;

(a4), for 2 device parameter values, be all available;

According to above-mentioned hypothesis, determine following control, comprise input voltage v, matrix H, E and the I of driving shaft motor _e:

$v=H\{\stackrel{\·\·}{r}-R(K_v{\stackrel{\·}{e}}_l+K_p{e}_l)-{\stackrel{\·\·}{\mathrm{\θI}}}_e{e}_w+{\stackrel{\·}{\mathrm{\θ}}}^2{e}_w-2{\stackrel{\·}{\mathrm{\θI}}}_e{\stackrel{\·}{e}}_w\}+E\stackrel{\·}{x},$

$H=\mathrm{diag}\left\{\frac{m_i+n_i{\left(\frac{2\mathrm{\π}}{p_i}\right)}^2}{\frac{2\mathrm{\π}}{p_i}}\right\},$

$E=\mathrm{diag}\left\{\frac{c_i+d_i{\left(\frac{2\mathrm{\π}}{p_i}\right)}^2}{\frac{{2\mathrm{\πk}}_i}{p_i}}\right\},$

$I_e=\left[\begin{array}{cc}0& 1\\ -1& 0\end{array}\right]---\left(1\right)$

M wherein _i(>0), c _i(>=0) and f _irespectively driving shaft i table quality, viscous friction coefficient and driving force, wherein K _vand K _pspeed and position feedback gain matrix; They are only had the diagonal matrix of positive element by hypothesis, equation is below set up:

$\mathrm{HR}({\stackrel{\·\·}{e}}_l+K_v{\stackrel{\·}{e}}_l+K_p{e}_l)=0---\left(2\right)$

Because matrix H and R are nonsingular, lower relation of plane need to be satisfied for realizing equation (2):

${\stackrel{\·\·}{e}}_l+K_v{\stackrel{\·}{e}}_l+K_p{e}_l=0---\left(3\right)$

From equation above, by feedback gain matrix K in suitable distribution equations (3) _vand K _preach a conclusion, when t levels off to when infinite, e _lleveling off to 0 is available, if for quadrature error component e _l2feedback gain is set up and is greater than tangential error component e _l1, for ideal curve, profile errors can be reduced sooner than tangent line tracking error, n _ifor driving the motor inertia of feed drive system; k _ifor torque-voltage transitions rate; p _iit is the joint of ball-screw; k _ifor torque-voltage transitions rate;

Step 2.2: parameter adjustment, specific as follows:

Mobile ideal position and local coordinate system, reduce the difference of quadrature error and real profile error, because a tracking error is along l ₁exist, at quadrature error component e _l2with profile errors e _cbetween may there is a unacceptable difference, for reducing this difference, at desirable contour curve, to physical location x, estimate nearest position, use r _arepresent; A local coordinate system is produced, and its direction is similar to the coordinate system of proximal most position, and its initial point approaches the ideal position of desirable contour curve, r _nand ∑ _n; With new definition, it is ∑ _nreplace rotation matrix R and tracking error e in equation (1) _wand e _l, produce e _l2and e _cdifference, the following describes that to produce be ∑ _nmethod:

Suppose along local coordinate system ∑ _ltransverse axis l ₁tracking error have a negative value because for reducing e _l1the controller gain distributing is normal than being e _l2that distributes is smaller, also supposes ideal position r and along desirable contour curve closest approach r _abetween distance and tracking error e _l1big or small approximately equivalent, in addition, along the ideal velocity of this section, approach constant, so, require by the time t of this section _dcan be estimated as follows:

$t_d=-e_{l1}/\sqrt{{\stackrel{\·}{r}}_x^2+{\stackrel{\·}{r}}_y^2}---\left(4\right)$

Coordinate system ∑ _amiddle initial point r _aand inclination angle [theta] _a, be the proximal most position to x at desirable contour curve, can be estimated as follows:

r _a=r(t-t _d),θ _a=θ(t-t _d) (5)

Wherein r () and θ () show the function of time;

The ideal position r revising _nbe expressed as:

r _n=r+R _ad _r,d _r=[0,-d _ra2] ^T (6)

Wherein, d _ra2for coordinate system ∑ _afrom r _avector d to r _rasecond element, R _ait is ∑ _ato ∑ _wrotation matrix, be θ _aa function, control inputs equation (1) is replaced by following formula:

$v=H\{{\stackrel{\·\·}{r}}_n-R_a(K_v{\stackrel{\·}{e}}_n+K_p{e}_n)-{\stackrel{\·\·}{\mathrm{\θ}}}_a{I}_e{e}_{\mathrm{wn}}+{\stackrel{\·}{\mathrm{\θ}}}_a^2{e}_{\mathrm{wn}}-{2\stackrel{\·}{\mathrm{\θ}}}_a{I}_e{\stackrel{\·}{e}}_{\mathrm{wn}}\}+E\stackrel{\·}{x}---\left(7\right)$

Wherein, the tracking error vector e between physical location and the ideal position of modification _wn=x-r _n;

So, obtain control power below, e _n1and e _n2speed of convergence can be individually adjusted:

${\stackrel{\·\·}{e}}_n+K_v{\stackrel{\·}{e}}_n+K_p{e}_n=0;$

Step 3: carry out Robust Control Algorithm, output controlled quentity controlled variable v, drives digital control platform;

In actual processing, comprise that the interference of non-linear friction and cutting force and equipment modeling mistake exists really, therefore, controller equation should be expanded and consider to disturb and equipment modeling error, and the feeding that contains BOUNDED DISTURBANCES vector w drives dynamics,

$H\stackrel{\·\·}{x}+E\stackrel{\·}{x}+w=v---\left(9\right)$

Controller is below proposed:

$v=\hat{H}\{{\stackrel{\·\·}{r}}_n-R_a(K_v{e}_v+K_p{e}_n)-{\stackrel{\·\·}{\mathrm{\θ}}}_a{I}_e{e}_{\mathrm{wn}}+{\stackrel{\·}{\mathrm{\θ}}}_a^2{e}_{\mathrm{wn}}-{2\stackrel{\·}{\mathrm{\θ}}}_a{I}_e{\stackrel{\·}{e}}_{\mathrm{wn}}\}+\hat{E}\stackrel{\·}{x}+\mathrm{\δv}$

$\hat{H}=\mathrm{diag}\left\{\frac{{\hat{m}}_i+{\hat{n}}_i{\left(\frac{2\mathrm{\π}}{{\hat{p}}_i}\right)}^2}{\frac{{2\mathrm{\π}\hat{k}}_i}{{\hat{p}}_i}}\right\}$

$\hat{E}=\mathrm{diag}\left\{\frac{{\hat{c}}_i+{\hat{d}}_i{\left(\frac{2\mathrm{\π}}{{\hat{p}}_i}\right)}^2}{\frac{{2\mathrm{\π}\hat{k}}_i}{{\hat{p}}_i}}\right\}$

Wherein

the nominal value that represents z, signal in formula, Λ is a diagonal matrix and the constant that only has positive element, signal e _vbe used the replacement as follow-up stability analysis

rate signal, symbol δ v is an input vector realizing robust control, e _nfor ∑ _ato ∑ _wthe transposition of rotation matrix and the tracking error vector e between physical location and the ideal position of modification _wnproduct;

By approximate distribution equations feedback gain matrix K _v, K _pand Λ, when leveling off to when infinite, t can obtain e _nlevel off to zero; In addition, along ∑ _nthe error convergence speed of each axle can independently be adjusted;

Robust control obtains following equation:

Wherein || a|| is the Euclid norm of a, and ε is little normal number;

In order to show the robust stability of the system that proposes, Leah Pu Nuofu function candidate be employed;

E reaches a conclusion _vand e _nconsistent final bounded, so, from also be consistent final bounded.

2. the control system that directly drives digital control platform robust control method to adopt based on coordinate transform and parameter adjustment as claimed in claim 1, is characterized in that: comprise voltage-regulating circuit, rectification filtering unit, IPM inversion unit, DSP, Hall element, grating scale, current sampling circuit, position sampling circuit, IPM isolation drive holding circuit;

Alternating voltage exports rectification filtering unit input end to, rectification filtering unit output terminal access IPM inversion unit, IPM is connected with motor, motor fuselage is equipped with grating scale, grating scale link position sample circuit input end, Hall element gathers motor current signal, export current sampling circuit to, current sampling circuit output terminal and position sampling circuit output end all access DSP, DSP outputs signal to voltage-regulating circuit input end and IPM isolation drive holding circuit, voltage-regulating circuit is adjusted alternating voltage, IPM isolation drive holding circuit access IPM inversion unit.

Images