CN109039173A - A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing - Google Patents
A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing Download PDFInfo
- Publication number
- CN109039173A CN109039173A CN201810900809.8A CN201810900809A CN109039173A CN 109039173 A CN109039173 A CN 109039173A CN 201810900809 A CN201810900809 A CN 201810900809A CN 109039173 A CN109039173 A CN 109039173A
- Authority
- CN
- China
- Prior art keywords
- pmlsm
- particle
- control
- speed
- signal
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P6/00—Arrangements for controlling synchronous motors or other dynamo-electric motors using electronic commutation dependent on the rotor position; Electronic commutators therefor
- H02P6/14—Electronic commutators
- H02P6/16—Circuit arrangements for detecting position
- H02P6/17—Circuit arrangements for detecting position and for generating speed information
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Feedback Control In General (AREA)
Abstract
The invention belongs to numerical control processing control technology field more particularly to a kind of iterative learning control methods and system of the PMLSM based on hybridization particle group optimizing.It is directed to repeatability or periodic controlled device, has stringent mathematical description, perfect theoretical system, is not entirely dependent on the accurate model of system, can control nonlinear system.Including main circuit, control circuit and control object.The control object is the three-phase PMLSM that fuselage is equipped with grating scale.The main circuit: for 220V alternating current to be converted into the three-phase alternating current of driving PMLSM.The control circuit: it realizes to the iterative learning position control based on hybridization particle group optimizing of PMLSM, PI speed control.
Description
Technical field
The invention belongs to numerical control processing control technology field more particularly to a kind of PMLSM based on hybridization particle group optimizing
Iterative learning control method and system.
Background technique
Since nearly half a century, as the performance of power electronic devices is continuously improved, it is straight drive control technology increasingly at
It is ripe, the extensive concern of people is received in the permanent magnetic linear synchronous motor of the upper extensive application of numerically-controlled machine tool, in ontology and control
Tactful aspect expands a large amount of relevant researchs, and achieves quite significant achievement.High-grade, digitally controlled machine tools use linear motor
Driving is following development trend, and high thrust linear motor is becoming the key foundation component of high-grade, digitally controlled machine tools, country
Support and propulsion energetically will be given to linear motor control and the research of actuation techniques, so the control that research linear motor is new
Technology is of great significance to China is improved in the theoretical research in linear motor field and industrial application level.
Permanent magnet linear synchronous motor utilizes high-energy permanent magnet, saves intermediate conversion mechanism, loss big with thrust strength
It is low, operational reliability is high, time constant is small, device is simple, response it is fast the features such as, make the quick-reaction capability and fortune of feed system
Dynamic precision is greatly improved.But due to two end regions of permanent magnet linear synchronous motor iron core and winding and its
The Distribution of Magnetic Field in middle position is dramatically different, along with uncertain factors such as Parameter Perturbations, is difficult to establish accurate permanent-magnet linear
The mathematical model of synchronous motor.Moreover, linear motor uses direct drive mode, the load disturbance of system, itself and outside are disturbed
The uncertain factors such as dynamic, Parameter Perturbation will bear directly against motor itself, without the buffering course of any centre, directly affect
Feed system performance is driven to straight-line electric, this considerably increases the control difficulty of linear motor.
It is a kind of feed forward control method that iterative learning, which controls (ILC), for having repeatability or periodic controlled device,
With stringent mathematical description, perfect theoretical system is not entirely dependent on the accurate model of system, can control nonlinear system
System.ILC algorithm principle is simply easily achieved, and has good robustness, therefore obtained many applications.But in tradition
ILC control process in, iterative learning control gain be fixed and invariable, and in iterative learning procedure error signal exist
Divergent component accumulation will lead to system convergence be deteriorated even dissipate.It is insufficient in order to make up conventional iterative study control, meet
The servo-system performance requirement of the high-precision of Numeric Control Technology, high speed, this just need to design the high speed suitable for PMLSM,
The servo-control system of high-precision and strong robustness, so the invention proposes a kind of PMLSM based on hybridization particle group optimizing
Iterative learning control method and system.
Summary of the invention
The present invention provides a kind of based on the PMLSM's for hybridizing particle group optimizing aiming at defect of the existing technology
Iterative learning control method and system.It is directed to repeatability or periodic controlled device, have stringent mathematical description,
Perfect theoretical system is not entirely dependent on the accurate model of system, can control nonlinear system.
To achieve the above object, the present invention adopts the following technical scheme that, including main circuit, control circuit and control object；
The control object is the three-phase PMLSM that fuselage is equipped with grating scale；
The main circuit: for 220V alternating current to be converted into the three-phase alternating current of driving PMLSM；
The control circuit: iterative learning position control, the PI speed based on hybridization particle group optimizing to PMLSM are realized
Control.
As a preferred solution of the present invention, the main circuit includes regulating circuit, rectification filtering unit and IPM inversion
Unit；
The rectification filtering unit: by being connected with three-phase alternating-current supply, the AC conversion for that will change is to stablize
Direct current；
The IPM inversion unit: the DC inverter for exporting current rectifying and wave filtering circuit supplies PMLSM at alternating current.
As another preferred embodiment of the invention, the input terminal of the voltage regulating module connects 220V alternating current, pressure regulation mould
The input terminal of the output end connection rectification filtering unit of block, the input of the output end connection IPM inversion unit of rectification filtering unit
The output end at end, IPM inversion unit connects PMLSM；A/D of the IPM inverter circuit through current sampling circuit connection dsp processor is logical
Road, the output end of PMLSM are connected to digital I/O mouthfuls of dsp processor through grating scale, position and speed detection circuit, dsp processor
Output end connection IPM isolation drive protection circuit input terminal, IPM isolation drive protect circuit output end connection IPM it is inverse
Become the control signal of unit.
As another preferred embodiment of the invention, the control circuit is used to control the switching tube in IPM inversion unit
On-off comprising dsp processor, current sampling circuit, position and speed sample circuit, IPM protective separation driving circuit.
As another preferred embodiment of the invention, the dsp processor: according to position, speed and the electric current received
Signal executes HPSO-ILC control algolithm, generates the switching tube on-off in driving signal control IPM inversion unit；
Current sampling circuit: the sampled signal of Hall current sensor is converted into the level signal of 0-3V；
Position and speed sample circuit: converting the collected rotor position speed signal of grating scale to can be by dsp processor
The digital quantity of identification；
IPM protective separation driving circuit: it is generated according to dsp processor in HPSO-ILC position control, PI speed control
Different pwm signals driving IPM inversion unit work and protection IPM inversion unit.
As another preferred embodiment of the invention, the process of signal is handled in the dsp processor are as follows: given
It after PMLSM position signal, is made the difference with the actual position signal detected through grating scale, generates position error signal, position is missed
Current controling signal, electric current control is calculated through HPSO-ILC controller in input quantity of the difference signal as HPSO-ILC controller
Signal processed generates pwm pulse sequence through DSP, and pwm pulse sequence controls the conducting and shutdown of six IGBT of IPM inverter circuit,
It obtains meeting the three-phase alternating current needed, send to PMLSM and control its movement.
A kind of PMLSM control method, comprising the following steps:
Step 1: the given position PMLSM input signal；
Step 2: initialization hybridization particle swarm optimization algorithm: initialization HPSO algorithm, determine algorithm population dimension D,
Population inertia weight range [ωMin,ωmax], accelerator coefficient c1And c2, position range [xmin, xmax], velocity interval [vmin,
vmax], probability of crossover bc, hybridization pond ratio bs, maximum number of iterations M, optimum solution stop condition foc.Random initializtion particle speed
Degree, position, and using first generation particle as the personal best particle p of particlebest, the individual optimal adaptation value of adaptive value conduct,
The smallest adaptive value is used as overall situation optimum position g as global optimal adaptation value, position in particlebest；
Step 3: the particle of generation is assigned to the iterative learning controller gain alpha of PMLSM respectively, beta, gamma, and
Butterworth zero phase low-pass filter bandwidth fc；
Step 4: determining the output signal u of iterative learning controllerj+1(t).By given desired output signal xd(t), it deposits
The control of the iteration j study of storage memory inputs uj(t), the reality output x of+1 iterative learning of jthj+1(t) and jth+
Tracking error e when 1 iterative learningj+1(t), it obtains the output signal of iteration controller and is sent into controlled device, wherein ej+1
(t)=xd(t)-xj+1(t)；
Step 5: more new particle individual optimal value and global optimum: correcting iterative learning controller using HPSO algorithm
Gain alpha, beta, gamma and Butterworth zero phase low-pass filter bandwidth fc.Update position and the speed of each particle of population
Degree, and then calculate the adaptive value of each particle.
Step 5-1: updating the position and speed of population particle, and formula is as follows:
xI, j(k+1)=xI, j(k)+vI, j(k+1), j=1,2 ..., d (1)
vI, j(k+1)=ω vI, j(k)+c1r1[pI, j-xI, j(k)]+c2r2[pG, j-xI, j(k)] (2)
In formula: ω is inertia weight；c1And c2For normal number, referred to as accelerator coefficient；r1And r2It is the random number of [0,1]；
xI, jIt (k) is particle position；vI, jIt (k) is particle rapidity；pI, jFor particle personal best particle；pG, jFor population optimal location；
Step 5-2: the adaptive value of more each particle and individual optimal adaptation value pbestIf control effect is more excellent, just
Update its individual optimum position and individual optimal adaptation value；
Step 5-3: the adaptive value of more each particle and global optimal adaptation value gbestIf control effect is more excellent, just
Update its global optimum position and global optimal adaptation value；
Step 5-4: choosing the particle of specified quantity according to probability of crossover, and puts it into hybridization pond, the particle in pond
Hybridization generates same number of filial generation particle two-by-two at random, and calculates the position and speed of filial generation.It is replaced using filial generation particle
Parent particle；
Wherein the adaptive value of particle is systematic error absolute time integral function ITAE, and calculation formula is as follows:
In formula: t is the time；E (t) is system given value and the deviation that system exports；
Step 5-5: it when algorithm reaches its stop condition, then stops search and exports result；Otherwise 5-1 step is returned to continue
Search.
As another preferred embodiment of the invention, the step 4 carries out as follows:
Real-time perfoming current sample and position sampling in step 4-1:PMLSM control system motion process；
Step 4-2:DSP processor is generated according to the current sampling data and position sampled data at current time to PMLSM
Carry out the pwm signal of position control；
Step 4-2-1: HPSO-ILC position control is carried out to PMLSM according to position sampled data: by desired locations and position
It sets after sampled data makes the difference, obtains position deviation, obtain desired speed after HPSO-ILC is calculated；
Step 4-2-2: PI speed control is carried out to PMLSM according to position sampled data: after the sampled data differential of position,
Obtain actual speed, then after the resulting desired speed of step 4-2-1 and actual speed are made the difference, it is inclined to obtain speed after PI is calculated
Difference obtains expectation electric current；
Step 4-2-3: the pwm signal that position control is carried out to PMLSM is generated according to current sampling data: by step 4-2-
2 resulting expectation electric currents make 2/3 transformation, then after transformation results and current sampling data are made the difference, obtain driving IPM inversion unit
Pwm signal, execute step 4-3；
Step 4-3:PMLSM works according to pwm signal: IPM inversion unit is believed according to the resulting PWM of step 4-2-3
Number, run PMLSM, real-time perfoming current sample and position sampling in PMLSM motion process；
Step 4-4: return step 4-2-2.
Further, in the step 2, the random value of speed is limited in [vmin, vmax], the random value of position is limited in
[xmin, xmax].Population inertia weight is in [ωmin, ωmax] successively decrease in range by linear relationship.
Further, in the step 4, the output signal of iterative learning controller:
In formula: α, β and γ are respectively ratio, integral and the differential study gain parameter of ILC controller.
It is further improved, in the step 5-1:
Inertia successively decreases weight:
In formula: ωmaxIndicate inertia weight maximum value；ωminIndicate inertia weight minimum value；K indicates current iteration number.
It is further improved, in the step 5-4: the calculation formula of the position and speed of filial generation are as follows:
In formula: mx represents the position of parent particle；Nx represents the position of filial generation particle；The speed of mv expression parent particle；
The speed of nv expression filial generation particle；I is the random number between 0 to 1.Keep pbestAnd gbestIt is constant.
Beneficial effect of the present invention compared with prior art.
ILC of the present invention is a kind of intelligent control method, mainly for having repeatability or periodic controlled device, is had
Stringent mathematical description, perfect theoretical system are not entirely dependent on the accurate model of system, can control nonlinear system.
But in traditional ILC control process, iterative learning control gain is fixed and invariable, and in iterative learning procedure accidentally
The accumulation of divergent component existing for difference signal, which will lead to system convergence and be deteriorated, even to be dissipated.The present invention will hybridize Particle Swarm Optimization
Method is combined with Iterative Learning Control Algorithm and Butterworth zero phase low-pass filter, passes through Crossbreeding Particle Swarm
The gain of real-time optimization iterative learning controller and Butterworth zero phase low-pass filter bandwidth.Next iteration it
It is preceding to filter out divergent component using Butterworth filter, improve system convergence speed.Draw in Crossbreeding Particle Swarm simultaneously
Enter the inertia weight of linear decrease.Compared with standard particle group's algorithm, Crossbreeding Particle Swarm has fast convergence rate, precision
Height is not easy the characteristics of falling into " precocity ", to realize good control effect.
Detailed description of the invention
The present invention will be further described with reference to the accompanying drawings and detailed description.The scope of the present invention not only limits to
In the statement of the following contents.
Fig. 1 is the specific embodiment of the invention based on Crossbreeding Particle Swarm optimization PMLSM iterative learning control principle
Block diagram.
Fig. 2 is ILC position control functional block diagram of the present invention.
Fig. 3 is PMLSM Control system architecture block diagram of the present invention.
Fig. 4 is PMLSM control system hardware main circuit schematic diagram of the present invention.
Fig. 5 is PMLSM control system position signal sample circuit schematic diagram of the present invention.
Fig. 6 is PMLSM control system current sampling circuit schematic diagram of the present invention.
Fig. 7 is PMLSM control system IPM isolation drive protection circuit diagram of the present invention.
Fig. 8 is the control flow chart of dsp processor of the present invention.
Fig. 9 is the flow chart of present invention protection interrupt processing.
Figure 10 is the flow chart of T1 interrupt processing of the present invention.
Figure 11 is that the present invention is based on the PMLSM iterative learning control method flow charts of Crossbreeding Particle Swarm optimization.
Specific embodiment
It elaborates with reference to the accompanying drawing to the specific embodiment of the invention.
Present embodiment based on hybridization particle group optimizing PMLSM iterative learning control method principle as shown in Figure 1,
Including the use of the particle group optimizing Butterworth zero phase low-pass filter and PID type iterative learning controller of hybridization, PI speed
Spend controller and controlled device.The effect of Butterworth zero phase low-pass filter is filtered out before next iteration
Divergent component in error signal；ILC positioner makes controlled device limited to obtain ideal control input signal
High-precision pursuit path is exported in time and section；PI speed control is to improve speed loop efficiently against disturbance
Response characteristic；Controlled device is PMLSM.
Based on the PMLSM iterative learning control method of Crossbreeding Particle Swarm optimization, as shown in figure 11, including following step
It is rapid:
Step 1: the given position PMLSM input signal；
Step 2: initialization hybridization particle swarm optimization algorithm: initialization HPSO algorithm, determine algorithm population dimension D,
Population inertia weight range [ωmin, ωmax], accelerator coefficient c1And c2, position range [xmin, xmax], velocity interval [vmin,
vmax], probability of crossover bc, hybridization pond ratio bs, maximum number of iterations M, optimum solution stop condition foc.Random initializtion particle speed
Degree, position, and using first generation particle as the personal best particle p of particlebest, the individual optimal adaptation value of adaptive value conduct,
The smallest adaptive value is used as overall situation optimum position g as global optimal adaptation value, position in particlebest, the random value of speed
It is limited in [vmin, vmax], the random value of position is limited in [xmin, xmax], population inertia weight is in [ωmin, ωmax] in range
Successively decrease by linear relationship；
Step 3: the particle of generation is assigned to the iterative learning controller gain alpha of PMLSM respectively, beta, gamma, and
Butterworth zero phase low-pass filter bandwidth fc；
Step 4: determining the output signal u of iterative learning controllerj+1(t).By given desired output signal xd(t), it deposits
The control of the iteration j study of storage memory inputs uj(t), the reality output x of+1 iterative learning of jthj+1(t) and jth+
Tracking error e when 1 iterative learningj+1(t), it obtains the output signal of iteration controller and is sent into controlled device, wherein ej+1
(t)=xd(t)-xj+1(t)；
Real-time perfoming current sample and position sampling in step 4-1:PMLSM control system motion process；
Step 4-2:DSP processor is generated according to the current sampling data and position sampled data at current time to PMLSM
Carry out the pwm signal of position control；
Step 4-2-1: HPSO-ILC position control is carried out to PMLSM according to position sampled data: by desired locations and position
It sets after sampled data makes the difference, obtains position deviation, obtain desired speed after HPSO-ILC is calculated；
ILC position control is substantially a kind of feedforward control, mainly in periodical or repeatability system, benefit
The control input signal of next time is corrected and formed with each output error signal and control input signal, is repeatedly run in this way
System reduces the error exported every time constantly, makes reality output constantly close to desired output.
ILC position control principle is as shown in Fig. 2, signal all in figure is all defined on signal all in finite interval figure
It is all defined on finite interval [0, T], subscript j indicates iteration j, ydIt (t) is the desired output of controlled system, yjIt (t) is the
The reality output of j iteration, ejIt (t) is the output error e of iteration jj(t)=yd(t)-yj(t), ILC controller jth+1
The control of secondary iteration inputs, and can generally indicate are as follows:
uj+1(t)=L [uj(t), ej(t)] (1)
In formula: L [] is law of learning function, and common law of learning has PID type law of learning, Optimal Learning rule, feedback-feedforward
Law of learning, adaptive learning rule, High-level Learning rule and discrete system law of learning etc..By selecting suitable law of learning, so that working as
When the number of iterations j → ∞.yj(t)→yd(t)。
In formula: α, β and γ are respectively ratio, integral and the differential study gain parameter of ILC controller.
Step 4-2-2: PI speed control is carried out to PMLSM according to position sampled data: after the sampled data differential of position,
Obtain actual speed, then after the resulting desired speed of step 4-2-1 and actual speed are made the difference, it is inclined to obtain speed after PI is calculated
Difference obtains expectation electric current.
It, will be by the output signal and feedback speed of HPSO-ILC positioner by PI controller design in speed ring
Input deviation e (t) of the comparison result of signal as PI speed control, the output signal u (t) of PI speed control is as control
Input signal processed is sent into controlled device.
Step 4-2-3: the pwm signal that position control is carried out to PMLSM is generated according to current sampling data: by step 4-2-
2 resulting expectation electric currents make 2/3 transformation, then after transformation results and current sampling data are made the difference, obtain driving IPM inversion unit
Pwm signal, execute step 4-5.
Step 4-3:PMLSM works according to pwm signal: IPM inversion unit is believed according to the resulting PWM of step 4-2-3
Number, run PMLSM, real-time perfoming current sample and position sampling in PMLSM motion process.
Step 4-4: return step 4-2-2.
Step 5: iterative learning controller more new particle individual optimal value and global optimum: being corrected according to HPSO algorithm
Gain alpha, beta, gamma and Butterworth zero phase low-pass filter bandwidth fc.Update position and the speed of each particle of population
Degree, and then calculate the adaptive value of each particle.
Step 5-1: updating the position and speed of population particle, and formula is as follows:
xI, j(k+1)=xI, j(k)+vI, j(k+1), j=1,2 ..., d (3)
vI, j(k+1)=ω vI, j(k)+c1r1[pI, j-xI, j(k)]+c2r2[pG, j-xI, j(k)] (4)
In formula: ω is inertia weight；c1And c2For normal number, referred to as accelerator coefficient (usually taking 2)；r1And r2It is [0,1]
Random number；xI, jIt (k) is particle position；vI, jIt (k) is particle rapidity；pI, jFor particle personal best particle；pG, jMost for population
Excellent position.
Inertia successively decreases weight:
In formula: ωmaxIndicate inertia weight maximum value, ωminIndicate inertia weight minimum value, k indicates current iteration number.
Step 5-2: the adaptive value of more each particle and individual optimal adaptation value pbestIf control effect is more excellent, just
Update its individual optimum position and individual optimal adaptation value；
Step 5-3: the adaptive value of more each particle and global optimal adaptation value gbestIf control effect is more excellent, just
Update its global optimum position and global optimal adaptation value；
Step 5-4: choosing the particle of specified quantity according to probability of crossover, and puts it into hybridization pond, the particle in pond
Hybridization generates same number of filial generation particle two-by-two at random, and calculates the position and speed of filial generation.It is replaced using filial generation particle
Parent particle.
The adaptive value of particle is systematic error absolute time integral function ITAE, and calculation formula is as follows:
In formula: t is the time；E (t) is system given value and the deviation that system exports.
The calculation formula of the position and speed of filial generation are as follows:
In formula: mx represents the position of parent particle；Nx represents the position of filial generation particle；The speed of mv expression parent particle；
The speed of nv expression filial generation particle；I is the random number between 0 to 1.Keep pbestAnd gbestIt is constant.
Step 5-5: it when algorithm reaches its stop condition, then stops search and exports result；Otherwise 5-1 step is returned to continue
Search.
PMLSM control system used by above-mentioned PMLSM control method, as shown in Figure 3, comprising: for exchanging 220V
Electricity is converted into the main circuit of the three-phase alternating current of driving PMLSM；For according to current sampling signal and position sampled signal pair
PMLSM carries out HPSO-ILC position control, the control circuit of PI speed control；The input terminal of main circuit connects 220V alternating current,
The three-phase input end of the output end connection PMLSM of main circuit；The grating of the position sampling input terminal connection PMLSM of control circuit
Ruler, the input terminal of the current sample input terminal connection PMLSM of control circuit.
Main circuit is as shown in figure 4, include voltage regulating module, rectification filtering unit and IPM inversion unit.
The input terminal of voltage regulating module connects 220V alternating current, the input of the output end connection rectification filtering unit of voltage regulating module
End, the input terminal of the output end connection IPM inversion unit of rectification filtering unit, the output end of IPM inversion unit are separately connected
The three-phase input end of PMLSM.
Voltage regulating module is using three-phase intelligent voltage regulating module EUV -25A-II, it can be achieved that the isolation pressure regulation of 0~220V.Rectification
Filter unit uses the uncontrollable rectification of bridge-type, and bulky capacitor filtering cooperates resistance capaciting absorpting circuit appropriate, can get IPM inversion list
Constant DC voltage needed for member work.IPM inversion unit uses the 6MBP50RA060 intelligent power module of company of Fuji, it
Pressure resistance be 600V, maximum current 50A, maximum operating frequency 20kHz.IPM inversion unit is driven using four groups of independent 15V
Dynamic power supply power supply.P, N is DC bus-bar voltage input terminal, is connected with the output end of rectification filtering unit, the end P is positive, and N-terminal is
It is negative；B is discharge end, is connected with the collector of internal bleeder pipe；U, V, W are three-phase inversion voltage output ends.IPM inversion unit is defeated
For three-phase alternating current out by output terminal U, V, W are connected to PMLSM.
Control circuit includes: position sample circuit, current sampling circuit, IPM isolation drive protection circuit, dsp processor.
The grating scale of position sample circuit connection PMLSM；The input terminal of Hall current sensor connection PMLSM；Hall current sensor
Output end connection current sampling circuit input terminal, the output end point of the output end of current sampling circuit, position sample circuit
Not Lian Jie dsp processor input terminal, dsp processor output end connection IPM isolation drive protection circuit input terminal, IPM
Isolation drive protects the control signal of the output end connection IPM inversion unit of circuit.
Position sample circuit: effect is to acquire PMLSM position signal by grating scale.
Current sampling circuit: effect is that the current-mode analog quantity for detecting Hall current sensor is changed into digital quantity, defeated
Enter dsp processor to be handled.
Speed sampling circuit: the speed signal that grating scale acquires is changed into digital quantity, is inputted at dsp processor
Reason.
IPM isolation drive protects circuit: effect is according to dsp processor in HPSO-ILC position control, PI speed control
The different pwm signals driving IPM inversion unit work of Shi Shengcheng and protection IPM inversion unit；
Dsp processor: effect is to carry out the position HPSO-ILC to PMLSM according to current sampling signal and position sampled signal
Control, PI speed control.
Position sample circuit is as shown in Figure 5: when PMLSM movement, the grating scale of fuselage installation can export and location information
Related three road pulse signal is used to detect the pulse signal A and B of location information including two-way, and all the way for returning to zero ginseng
The signal Z examined.This three roads pulse signal is isolated by high speed photo coupler 6N137.Because of the San Lumai of grating scale output
Rushing signal is 5V, and the I/O of dsp processor mouth voltage is 3.3V, so needing to turn signal by 5V by bleeder circuit
It is changed to 3.3V.Pulse signal A and B after conversion is finally connected respectively to the two-way quadrature coding pulse interface of dsp processor
Pulse signal Z after conversion is connected to the capturing unit CAP3 of DPS by QEP1 and QEP2.DSP trapped inside unit can be used soft
Part is defined as quadrature coding pulse input unit, can count later to pulse, may determine that PMLSM according to pulse train
The direction of motion and position.
(A, B biphase current detection circuit are identical, therefore only provide A phase current sensing electricity as shown in Figure 6 for current sampling circuit
Road): current detection circuit is the three phase promoter electric currents of PMLSM to be sent into DSP to be converted into digital form and carry out after sensor
A series of transformation.Since this system is three-phase balanced system, i.e., three-phase current vector sum is zero, therefore only needs to detect A, B
Biphase current, so that it may obtain three-phase current.Present embodiment is using two CSM025PT5 series Hall current sensor detections
A, B biphase current, the current range that it can be acquired are -16A~+16A, and the voltage range of output is 0~5V.Wherein pin 5,6
Scholar's 15V voltage is connect respectively, and what pin 1 exported is the A phase current of Hall current sensor detection.Because Hall current sensor is defeated
Voltage range out is 0~5V, and the A/D module samples voltage range of dsp processor is 0~3V, and operational amplifier can be used
Signal is adjusted to 0~3V by adjusting resistance VR1 by 0P07GS.The power supply of amplifier connects scholar's 15V voltage, between voltage and ground
Connect decoupling capacitance.Circuit input end connects capacitor filtering, to remove high-frequency signal interference, improves sampling precision.After finally adjusting
A, B biphase current be connected respectively to AD0, AD1 pin of dsp processor.
IPM isolated drive circuit is as shown in Figure 7: using eight single-wire drive device 74LS240, the input terminal of driver is connected to
PWM module inside dsp processor, is controlled by dsp processor.The failure output terminal of IPM inversion unit passes through photoelectric coupling circuit
HCPL4506 is connected to (PDPINTA)-pin of dsp processor, and DPS in time will when ensuring that IPM inversion unit breaks down
All incident management outputs are set to high-impedance state, and DSP stops to driver output pwm signal, to protect IPM inversion unit.
The core dsp processor of control circuit uses TMS320F2812, and matched development board includes the read-only storage of target
Device, eCAN interface, serial boot ROM, user lamp, reset circuit, can be configured to RS232/RS422/ at analog interface
The outer 256*16 RAM of asynchronous serial port, SPI synchronous serial interface and the piece of RS485.
When PMLSM is worked normally, the quadrature coding pulse circuit of TMS320F2812 chip task manager EVA is enabled
With capturing unit CAP3, the actual position information of PMLSM is received.16 channel A/D modules of enabled TMS320F2812 chip, connect
Receive the actual current information of control PMLSM.The PWM module of enabled TMS320F2812 chip, exports PWM wave, controls IPM inversion
Unit, to realize the control to PMLSM.
After desired locations and the resulting physical location of detection are made the difference, HPSO-ILC control program is called to obtain expectation speed
Degree.Actual speed is obtained after will test resulting physical location differential.Desired speed is made the difference with resulting actual speed is calculated
Afterwards, it calls PI to control program, obtained expectation electric current, and 2/3 transformation is carried out to it.Utilize transformed expectation electric current and detection
Resulting actual current generates PWM wave, and output protects circuit to IPM isolation drive.
By (PDPINTA) of TMS320F2812 chip-Pin is connected with the failure output terminal of IPM inversion unit, protection
With monitoring whole system.When the failures such as over-voltage, overcurrent, under-voltage occurs in system, TMS320F2812 chip can be timely
By (PDPINTA)-Pin is set to high-impedance state, blocks PWM output signal, protects IPM unit.
The core of control circuit is TMS320LF2812DSP processor, including HPSO-ILC position control module, PI speed
Control module and pwm signal generation module.It is responsible for conversion A/D conversion, the calculating of motor speed, the PI adjusting of speed and position
HPSO-ILC is adjusted, and finally obtains the control signal of Voltage space vector PWM, drives IPM inversion unit.Dsp controller is also negative
Duty protection and monitoring whole system, once the failures such as over-voltage, overcurrent, under-voltage occurs in system, DSP will block PWM output letter
Number, to protect IPM inversion unit.
The control flow of dsp processor is as shown in Figure 8, comprising the following steps:
1, system initialization；
2, DSP is initialized；
3, INTI, INTZ is allowed to interrupt；
4, starting TI underflow is interrupted；
5, interrupt latency；
6, TI interrupt processing；
7, interrupt processing is protected；
8, terminate.
Wherein, protection interrupt processing is as shown in Figure 9, comprising the following steps:
1, forbid all interruptions；
2, IPM is blocked；
3, it interrupts and returns.
T1 interrupt processing is as shown in Figure 10, comprising the following steps:
1, it keeps the scene intact；
2, position sample, compared with given position signal after obtain position deviation；
3, the HPSO-ILC of calling station controls program；
4, calculate motor speed, compared with the pi regulator output signal of position after obtain velocity deviation；
5, the PI of speed is called to control program；
6, current sample is carried out；
7,3/2 transformation is carried out to current sampling data；
8, q shaft current calculating torque is utilized；
9,2/3 transformation is carried out to the electric current of output；
10, the current value for using transformation to obtain obtains pwm signal as carrier wave and triangular modulation；
11, it interrupts and returns.
It is understood that being merely to illustrate the present invention above with respect to specific descriptions of the invention and being not limited to this
Technical solution described in inventive embodiments, those skilled in the art should understand that, still the present invention can be carried out
Modification or equivalent replacement, to reach identical technical effect；As long as meet use needs, all protection scope of the present invention it
It is interior.
Claims (10)
1. a kind of PMLSM iterative learning control systems based on hybridization particle group optimizing, including main circuit, control circuit and control
Object；It is characterized in that, the control object is the three-phase PMLSM that fuselage is equipped with grating scale；
The main circuit: for 220V alternating current to be converted into the three-phase alternating current of driving PMLSM；
The control circuit: it realizes to the iterative learning position control based on hybridization particle group optimizing of PMLSM, PI speed control
System.
2. a kind of PMLSM iterative learning control systems based on hybridization particle group optimizing according to claim 1, feature
Be: the main circuit includes regulating circuit, rectification filtering unit and IPM inversion unit；
The rectification filtering unit: by being connected with three-phase alternating-current supply, AC conversion for that will change is stable straight
Galvanic electricity；
The IPM inversion unit: the DC inverter for exporting current rectifying and wave filtering circuit supplies PMLSM at alternating current.
3. a kind of PMLSM iterative learning control systems based on hybridization particle group optimizing according to claim 1, feature
Be: the input terminal connection 220V alternating current of the voltage regulating module, the output end of voltage regulating module connect the defeated of rectification filtering unit
Enter end, the input terminal of the output end connection IPM inversion unit of rectification filtering unit, the output end connection of IPM inversion unit
PMLSM；A/D channel of the IPM inverter circuit through current sampling circuit connection dsp processor, the output end of PMLSM through grating scale,
Position and speed detection circuit is connected to digital I/O mouthfuls of dsp processor, and the output end connection IPM isolation drive of dsp processor is protected
The input terminal of protection circuit, IPM isolation drive protect the control signal of the output end connection IPM inversion unit of circuit.
4. a kind of PMLSM iterative learning control systems based on hybridization particle group optimizing according to claim 1, feature
Be: the control circuit is used to control the switching tube on-off in IPM inversion unit comprising dsp processor, current sample electricity
Road, position and speed sample circuit, IPM protective separation driving circuit.
5. a kind of PMLSM iterative learning control systems based on hybridization particle group optimizing according to claim 1, feature
It is: the dsp processor: according to position, speed and the current signal received, executes HPSO-ILC control algolithm, generate
Driving signal controls the switching tube on-off in IPM inversion unit；
Current sampling circuit: the sampled signal of Hall current sensor is converted into the level signal of 0-3V；
Position and speed sample circuit: it converts the collected rotor position speed signal of grating scale to can be identified by dsp processor
Digital quantity；
IPM protective separation driving circuit: it is generated not according to dsp processor in HPSO-ILC position control, PI speed control
With pwm signal driving IPM inversion unit work and protection IPM inversion unit.
6. a kind of PMLSM iterative learning control systems based on hybridization particle group optimizing according to claim 1, feature
It is: handles the process of signal in the dsp processor are as follows: after given PMLSM position signal, is detected with through grating scale
Actual position signal make the difference, generate position error signal, using position error signal as the input quantity of HPSO-ILC controller,
Current controling signal is calculated through HPSO-ILC controller, current controling signal generates pwm pulse sequence, pwm pulse through DSP
Sequence controls the conducting and shutdown of six IGBT of IPM inverter circuit, obtains meeting the three-phase alternating current needed, send to PMLSM
Control its movement.
7. the control method of the PMLSM iterative learning control systems according to claim 1 based on hybridization particle group optimizing,
It is characterized by comprising following steps:
Step 1: the given position PMLSM input signal；
Step 2: initialization hybridization particle swarm optimization algorithm: initialization HPSO algorithm determines population dimension D, the particle of algorithm
Group's inertia weight range [ωmin, ωmax], accelerator coefficient c1And c2, position range [xmin, xmax], velocity interval [vmin, vmax]、
Probability of crossover bc, hybridization pond ratio bs, maximum number of iterations M, optimum solution stop condition foc.Random initializtion particle speed, position
It sets, and using first generation particle as the personal best particle p of particlebest, adaptive value is as individual optimal adaptation value, in particle
The smallest adaptive value is used as overall situation optimum position g as global optimal adaptation value, positionbest；
Step 3: the particle of generation is assigned to the iterative learning controller gain alpha of PMLSM, beta, gamma and Butterworth respectively
Zero phase low-pass filter bandwidth fc；
Step 4: determining the output signal u of iterative learning controllerj+1(t).By given desired output signal xd(t), storage note
The control input u for the iteration j study recalledj(t), the reality output x of+1 iterative learning of jthj+1(t) and jth+1 time
Tracking error e when iterative learningj+1(t), it obtains the output signal of iteration controller and is sent into controlled device, wherein ej+1(t)
=xd(t)-xj+1(t)；
Step 5: more new particle individual optimal value and global optimum: correcting iterative learning controller gain using HPSO algorithm
α, beta, gamma and Butterworth zero phase low-pass filter bandwidth fc.The position and speed of each particle of population is updated,
And then calculate the adaptive value of each particle.
Step 5-1: updating the position and speed of population particle, and formula is as follows:
xI, j(k+1)=xI, j(k)+vI, j(k+1), j=1,2 ..., d (1)
vI, j(k+1)=ω vI, j(k)+c1r1[pI, j-xI, j(k)]+c2r2[pG, j-xI, j(k)] (2)
In formula: ω is inertia weight；c1And c2For normal number, referred to as accelerator coefficient；r1And r2It is the random number of [0,1]；xI, j(k)
For particle position；vI, jIt (k) is particle rapidity；pI, jFor particle personal best particle；pG, jFor population optimal location；
Step 5-2: the adaptive value of more each particle and individual optimal adaptation value pbestIf control effect is more excellent, it is just updated
Individual optimum position and individual optimal adaptation value；
Step 5-3: the adaptive value of more each particle and global optimal adaptation value gbestIf control effect is more excellent, it is just updated
Global optimum position and global optimal adaptation value；
Step 5-4: choosing the particle of specified quantity according to probability of crossover, and puts it into hybridization pond, and the particle in pond is random
Hybridization generates same number of filial generation particle two-by-two, and calculates the position and speed of filial generation.Parent is replaced using filial generation particle
Particle；
Wherein the adaptive value of particle is systematic error absolute time integral function ITAE, and calculation formula is as follows:
In formula: t is the time；E (t) is system given value and the deviation that system exports；
Step 5-5: it when algorithm reaches its stop condition, then stops search and exports result；Otherwise 5-1 step is returned to continue to search
Rope.
8. the control method of the PMLSM iterative learning control systems according to claim 1 based on hybridization particle group optimizing,
It is characterized by: the step 4 carries out as follows:
Real-time perfoming current sample and position sampling in step 4-1:PMLSM control system motion process；
Step 4-2:DSP processor is generated according to the current sampling data and position sampled data at current time and is carried out to PMLSM
The pwm signal of position control；
Step 4-2-1: HPSO-ILC position control is carried out to PMLSM according to position sampled data: desired locations and position are adopted
After sample data make the difference, position deviation is obtained, obtains desired speed after HPSO-ILC is calculated；
Step 4-2-2: PI speed control is carried out to PMLSM according to position sampled data: after the sampled data differential of position, being obtained
Actual speed, then after the resulting desired speed of step 4-2-1 and actual speed are made the difference, velocity deviation is obtained after PI is calculated,
Obtain expectation electric current；
Step 4-2-3: the pwm signal that position control is carried out to PMLSM is generated according to current sampling data: by step 4-2-2 institute
Expectation electric current make 2/3 transformation, then after transformation results and current sampling data are made the difference, obtain driving IPM inversion unit
Pwm signal executes step 4-3；
Step 4-3:PMLSM works according to pwm signal: IPM inversion unit makes according to the resulting pwm signal of step 4-2-3
PMLSM is run, real-time perfoming current sample and position sampling in PMLSM motion process；
Step 4-4: return step 4-2-2.
9. the control method of the PMLSM iterative learning control systems according to claim 1 based on hybridization particle group optimizing,
It is characterized by: the random value of speed is limited in [v in the step 2min, vmax], the random value of position is limited in [xmin,
xmax].Population inertia weight is in [ωmin, ωmax] successively decrease in range by linear relationship.
10. the controlling party of the PMLSM iterative learning control systems according to claim 1 based on hybridization particle group optimizing
Method, it is characterised in that: in the step 4, the output signal of iterative learning controller:
In formula: α, β and γ are respectively ratio, integral and the differential study gain parameter of ILC controller；
In the step 5-1: inertia successively decreases weight are as follows:
In formula: ωmaxIndicate inertia weight maximum value；ωminIndicate inertia weight minimum value；K indicates current iteration number；
In the step 5-4: the calculation formula of the position and speed of filial generation are as follows:
In formula: mx represents the position of parent particle；Nx represents the position of filial generation particle；The speed of mv expression parent particle；Nv table
Show the speed of filial generation particle；I is the random number between 0 to 1.Keep pbestAnd gbestIt is constant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810900809.8A CN109039173A (en) | 2018-08-09 | 2018-08-09 | A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810900809.8A CN109039173A (en) | 2018-08-09 | 2018-08-09 | A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109039173A true CN109039173A (en) | 2018-12-18 |
Family
ID=64633185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810900809.8A Pending CN109039173A (en) | 2018-08-09 | 2018-08-09 | A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109039173A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110829921A (en) * | 2019-11-15 | 2020-02-21 | 江南大学 | Iterative feedback setting control and optimization method for permanent magnet synchronous motor |
| CN111200378A (en) * | 2020-02-13 | 2020-05-26 | 广州大学 | Piezoelectric motor energy-saving control method based on iterative learning |
| WO2020181934A1 (en) * | 2019-03-11 | 2020-09-17 | 阿里巴巴集团控股有限公司 | Method and device for determining position of target object on the basis of particle swarm algorithm |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103544526A (en) * | 2013-11-05 | 2014-01-29 | 辽宁大学 | Improved particle swarm algorithm and application thereof |
| CN103560721A (en) * | 2013-11-16 | 2014-02-05 | 沈阳工业大学 | Device and method for controlling gantry numerical control milling machine through double line permanent magnet synchronous motors |
| CN104881512A (en) * | 2015-04-13 | 2015-09-02 | 中国矿业大学 | Particle swarm optimization-based automatic design method of ripple-free deadbeat controller |
| CN106773649A (en) * | 2016-12-21 | 2017-05-31 | 成都千嘉科技有限公司 | A kind of automatic control valve for gaseous fuel door intelligent control method based on PSO pid algorithms |
-
2018
- 2018-08-09 CN CN201810900809.8A patent/CN109039173A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103544526A (en) * | 2013-11-05 | 2014-01-29 | 辽宁大学 | Improved particle swarm algorithm and application thereof |
| CN103560721A (en) * | 2013-11-16 | 2014-02-05 | 沈阳工业大学 | Device and method for controlling gantry numerical control milling machine through double line permanent magnet synchronous motors |
| CN104881512A (en) * | 2015-04-13 | 2015-09-02 | 中国矿业大学 | Particle swarm optimization-based automatic design method of ripple-free deadbeat controller |
| CN106773649A (en) * | 2016-12-21 | 2017-05-31 | 成都千嘉科技有限公司 | A kind of automatic control valve for gaseous fuel door intelligent control method based on PSO pid algorithms |
Non-Patent Citations (3)
| Title |
|---|
| YI-CHENG HUANG; YI-WEI SU: "Study on Iterative Learning Control bandwidth tuning using particle swarm optimization technique for high precision motion", 《2014 INTERNATIONAL CONFERENCE ON INFORMATION SCIENCE, ELECTRONICS AND ELECTRICAL ENGINEERING》 * |
| 孙璐: "基于经验模态分解算法的直驱XY平台迭代学习控制", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 * |
| 蔡莎莎 等: "基于杂交粒子群算法的汽轮机调速系统参数辨识", 《电力学报》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020181934A1 (en) * | 2019-03-11 | 2020-09-17 | 阿里巴巴集团控股有限公司 | Method and device for determining position of target object on the basis of particle swarm algorithm |
| CN110829921A (en) * | 2019-11-15 | 2020-02-21 | 江南大学 | Iterative feedback setting control and optimization method for permanent magnet synchronous motor |
| CN111200378A (en) * | 2020-02-13 | 2020-05-26 | 广州大学 | Piezoelectric motor energy-saving control method based on iterative learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109039173A (en) | A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing | |
| CN101499753B (en) | Brushless servo control system and driving apparatus based on permanent magnet brushless DC motor | |
| CN103916062B (en) | A kind of vector controlled YE based on DSP | |
| CN103560722B (en) | A kind of permanent magnet linear synchronous motor controls device and method | |
| CN105353682B (en) | Aircraft power system experiment platform device | |
| CN101951182B (en) | Voltage compensation method of pulse width modulation output based on voltage equivalent | |
| CN107132759A (en) | A kind of linear electric motors driving FTS is based on ESO sliding formworks and improves repetitive control | |
| CN104113078A (en) | Photovoltaic direct drive system and control method thereof | |
| CN106911277A (en) | Control system for permanent-magnet synchronous motor based on matrix converter | |
| CN100595700C (en) | Reschedule decision making system of semiconductor manufacturing system | |
| CN104660130B (en) | A kind of error-tolerance type Instantaneous torque control asynchronous machine power generation system and its control method | |
| CN102263536A (en) | Stepping motor subdivided driving device for current vector constant amplitude control | |
| CN105676780B (en) | XY motion platform profile control apparatus based on fuzzy cerebellar model articulation controller | |
| CN105048896B (en) | A kind of brshless DC motor Direct Torque adaptive fuzzy control method | |
| CN113285481B (en) | Grid-connected converter inductance parameter online estimation method, prediction control method and system | |
| CN114172443B (en) | Permanent magnet motor driving system current sensor fault online diagnosis method | |
| CN104834219A (en) | PMLSM driven XY platform control method based on empirical mode decomposition, and system thereof | |
| CN201403064Y (en) | Brushless servo system and driving device based on permanent magnet brushless direct current motor | |
| CN106067744A (en) | A kind of novel intelligent motor controller | |
| CN110429835B (en) | LCL filtering-based RBFNN segmented online optimization passive control system and method | |
| CN104868761A (en) | Novel triangular wave comparison control method for three-phase PWM rectifier | |
| CN104065324B (en) | Three phase alternating current motor power driving controller based on preposing conversion device cascade three-level inverter | |
| CN105915145B (en) | A kind of permanent magnet linear synchronous motor control device and method | |
| CN105048508A (en) | Method and system for controlling T-type three-level single-phase grid-connected inverter | |
| CN102522948B (en) | Hybrid intelligent adjusting method of torque hysteresis width in DTC (Direct Torque Control) system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |
|
| RJ01 | Rejection of invention patent application after publication |