CN103780187B - Permanent magnet synchronous motor high-dynamic response current method and system - Google Patents

Abstract

This invention discloses a method and a system for permanent magnet synchronous motor high-dynamic response current control. The method comprises: preferably adopting an improved dutyfactor updating strategy, namely updating and calculating a new dutyfactor at the peak value of a calculator in first half control cycle and combining the fact that adopting various current inner ring adjustment strategies according to the characteristics of a q axis and a d axis, realizing feedback values following upper command by half a control cycle ahead of time; adopting predictive control in a torque current ring to connect the pure integration to eliminate the influence caused by inaccurate parameters and adopting PI control in exciting current to grantee the stability and easy reliability of the system. As a result, the invention can further shorten control delay and sensitivity of system parameters.

Description

Permagnetic synchronous motor high dynamic response current control method and system

Technical field

The present invention relates to a kind of permagnetic synchronous motor technology, particularly permagnetic synchronous motor high dynamic response current control side Method and system.

Background technology

Due to high efficiency, high power density, non-carbonate the advantages of, permagnetic synchronous motor obtains in high performance servo occasion Extensively apply, the application such as numerical control machining center, industrial robot driving requires quick electromagnetic torque response whole to ensure The high dynamic performance of individual system, therefore, the electric current loop dynamic characteristic directly related with electromagnetic torque has become measurement servo performance One of most important index.

Permagnetic synchronous motor is a multivariate, close coupling, nonlinear system, for the ease of research, often does as follows Assume：Three-phase stator winding is full symmetric；Ignore saturation unshakable in one's determination, disregard vortex and magnetic hystersis loss；There is no Damper Winding on rotor, Rotor every phase magnetic potential of air gap is in Sine distribution in space, and under synchronous rotating frame, PMSM stator voltage equation is：

u_d=Ri_d+L_ddi_d/dt-ωL_qi_q(1)

u_q=Ri_q+L_qdi_q/dt+ωL_di_d+ωψ_f(2)

In formula, ud, uq are stator d, q shaft voltages, and id, iq are stator currents, and R is stator resistance, and Ld, Lq are stator electricity Sense, ψ f is permanent magnet flux linkage, and ω is rotor angular rate.

As shown in figure 1, being the block diagram of existing permanent magnet synchronous motor vector control system, system is twin nuclei, respectively For speed outer shroud and current inner loop, to durface mounted permanent magnet synchronous motor, typically adopt the control strategy of id=0；In order that system Have dynamic property it is desirable to electric current loop has quick dynamic responding speed.

The general control method to permagnetic synchronous motor electric current loop mainly has PID (or PI) actuator, Hysteresis control, prediction Control etc., PID regulator structure is simple, reliable and stable, is most widely used at present in servosystem, PID regulator principle On be a kind of linear regulator, improve PID regulator gain can improve dynamic performance, but excessive gain can affect be System stability, brings hyperharmonic noise, is difficult to take into account the rapidity of response and stability, in high performance servo in practical application Using the electric current loop dynamic property required for PID regulator difficult to reach.

Hysteresis control rapidity is good, but this Bang-Bang control mode has that ripple is big, switching frequency is fixing etc. lacks Fall into, be not suitable for high performance control occasion, for permagnetic synchronous motor Digitizing And Control Unit, have scholar to control time delay from minimizing Angle carry " double sampled double update (double sampling anddouble update, DSDU) " strategy, in a cycle Interior double sampling electric current, updates PWM duty cycle twice, improves electric current dynamic responding speed, its cost is that amount of calculation doubles, to control The requirement of coremaking piece processing speed is higher, has scholar to adopt the mode of feedforward to realize d, q axle full decoupled dynamic with raising electric current Control performance.

PREDICTIVE CONTROL can realize the quick tracking to command signal non-overshoot, but it relies on the perfect number of controlled device Learn model, this is proposed with various improved methods, such as obtain the accurate mathematical of controlled device by the method for on-line parameter identification Model, eliminates the inaccurate impact to PREDICTIVE CONTROL effect of parameter, has scholar to be based on dead beat PREDICTIVE CONTROL principle, proposes a kind of Robust Current control algolithm, the stability when model parameter is inaccurate for the system that enhances, there is scholar then using the electric current loosening Deviation constraint condition and smooth output voltage Forecasting Methodology, system robustness when enhancing motor inductances parameter mismatch, but one Determine the dynamic property of PREDICTIVE CONTROL is reduced on degree, for Digitizing And Control Unit, based on track with zero error scheme, using list The single renewal of sampling can be by current loop control delay decrease to 2 controlling cycles.

Existing current control method, is illustrated in figure 1 permanent magnet synchronous motor vector control system block diagram：Ignore position ring In the case of, permagnetic synchronous motor vector controlled structure generally includes speed outer shroud and current inner loop, and speed detector is (such as Code-disc) obtain motor speed, it is compared feeding speed ring pi regulator afterwards with speed command, after overregulating, speed ring is defeated Go out the iqr that q shaft current instruction is in Fig. 1；D axle aspect, due to generally adopting id=0 control mode, therefore d shaft current instruction Idr is usually arranged as zero.Two current-orders relatively and are input to corresponding electricity with current feedback values iq, id of corresponding axle respectively Stream ring actuator, this two actuators export voltage uq, the ud needing to apply on corresponding axle more respectively.Such uq, ud voltage Calculate three-phase dutycycle through steps such as coordinate inverse transformations, eventually become suitable three-phase voltage waveform and be applied on motor, Realize the control to motor；The actuator (speed ring one, electric current loop two) adopting in the prior art is not necessarily intended to use PI To realize, they can be realized by other control methods, why be drawn as pi regulator, be because pi regulator in Fig. 1 Simply, easily realize, thus be widely used and be familiar with by majority.

Existing basic PMSM electric current loop forecast Control Algorithm, as shown in Fig. 2 illustrate electric current loop digital control when PWM updates sequential, and after DSP enters to become owner of interruption, detection obtains electric current currency (id (k), iq (k)) and angular velocity omega.Run speed Degree ring obtains given value of current value (idr, iqr), and further running current ring obtains the voltage vector of needs applying；Limited by hardware System, the calculated voltage vector (ud (k+1), uq (k+1)) of k-th controlling cycle (kTs) generally will wait until (k+1) individual control Apply again during cycle ((k+1) Ts) processed.Until this end cycle, electric current is possible to follow the tracks of the current-order in this cycle upper Signal (idr, iqr).

Control time delay be there is on digital control approach principle：Because in a sampling period (controlling cycle in other words), By the controlled quentity controlled variable (voltage ud, uq such as herein) to required for calculate for the sampling (the such as sampling to electric current id, iq herein) Certainly it is (no matter how soon DSP operation speed had it is always necessary to the time) requiring time for, this is in digital control approach principle Just unavoidable, additionally, this special occasions for motor control herein, interpretation process above is it has been mentioned that " uq, ud Voltage calculates three-phase dutycycle through steps such as coordinate inverse transformations, eventually becomes suitable three-phase voltage waveform and is applied to motor On, realize the control to motor." hardware limitation of this voltage applying mode and DSP determines controlled quentity controlled variable (i.e. electricity herein Pressure) applying must carry out in the controlling cycle of a complete length；, such as DSP enters current period taking Fig. 2 as a example (kTs) start after (i.e. " this cycle " mentioned above) to apply the voltage (ud (k), uq (k)) of certain value, this applies process must A complete sampling period or controlling cycle must be continued, can not arbitrarily change halfway；If it is intended to applying new voltage Value, such as (ud (k+1), uq (k+1)), then just can must carry out after next cycle starts, this controlled quentity controlled variable (i.e. voltage (ud (k+1), uq (k+1))) make it possible to the command value that value of feedback follows the tracks of this cycle in next end cycle；Therefore " electric current Instruction tracking time delay be two controlling cycles the soonest ", include unalterable cycle and play major control act on next Cycle.

Content of the invention

The invention aims to solving above-mentioned the deficiencies in the prior art and providing one kind can shorten current control and prolong When the time, reduce the permagnetic synchronous motor high dynamic response current control method of systematic parameter sensitivity and system simultaneously.

To achieve these goals, the permagnetic synchronous motor high dynamic response current control method designed by the present invention, bag Include the speed outer shroud to q axle and d axle to adjust and current inner loop regulation, in terms of q axle, Negotiation speed detection obtains motor speed, It is compared with speed command and sends into the regulation of speed outer shroud afterwards after overregulating, speed outer shroud output q shaft current instruction iqr；In d Axle aspect, due to generally adopting id=0 control mode, therefore d shaft current instruction idr is usually arranged as zero, two current-orders Iqr relatively and is input to corresponding current inner loop regulation after overregulating with current feedback values iq, id of corresponding axle respectively to idr Export voltage uq, the ud needing to apply on corresponding axle respectively, such uq, ud voltage calculates through steps such as coordinate inverse transformations Three-phase dutycycle, eventually becomes suitable three-phase voltage waveform and is applied on motor, realizes the control to motor, described dutycycle More New Policy using in front half controlling cycle at enumerator peak value update calculate new dutycycle, and combine to q, D axle takes the different current inner loop of suitable features to adjust control strategy, realizes command value on feedback value trace and shifts to an earlier date half Controlling cycle.

In order to further reduce control time delay, improve dynamic response, the current inner loop of described q axle adjusts control strategy Using fundamental forecasting control, and id no contributes to electromagnetic torque, not high to its dynamic performance requirements, in the electric current of described d axle Ring adjusts control strategy and adopts PI to adjust, the problems such as to avoid PREDICTIVE CONTROL to lead to steady-state error because parameter setting is inaccurate.

Because the use of plant model during PREDICTIVE CONTROL causes PREDICTIVE CONTROL more sensitive to systematic parameter, if ginseng Number somewhat has inaccurate, then the voltage calculating is not just the magnitude of voltage it is anticipated that intentionally getting, therefore basic Pure integral element in parallel in forecast Control Algorithm.

By shifting to an earlier date half controlling cycle, control with fundamental forecasting and be combined so that k-th controlling cycle (kTs) is counted The voltage vector (ud (k+1), uq (k+1)) obtaining can during (k+0.5) individual controlling cycle ((k+0.5) Ts) again Apply ", the computing formula of therefore described q shaft voltage is：

$\frac{(U_q\left(k\right)+2U_q(k+1))}{3}=R\frac{(i_q\left(k\right)+i_{q}r)}{2}+$

$L_q\frac{(i_{q}r-i_q\left(k\right))}{1.5T_s}+\mathrm{\ω}{L}_d\frac{(i_d\left(k\right)+i_{d}r)}{2}+\mathrm{\ω}{\mathrm{\ψ}}_f---\left(5\right)$

Wherein, (id (k), iq (k)) is the sampled value of current period, and (idr, iqr) is expected value, R is resistance, Lq is q The inductance of axle, ψ_fPermanent magnet flux linkage for described permagnetic synchronous motor.

A kind of system of permagnetic synchronous motor high dynamic response current control, including the speed inspection being coupled together by signal Survey device, speed PI adjusting means, electric current PI adjusting means, fundamental forecasting control device, inverter and coordinate transformation device, Adding between fundamental forecasting control device outfan and the input of inverter has the output signal of pure integration control device；Described Speed detector obtain the tach signal of permagnetic synchronous motor and be compared with previously given speed command signal and be superimposed Afterwards, it is connected with the input of speed PI adjusting means；Described coordinate transformation device is connected with permagnetic synchronous motor and obtains q, d axle Current feedback values iq, id；The outfan of described speed PI adjusting means is compared superposition with corresponding current feedback values iq Afterwards, it is connected with the input of pure integration control device；The outfan of described speed PI adjusting means also with fundamental forecasting control The input of device connects；The input of described fundamental forecasting control device is also associated with corresponding current feedback values iq；Described After the output end signal of the outfan of fundamental forecasting control device and pure integration control device is overlapped；Obtain the voltage of q axle Uq is simultaneously connected with the input of inverter；After d shaft current instruction idr corresponding current feedback values id is compared superposition, It is connected with the input of electric current PI adjusting means, obtain the voltage ud of d axle；And be connected with the input of inverter；Described inversion The outfan of device connects permagnetic synchronous motor；

Permagnetic synchronous motor high dynamic response current control method and system that the present invention obtains, first, propose one kind and change The dutycycle entered more New Policy, on the premise of unitary sampling single PWM fiducial value updates within the electronic power switch cycle, makes electricity Flow control time delay shortens to 1.5 controlling cycles further；Secondly, adopt PREDICTIVE CONTROL link integration in parallel in torque current ring The impact that device elimination brings because parameter is inaccurate, simultaneously in order to keep system stability and easy implementation, adopts in exciting current ring With PI control it is achieved that the present invention can shorten control time delay further, reduce systematic parameter sensitivity simultaneously.

Brief description

Fig. 1 is permanent magnet synchronous motor vector control system block diagram in prior art；

Fig. 2 is that in prior art, PWM duty cycle updates sequential chart；

Fig. 3 is the system block diagram of the high dynamic response current control for the present invention；

Fig. 4 is that the PWM duty cycle for the present invention updates sequential chart；

Fig. 5 is that R increases 50% two kind of control method simulation result figure；

Fig. 6 is that ψ f increases 50% two kind of control method simulation result figure；

Fig. 7 is the simulation result figure of two kinds of control method transient processes；

Fig. 8 is that resistance increases 50% two kind of control method experimental result picture；

Fig. 9 is the transient-wave figure of pi regulator control strategy；

Figure 10 is the transient-wave figure of fundamental forecasting control method；

Figure 11 is the transient-wave figure of control method proposed by the present invention.

In figure：Speed detector 1, speed PI adjusting means 2, electric current PI adjusting means 3, fundamental forecasting control device 4, Inverter 5, coordinate transformation device 6, pure integration control device 7.

Specific embodiment

The present invention is further described with reference to the accompanying drawings and examples.

Embodiment：

As shown in Figure 3, Figure 4, the permagnetic synchronous motor high dynamic response current control method that the present invention provides, including to speed Ring of being outside one's consideration is adjusted and current inner loop is adjusted, and in terms of q axle, Negotiation speed detection obtains motor speed, by itself and speed command ratio Send into speed outer shroud more afterwards and adjust after overregulating, speed outer shroud output q shaft current instruction iqr；In terms of d axle, due to usual Using id=0 control mode, therefore d shaft current instruction idr is usually arranged as zero, two current-orders iqr, idr respectively with phase Answer current feedback values iq, id of axle relatively and be input to corresponding current inner loop and adjust and export respectively on corresponding axle after overregulating Need voltage uq, the ud applying, such uq, ud voltage calculates three-phase dutycycle through steps such as coordinate inverse transformations, finally Become suitable three-phase voltage waveform to be applied on motor, realize the control to motor, the more New Policy of described dutycycle adopts Front half controlling cycle updates at enumerator peak value and calculates new dutycycle, and combination is taken to q, d axle and is suitable for respectively Adjust control strategy from the different current inner loop of feature, realize command value on feedback value trace and shift to an earlier date half controlling cycle；Described The current inner loop of q axle adjust control strategy and adopt fundamental forecasting to control, the current inner loop of described d axle adjusts control strategy and adopts PI is adjusted；The ultimate principle that fundamental forecasting controls is if the current period sampled value (id (k), iq (k)) with electric current and expectation Value (idr, iqr) is two points of the whole story, and acquiescence rotating speed keeps constant within the 2Ts time period, and formula (1) and formula (2) are carried out Discretization, obtains formula (3) and formula (4)；Due to entering after k-th controlling cycle, (id (k), iq (k)), (idr, iqr), (ud (k), uq (k)), ω etc. are known, therefore can obtain (ud (k+1), uq (k+1)), due to (ud (k+1), uq (k+1)) Solution procedure to meet permagnetic synchronous motor mathematical model be formula (1) and formula (2), therefore this magnitude of voltage can be in theory Motor d, q shaft current is adjusted to the expectation state (idr, iqr) from current state (id (k), iq (k)), and time delay is two and controls week Phase.

$\frac{(U_d\left(k\right)+U_d(k+1))}{2}=R\frac{(i_d\left(k\right)+i_{d}r)}{2}+$

$L_d\frac{(i_{d}r-i_d\left(k\right))}{2T_s}-\mathrm{\ω}{L}_q\frac{(i_q\left(k\right)+i_{q}r)}{2}---\left(3\right)$

$\frac{(U_q\left(k\right)+U_q(k+1))}{2}=R\frac{(i_q\left(k\right)+i_{q}r)}{2}+$

$L_q\frac{(i_{q}r-i_q\left(k\right))}{{2T}_s}+\mathrm{\ω}{L}_d\frac{(i_d\left(k\right)+i_{d}r)}{2}+{\mathrm{\ω\ψ}}_f---\left(4\right)$

By shifting to an earlier date half controlling cycle, the new dutycycle obtaining in front half controlling cycle is at enumerator peak value Update, no longer wait until that next cycle updates after starting again, and be combined with fundamental forecasting control, such way does not increase Any software and hardware expense, is only required in front half computation of Period and goes out new dutyfactor value, so response to current-order shifts to an earlier date Half controlling cycle, on the premise of not improving sample rate and PWM renewal frequency, makes total control delay decrease to 1.5 Controlling cycle, obtains shown in discrete voltage equation such as formula (5) according to above-mentioned formula (4),

$\frac{(U_q\left(k\right)+2{U_q}_{}(k+1))}{3}=R\frac{(i_q\left(k\right)+i_{q}r)}{2}+$ $L_q\frac{(i_{q}r-i_q\left(k\right))}{1.5T_s}+\mathrm{\ω}{L}_d\frac{(i_d\left(k\right)+i_{d}r)}{2}+{\mathrm{\ω\ψ}}_f---\left(5\right)$

Wherein, (id (k), iq (k)) is the sampled value of current period, and (idr, iqr) is expected value, R is resistance, Lq is q The inductance of axle, ψ_fPermanent magnet flux linkage for described permagnetic synchronous motor.

Because, in practical application, the permagnetic synchronous motor mathematical model parameter that fundamental forecasting control uses differs and fixes Really, fundamental forecasting control effect will be made not reach perfect condition；It is therefore desirable to the sensitivity to parameter that analysis fundamental forecasting controls, Because d shaft current is controlled using pi regulator, there is not parameter sensitivity sex chromosome mosaicism, for this only for q shaft torque amperometry, false If resistance arranges value has error delta R, the voltage Uq (k+1) that calculates by formula (4) should be and real as shown in formula (6) The electric current on border still presses the rule change of formula (4), such as shown in formula (7).

$\frac{(U_q\left(k\right)+{\hat{U}}_q(k+1))}{2}=(R+\mathrm{\ΔR})\frac{(i_q\left(k\right)+i_{q}r)}{2}$

$+L_q\frac{(i_{q}r-i_q\left(k\right))}{2T_s}+\mathrm{\ω}{L}_d\frac{(i_d\left(k\right)+i_{d}r)}{2}+{\mathrm{\ω\ψ}}_f---\left(6\right)$

$\frac{(U_q\left(k\right)+{\hat{U}}_q(k+1))}{2}=R\frac{(i_q\left(k\right)+{\hat{i}}_{q}r)}{2}$

$+L_q\frac{({\hat{i}}_{q}r-i_q\left(k\right))}{2T_s}+{\mathrm{\ωL}}_d\frac{(i_d\left(k\right)+i_{d}r)}{2}+{\mathrm{\ω\ψ}}_f---\left(7\right)$

Formula (7) subtracts formula (6), can derive following relation

$\frac{{\hat{i}}_{q}r-i_{q}r}{i_{q}r}=\frac{\frac{\mathrm{\ΔR}}{R}}{\frac{1}{2}+\frac{L_q}{2T_{s}R}}---\left(8\right)$

In the same manner, can derive that magnetic linkage ψ f and inductance L (thinking Ld=Lq=L) has q axle in the case of error delta ψ f and Δ L The error of electric current, respectively as shown in formula (9) and (10)

${\hat{i}}_{q}r-i_{q}r=\frac{\mathrm{\ω\Δ}{\mathrm{\ψ}}_f}{\frac{R}{2}+\frac{L_q}{2T_s}}---\left(9\right)$ ${\hat{i}}_{q}r-i_{q}r=\frac{\frac{\mathrm{\ΔL}}{L}(i_{q}r-i_q\left(k\right))}{1+\frac{{\mathrm{RT}}_s}{L}}---\left(10\right)$

Knowable to formula (8) to formula (10), the deviation of resistance will cause the steady-state error of q shaft current, the two relative Error is linear；Therefore the impact to electric current for the magnetic linkage deviation and motor operating state have direct relation, show public Formula comprises motor speed in (9), and rotating speed is higher, and the impact that magnetic linkage deviation is brought is also bigger, understands inductance deviation by formula (10) The impact bringing is mainly reflected in transient process, and excessive inductance will result in current overshoot.

Therefore in order to eliminate the error being likely to occur because parameter is inaccurate, pure integration ring in parallel in fundamental forecasting control Section, ensures the accurate tracking to command signal for the q shaft current using integral term to the accumulative action of error, in the same manner, changes to using The forecast Control Algorithm entering PWM more New Policy is that formula (5) is analyzed, shown in result such as formula (11) formula (13), As can be seen that the shortening of control time delay can reduce the sensitivity to parameter to resistance, magnetic linkage, but the impact of inductance error is substantially not Become.

$\frac{{\hat{i}}_{q}r-i_{q}r}{i_{q}r}=\frac{\frac{\mathrm{\ΔR}}{R}}{\frac{1}{2}+\frac{L_q}{1.5T_{s}R}}---\left(11\right)$

${\hat{i}}_{q}r-i_{q}r=\frac{\mathrm{\ω\Δ}{\mathrm{\ψ}}_f}{\frac{R}{2}+\frac{L_q}{1.5T_s}}---\left(12\right)$

${\hat{i}}_{q}r-i_{q}r=\frac{\frac{\mathrm{\ΔL}}{L}(i_{q}r-i_q\left(k\right))}{1+\frac{1.5R{T}_s}{2L}}---\left(13\right)$

A kind of system of permagnetic synchronous motor high dynamic response current control, including the speed inspection being coupled together by signal Survey device 1, speed PI adjusting means 2, electric current PI adjusting means 3, fundamental forecasting control device 4, inverter 5 and coordinate transform dress Put 6, adding between the input of fundamental forecasting control device 4 outfan and inverter 5 has the output of pure integration control device 7 Signal；Described speed detector 1 obtains the tach signal of permagnetic synchronous motor and previously given speed command signal enters After row compares superposition, it is connected with the input of speed PI adjusting means 2；Described coordinate transformation device 6 and permagnetic synchronous motor Connect current feedback values iq, the id obtaining q, d axle；The outfan of described speed PI adjusting means 2 and corresponding current feedback values After iq is compared superposition, it is connected with the input of pure integration control device 7；The outfan of described speed PI adjusting means 2 Also it is connected with the input of fundamental forecasting control device 4；The input of described fundamental forecasting control device 4 is also associated with corresponding Current feedback values iq；The output end signal of the outfan of described fundamental forecasting control device 4 and pure integration control device 7 is folded Plus after；Obtain the voltage uq of q axle and be connected with the input of inverter 5；D shaft current instructs the corresponding current feedback of idr After value id is compared superposition, it is connected with the input of electric current PI adjusting means 3, obtain the voltage ud of d axle；And with inverter 5 Input connect；The outfan of described inverter 5 connects permagnetic synchronous motor；Described speed PI adjusting means 2 is by speed The motor speed of the permagnetic synchronous motor that detection means 1 detects is compared with previously given speed command signal after being superimposed Signal carry out PI regulation, obtain q shaft current instruction iqr；Described electric current PI adjusting means 3 is by d shaft current instruction idr therewith Corresponding current feedback values id is compared the signal after superposition and carries out current control, obtains the voltage ud of d axle；Described substantially pre- Survey control device 4 quickly to be followed the tracks of q shaft current instruction iqr corresponding current feedback values i；Described pure integration control After the signal that q shaft current instruction iqr corresponding current feedback values iq is compared superposition is integrated by device 7, and with The signal that fundamental forecasting control device 4 is followed the tracks of is overlapped, and obtains the voltage uq of q axle；Described inverter 5 is by the voltage of d, q axle Become suitable three-phase voltage waveform after signal ud, uq conversion to be applied on motor, realize the control to motor.

Control method according to the present invention carries out the simulation experiment result：Build permagnetic synchronous motor in Simulink environment Vector Control Model is simultaneously emulated, by the parameter of electric machine as shown in table 1, controlling cycle 125 μ s, current command signal be arranged to as To simulate real electrical machinery rotating, specially boost phase time 0.375s shown in Fig. 5, amplitude is 7A, smooths fall through 0.25s To 1.75A, then start reversely to accelerate through the constant-speed operation stage of 0.25s, and so on；Load torque be 2N m, direction with Rotary speed direction is contrary；Rotary inertia summation is 0.08408kg.m2.

Table 1 parameter of electric machine

Tab.l Parameters of the PMSM

The d shaft current that q shaft current PREDICTIVE CONTROL uses is its sampled value, and therefore control effect and d shaft current is concrete Control method is unrelated；In the case of illustrating resistance R increase 50% as shown in Figure 5, two kinds of forecast Control Algorithm are to q shaft current Control effect, contrast Fig. 4 (a), (b) understand, the effect of pure integral element in parallel in the improvement PREDICTIVE CONTROL that the present invention obtains disappears The impact bringing except parameter is inaccurate.

As shown in Figure 6：Show is that in the case that magnetic linkage ψ f increases 50%, two kinds of forecast Control Algorithm are to q shaft current Control error, it will also be appreciated that the improved forecast Control Algorithm that the present invention obtains eliminates the impact that magnetic linkage error is brought, such as Fig. 7 Shown：A () and (b) is the transient-wave to q shaft current for two kinds of forecast Control Algorithm, current-order is at a time from -1.5A Step to 1.5A.In figure abscissa unit be system controlling cycle be 125 μ s it can be seen that the present invention obtain improved pre- Survey control method and half controlling cycle is advanced by the response of command signal, time delay is 1.5 controlling cycles, with theory analysis Result is coincide；With respect to the time delay of 2 controlling cycles of fundamental forecasting control method, the response to current command signal substantially adds Hurry up.

Control method according to the present invention carries out experiment to be proved：In an experiment using a durface mounted permanent magnet synchronous motor, Parameter is consistent with table 1, and using the DSP experiment porch of TMS320LF28335, dominant frequency 8kHz is controlling cycle 125 μ s, sets experiment Condition with emulate consistent, motor drag dynamometer machine simulation constant torque load, then experimental data by communicate by way of by DSP It is sent to computer.

As shown in figure 8, illustrating in the case that resistance increases 50%, the control to q shaft current of two kinds of forecast Control Algorithm Effect is it can be seen that this invention removes the impact that brings of parameter error.

Obtain illustrating pi regulator, fundamental forecasting control method as Fig. 9, Figure 10, Figure 11 are followed successively by according to experimental result And improve forecast Control Algorithm waveform controlled to the transient state of q shaft current, current-order the 0.0375s moment from -1.5A step to 1.5A.It is 125 μ s that each point of in figure represents a controlling cycle, it can therefore be seen that pi regulator occurs in that common overshoot And vibration.And fundamental forecasting control method could will respond to command signal after next cycle starts, time delay is two The individual cycle, and the forecast Control Algorithm of our present invention, due to employing new dutycycle more New Policy, terminates in current period When electric current reached the intermediate value of step signal, the tracking to instruction is rapider, therefore proves this according to emulation and experiment Controlling cycle can be effectively shortened in invention.

Claims (3)

1. a kind of permagnetic synchronous motor high dynamic response current control method, including adjusting to the speed outer shroud of q axle and d axle and electricity Stream internal ring is adjusted, and in terms of q axle, Negotiation speed detection obtains motor speed, it is compared with speed command and sends into outside speed afterwards Ring is adjusted, after overregulating, speed outer shroud output q shaft current instruction iqr；In terms of d axle, d shaft current instruction idr is set to Zero, two current-order iqr relatively and are input in corresponding electric current with current feedback values iq, id of corresponding axle respectively to idr Ring adjusts and exports voltage uq, the ud needing to apply on corresponding axle after overregulating respectively, and such uq, ud voltage is anti-through coordinate Shift step calculates three-phase dutycycle, and finally output three-phase voltage waveform is applied on motor, realizes the control to motor, its Feature is：Described calculate three-phase dutycycle through coordinate inverse transformation step and be utilized in front half controlling cycle counting Calculate new dutycycle at device peak value, and combination takes to q, d axle each different current inner loop to adjust control, realizes feedback On value trace, command value shifts to an earlier date half controlling cycle；

The current inner loop of described q axle is adjusted to control and is controlled using fundamental forecasting, and the current inner loop of described d axle adjusts to control and adopts PI is adjusted；

Pure integral element in parallel in fundamental forecasting control.

2. permagnetic synchronous motor high dynamic response current control method according to claim 1, is characterized in that：By in advance Half controlling cycle and fundamental forecasting control and are combined so that the calculated voltage vector of k-th controlling cycle (kTs) (ud (k+1), uq (k+1)) can apply during (k+0.5) individual controlling cycle ((k+0.5) Ts) again, described q shaft voltage Computing formula be：

$\begin{array}{c}\frac{\left(U_q\left(k\right)+2U_q\left(k+1\right)\right)}{3}=R\frac{\left(i_q\left(k\right)+i_{q}r\right)}{2}+\\ L_q\frac{\left(i_{q}r-i_q\left(k\right)\right)}{1.5T_s}+{\mathrm{\ωL}}_d\frac{\left(i_d\left(k\right)+i_{d}r\right)}{2}+{\mathrm{\ω\ψ}}_f\end{array}---\left(5\right)$

Wherein, (id (k), iq (k)) is the sampled value of current period, and (idr, iqr) is expected value, R is resistance, Lq is q axle Inductance, ψ f are the permanent magnet flux linkage of described permagnetic synchronous motor.

3. a kind of permagnetic synchronous motor high dynamic response current control system, including the velocity measuring dress being coupled together by signal Put (1), speed PI adjusting means (2), electric current PI adjusting means (3), fundamental forecasting control device (4), inverter (5) and coordinate Converting means (6), is characterized in that：Adding between the input of fundamental forecasting control device (4) outfan and inverter (5) has The output signal of pure integration control device (7)；Described speed detector (1) obtain the tach signal of permagnetic synchronous motor with After previously given speed command signal is compared superposition, it is connected with the input of speed PI adjusting means (2)；Described seat Mark converting means (6) is connected current feedback values iq, the id obtaining q, d axle with permagnetic synchronous motor；Described speed PI adjusting means (2) outfan is compared after superposition with corresponding current feedback values iq, connects with the input of pure integration control device (7) Connect；The outfan of described speed PI adjusting means (2) is also connected with the input of fundamental forecasting control device (4)；Described base The input of this prediction control device (4) is also associated with corresponding current feedback values iq；Described fundamental forecasting control device (4) After the output end signal of outfan and pure integration control device (7) is overlapped；Obtain q axle voltage uq and with inverter (5) Input connect；After d shaft current instruction idr corresponding current feedback values id is compared superposition, adjust with electric current PI The input of device (3) connects, and obtains the voltage ud of d axle；And be connected with the input of inverter (5)；Described inverter (5) Outfan connects permagnetic synchronous motor.