CN101977009A - Control method of sine wave linear motor for feed driving of high-precision numerical control machine - Google Patents

Abstract

The invention discloses a control method of a sine wave linear motor for feed driving of a high-precision numerical control machine, comprising the specific steps as follows: a motion trail tracking module transmits the generated displacement error E and the error change rate dE to a position control loop; the position control loop adopts a self-adaption fuzzy control algorithm, and the output speed amount of the fuzzy controller and the measured speed signals are compared and transmitted to a speed control loop; the speed control loop transmits the generated error signal and the d-axle component error signals to a current control loop; the current control loop controls the current output by an inverter by using a space vector PWM (Pulse-Width Modulation) technology with a PI (Proportional-Integral) regulator so as to control the output torque of the motor; a feedback signal detection module outputs the feedback signals to the current control loop and regulates the output torque of the motor. The invention combines self-adaption control, fuzzy control and vector control of an alternative-current motor together, and improves the stability, the accuracy and the dynamic response capacity of a linear servo motor.

Description

The high-precision numerical control machine feeding drives the control method with sinusoidal wave linear electric motors

Technical field

This invention relates to a kind of high-precision numerical control machine feeding and drives the control method of using linear servo-actuator, specifically a kind of vector control method that has the linear servo-actuator of Adaptive Fuzzy Control algorithm.

Background technology

At present, because the demand of digital control processing field aspect speed, acceleration, precision and impulse stroke is more and more higher, the novel numerical control lathe trends towards adopting linear electric motors as the feeding mode of feeding drive motors with alternative traditional " rotating servo motor+ball-screw " gradually.But, because linear electric motors directly are connected with process tool, the disturbance of external loading, frictional force can directly be passed to linear electric motors, consider again linear electric motors when work various parameters variation and the influence of the uncertain factor of control system, this just has higher requirement to the control of linear electric motors.

Because sinusoidal wave linear permanent-magnet motor has high power density and high fast, accurately control performance, make it become feeding and drive the first-selection of using linear electric motors.Research for this motor control method is a focus all the time, for example adopt robust control technique to overcome the various disturbances of control system, but this method is based on accurate model parameter control method, and the model of linear electric motors has many uncertainties and some difficult parameters to measure.Someone considers to adopt traditional self-adaptation control method, comprises model reference adaptive control, self-regulation control and Sliding mode variable structure control etc., and the advantage of this class control method is the dependence that can break away from the motor accurate parameters.Intelligent control technology such as fuzzy control, ANN Control etc. are not studied the person in recent years and are applied in the servo-drive control, and have obtained desirable motion control performance.But the technology that the vector control that also adaptive control, fuzzy control has not been reached alternating current machine at present combines is used for Electric Machine Control.

Summary of the invention

In order to overcome under the externally interference effect that occurs in the direct feeding driving control system problems such as system's instability, control precision difference and dynamic response capability be low.The control method that the purpose of this invention is to provide a kind of sinusoidal wave linear servo-actuator based on the adaptive fuzzy algorithm, this control method combines the vector control that adaptive control, fuzzy control have reached alternating current machine, can overcome various disturbances and uncertain influence effectively, improve stability, accuracy and the dynamic response capability of linear servo-actuator control system.

The objective of the invention is to be achieved through the following technical solutions:

A kind of high-precision numerical control machine feeding drives the control method with sinusoidal wave linear electric motors, it is characterized in that: this control method is controlled by movement locus tracking module, Position Control ring, speed control ring, current regulator and feedback signal detection module offset of sinusoidal ripple linear electric motors, specifically may further comprise the steps:

1) the movement locus tracking module compares the displacement reference instruction of generation and the real-time displacement signal of collection, and sends displacement error E and the error rate dE that produces to the Position Control ring;

2) the Position Control ring has adopted the Adaptive Fuzzy Control algorithm, earlier with two input variable obfuscations, according to established fuzzy rule and utilize the product inference machine to carry out fuzzy reasoning, can obtain the output of fuzzy controller behind the ambiguity solution; Simultaneously, the displacement reference instruction also relatively through the real-time displacement signal of a second order reference model and collection, the displacement error of generation is used for the ambiguity solution parameter in the blur correction mode controller, to eliminate the dynamic uncertainty and the load variations of controlling object; The output speed amount of fuzzy controller and the rate signal of measurement compare and are transferred to the speed control ring;

3) the speed control ring adopts simple P adjuster, its output variable is the q axle component of current of electric, this component compares as the q axle component with reference to the amount winding current next with gathering conversion, and the error signal of generation is transferred to current regulator with the error signal of d axle component;

4) current regulator adopts the space vector PWM modulation technique that has pi regulator, two current components of d, q are respectively through being output as the component of voltage of two axles behind the PI controller, through exporting three-phase voltage after the coordinate transform based on the stator coordinate system, after the space vector PWM module is produced 6 road PWM output, the size of control inverter output current, thereby the output torque of control motor;

5) displacement signal of the linear electric motors that record of feedback signal detection module phase current signal that current transformer is measured and straight line photoelectric encoder is transferred to current regulator, and the output torque of motor is adjusted.

Among the present invention, be provided with the control signal feedback module in current regulator, the displacement signal of the linear electric motors that phase current signal that this feedback signal is measured by current transformer and straight line photoelectric encoder record is formed.

Adopt point-to-point (PTP) position control method in the movement locus tracking module.Fuzzy controller has adopted the average ambiguity solution method in monodrome fuzzy device, product inference method and center in the Position Control ring; Reference model adopts second-order system, and correction mechanism adopts the gradient descent method.

The present invention is the combination that adaptive control, fuzzy control have reached the vector control of alternating current machine.Beneficial effect of the present invention is as follows:

1, the space vector of voltage PWM method that current regulator adopts among the present invention can effectively reduce the switching loss of device for power switching, has improved the dynamic response capability of system.

2, the adaptive fuzzy control method that the position control ring adopts among the present invention can suppress various disturbances and uncertain influence to control system effectively, has improved the stability and the accuracy of servo-control system.

3, adopt the feeding mode of the direct feeding type of drive replacement of sinusoidal wave linear electric motors traditional " rotating servo motor+ball-screw " among the present invention, can improve speed, acceleration, dynamic response capability and the accuracy of servo system greatly.

The present invention combines the vector control that adaptive control, fuzzy control have reached alternating current machine, can overcome various disturbances and uncertain influence to control system effectively, has improved stability, accuracy and the dynamic response capability of linear servo-actuator.Being specially adapted to feed drive unit adopts in the high accuracy servo system of high-speed numeric control lathe of linear electric motors.

Description of drawings

Fig. 1 is a control block diagram of the present invention.

Fig. 2 is the graph of a relation between rest frame and the synchronous rotating frame.

Fig. 3 is the space vector of voltage realization flow figure based on DSP.

Fig. 4 is the fuzzy controller schematic diagram of structure.

Embodiment

A kind of high-precision numerical control machine feeding of the present invention drives the control method with sinusoidal wave linear electric motors, specifically may further comprise the steps:

The movement locus tracking module compares the displacement reference instruction of generation and the real-time displacement signal of collection, and sends displacement error E and the error rate dE that produces to the Position Control ring;

The Position Control ring has adopted the Adaptive Fuzzy Control algorithm, earlier with two input variable obfuscations, according to established fuzzy rule and utilize the product inference machine to carry out fuzzy reasoning, can obtain the output of fuzzy controller behind the ambiguity solution; Simultaneously, the displacement reference instruction also relatively through the real-time displacement signal of a second order reference model and collection, the displacement error of generation is used for the ambiguity solution parameter in the blur correction mode controller, to eliminate the dynamic uncertainty and the load variations of controlling object; The output speed amount of fuzzy controller and the rate signal of measurement compare and are transferred to the speed control ring;

The speed control ring adopts simple P adjuster, its output variable is the q axle component of current of electric, this component compares as the q axle component with reference to the amount winding current next with gathering conversion, and the error signal of generation is transferred to current regulator with the error signal of d axle component;

Current regulator adopts the space vector PWM modulation technique that has pi regulator, two current components of d, q are respectively through being output as the component of voltage of two axles behind the PI controller, through exporting three-phase voltage after the coordinate transform based on the stator coordinate system, after the space vector PWM module is produced 6 road PWM output, the size of control inverter output current, thereby the output torque of control motor;

The displacement signal of the linear electric motors that phase current signal that the feedback signal detection module is measured current transformer and straight line photoelectric encoder record is transferred to current regulator, and the output torque of motor is adjusted.

As shown in Figure 1, controlled device sine wave permanent magnet linear electric motors in the dynamic mathematical models of rotating under d, the q axis coordinate system synchronously are:

$\left\{\begin{array}{c}\frac{{\mathrm{di}}_d}{\mathrm{dt}}=-\frac{R_s}{L_d}{i}_d+\frac{\mathrm{\π}}{\mathrm{\τ}}\frac{L_q}{L_d}\frac{{\mathrm{dx}}_p}{\mathrm{dt}}{i}_q+\frac{1}{L_d}{v}_d\\ \frac{{\mathrm{di}}_q}{\mathrm{dt}}=-\frac{R_s}{L_q}{i}_q-\frac{\mathrm{\π}}{\mathrm{\τ}}\frac{L_d}{L_q}\frac{{\mathrm{dx}}_p}{\mathrm{dt}}{i}_d-\frac{\mathrm{\π}}{\mathrm{\τ}}\frac{{\mathrm{\λ}}_f}{L_q}\frac{{\mathrm{dx}}_p}{\mathrm{dt}}+\frac{1}{L_q}{v}_q\end{array}\right.---\left(1\right)$

The electromagnetic push that produces is:

$F_e=-\frac{3}{2}\frac{\mathrm{\π}}{\mathrm{\τ}}((L_d-L_q)i_d+{\mathrm{\λ}}_f)i_q---\left(2\right)$

If consider load resistance, the dynamic motion equation of permanent-magnetism linear motor is:

$F_e-F_L=M_m\frac{d^2{x}_{\mathrm{<figure-callout\; id="2"\; label="p\; dt"\; filenames="HSA00000300097900022.png"\; state="\{\{state\}\}">p</figure-callout>}}}{{\mathrm{dt}}^{}}+B_m\frac{{\mathrm{dx}}_p}{\mathrm{dt}}---\left(3\right)$

As shown in Figure 1, current regulator comprises that two pi regulators, coordinate transform (comprise Clark, Clark ^-1, Park, Park ^-1Four coordinate transforms), module such as SVPWM.By transform vector to highly the coupling highly nonlinear linear electric motors inductance characteristic carry out decoupling zero, adopted the vector control of id=0, as long as the direction that makes stator current like this is all the time on the q direction of principal axis, just can directly control the thrust of motor by the size of control stator current, at this moment, formula (2) becomes:

$F_e=-\frac{3}{2}\frac{\mathrm{\&pi;}}{\mathrm{\&tau;}}{\mathrm{\&lambda;}}_f{i}_q=K_t{i}_q---\left(4\right)$

Static 3 phases, 2 phase coordinate systems and the location diagram that rotates 2 phase coordinate systems synchronously are as shown in Figure 2.Fig. 3 is based on the realization flow figure of the space voltage vector of DSP.

As shown in Figure 1, the Position Control ring comprises fuzzy controller, adaptively correcting mechanism, reference model etc.The structure of fuzzy controller as shown in Figure 4, tracking error e and error rate de are defined as:

$\left\{\begin{array}{c}e\left(k\right)=x_p^{*}\left(k\right)-x_p\left(k\right)\\ \mathrm{de}\left(k\right)=e\left(k\right)-e(k-1)\end{array}\right.---\left(5\right)$

With tracking error e and error rate de input, speed v as fuzzy controller _fAs the output of fuzzy controller, e, de, v _fAll carried out normalized, normalization coefficient is respectively K _e, K _De, K _Vf, select the symmetric triangular shape function as membership function, determine the language value of input/output variable.

Fuzzy controller has adopted the average ambiguity solution method in monodrome fuzzy device, product inference method and center.

Reference model has adopted second-order system, and its transfer function is:

$\frac{x_m\left(s\right)}{x_p^{*}\left(s\right)}=\frac{{\mathrm{\&omega;}}_n^2}{{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="HSA00000300097900022.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+2\mathrm{\&zeta;}{\mathrm{\&omega;}}_{n}s+{\mathrm{\&omega;}}_n^2}---\left(6\right)$

According to the Electric Machine Control requirement, this second-order system be non-overshoot, response fast and steady-state error be zero system.Select damping coefficient ξ=1, then at t of following rise time of above-mentioned constraint _rWith natural frequency ω _nBetween relation as follows:

$(1-{\mathrm{\&omega;}}_n{t}_r)e^{-{\mathrm{\&omega;}}_n{t}_r}=0.1---\left(7\right)$

In case t _rGiven, just can obtain ω _nAnd can obtain the discrete form of reference model transfer function with hyperbolic transformation:

$\frac{x_m\left(z^{-1}\right)}{x_p^{*}\left(z^{-1}\right)}=\frac{a_0+a_1{z}^{-1}+a_2{z}^{-2}}{1+b_1{z}^{-1}+b_2{z}^{-2}}---\left(8\right)$

Difference equation is:

$x_m\left(k\right)=-b_1{x}_m(k-1)-b_2{x}_m(k-2)-a_0{x}_p^{*}\left(k\right)+a_1{x}_p^{*}(k-1)+a_2{x}_p^{*}(k-2)---\left(9\right)$

Correction mechanism has adopted the gradient descent method, by minimizing square ambiguity solution parameter of coming the blur correction mode controller of reference model output and electric mover displacement difference:

$J(k+1)=\frac{1}{2}{e}_m{(k+1)}^2=\frac{1}{2}{[x_m(k+1)-x_p(k+1)]}^2---\left(10\right)$

$\mathrm{\&Delta;}{c}_{m,n}(k+1)\&Proportional;-\frac{\&PartialD;J(k+1)}{\&PartialD;c_{m,n}\left(k\right)}=-\mathrm{\&alpha;}\frac{\&PartialD;J(k+1)}{\&PartialD;c_{m,n}\left(k\right)}---\left(11\right)$

Point-to-point Position Control mode has been adopted in the tracking of linear motor rotor movement locus, and the instantaneous position instruction during each sampling period can have the following step to obtain:

(1) calculate time time of removing accelerator earlier, calculate the total time of considering accelerator again:

T ₁＝Δx _p/V

T _o＝MAX(T ₁，T _acc)+T _acc

(2) ask for the multiple in sampling period: N '=[T _Acc/ t _d] and N=[T _o/ t _d]

(3) upgrade maximal rate: V '=Δ x _p/ (T ' _o-T ' _Acc)

(4) ask for acceleration: A=V '/T ' _Acc

(5) calculate the instantaneous position instruction:

A) accelerating region:

B) at the uniform velocity distinguish:

C) deceleration area:

The present invention combines the vector control that adaptive control, fuzzy control have reached alternating current machine, can overcome various disturbances and uncertain influence to control system effectively, has improved stability, accuracy and the dynamic response capability of linear servo-actuator.

Claims (3)

1. a high-precision numerical control machine feeding drives the control method with sinusoidal wave linear electric motors, it is characterized in that: this control method is controlled by movement locus tracking module, Position Control ring, speed control ring, current regulator and feedback signal detection module offset of sinusoidal ripple linear electric motors, specifically may further comprise the steps:

1) the movement locus tracking module compares the displacement reference instruction of generation and the real-time displacement signal of collection, and sends displacement error E and the error rate dE that produces to the Position Control ring;

2) the Position Control ring has adopted the Adaptive Fuzzy Control algorithm, earlier with two input variable obfuscations, according to established fuzzy rule and utilize the product inference machine to carry out fuzzy reasoning, can obtain the output of fuzzy controller behind the ambiguity solution; Simultaneously, the displacement reference instruction also relatively through the real-time displacement signal of a second order reference model and collection, the displacement error of generation is used for the ambiguity solution parameter in the blur correction mode controller, to eliminate the dynamic uncertainty and the load variations of controlling object; The output speed amount of fuzzy controller and the rate signal of measurement compare and are transferred to the speed control ring;

3) the speed control ring adopts simple P adjuster, its output variable is the q axle component of current of electric, this component compares as the q axle component with reference to the amount winding current next with gathering conversion, and the error signal of generation is transferred to current regulator with the error signal of d axle component;

4) current regulator adopts the space vector PWM modulation technique that has pi regulator, two current components of d, q are respectively through being output as the component of voltage of two axles behind the PI controller, through exporting three-phase voltage after the coordinate transform based on the stator coordinate system, after the space vector PWM module is produced 6 road PWM output, the size of control inverter output current, thereby the output torque of control motor;

5) displacement signal of the linear electric motors that record of feedback signal detection module phase current signal that current transformer is measured and straight line photoelectric encoder is transferred to current regulator, and the output torque of motor is adjusted.

2. high-precision numerical control machine feeding according to claim 1 drives the control method with sinusoidal wave linear electric motors, it is characterized in that: adopt point-to-point position control method in the movement locus tracking module.

3. high-precision numerical control machine feeding according to claim 1 drives the control method with sinusoidal wave linear electric motors, and it is characterized in that: fuzzy controller has adopted the average ambiguity solution method in monodrome fuzzy device, product inference method and center in the Position Control ring; Reference model adopts second-order system, and correction mechanism adopts the gradient descent method.

Images