阅读说明：本技术 一种抑制开关磁阻直线电机电磁力脉动的方法 (Method for inhibiting electromagnetic force pulsation of switched reluctance linear motor ) 是由 陈昊 华孟玉 田嘉成 刘劲夫 刘闯 张珂 巩士磊 于 2021-03-26 设计创作，主要内容包括：本发明公开了一种抑制开关磁阻直线电机电磁力脉动的方法,属于电机控制领域。由于SRLM的非线性特性、双凸极结构和纵向端部效应,SRLM在运行过程中会产生力脉动,引起振动和噪声,限制了SRLMs的发展。为了抑制开关磁阻直线电机电磁力脉动,本发明以三相双侧SRLM为原型,以RT-LAB为控制器。首先对SRLM下一个周期的电磁力进行预测,然后利用模糊调节器使预测值跟随PI调节器的速度闭环给出的参考值。通过对电磁力的预测,可以有效地解决电机负电磁力问题。同时,电磁力的预测也改善了电机的换向问题。模糊控制非常适合于难以用数学模型描述的SRLM。最后对系统的负载能力进行测试,发现在增加负载后,系统仍然可以抑制电磁力脉动。(The invention discloses a method for inhibiting electromagnetic force pulsation of a switched reluctance linear motor, and belongs to the field of motor control. Due to the nonlinear characteristics, double-salient structure and longitudinal end effects of SRLMs, SRLMs can produce force pulsations during operation, causing vibration and noise, limiting the development of SRLMs. In order to inhibit electromagnetic force pulsation of the switched reluctance linear motor, the invention takes a three-phase double-sided SRLM as a prototype and an RT-LAB as a controller. The electromagnetic force of the next period of the SRLM is predicted, and then the predicted value is made to follow a reference value given by a speed closed loop of a PI regulator by using a fuzzy regulator. By predicting the electromagnetic force, the problem of negative electromagnetic force of the motor can be effectively solved. Meanwhile, the prediction of the electromagnetic force also improves the commutation problem of the motor. Fuzzy control is well suited for SRLMs that are difficult to describe with mathematical models. Finally, the load capacity of the system is tested, and the system can still restrain electromagnetic force pulsation after the load is increased.)

1. A method for inhibiting electromagnetic force pulsation of a switched reluctance linear motor is characterized by comprising the following steps: the core part of the control algorithm consists of two parts, one part is predictive control, the predictive control has the advantages of good stability, easy modeling, capability of effectively processing multivariable problems and the like, and meanwhile, the stability of the system can be improved, and the method is suitable for motor control. The other part is fuzzy control, which does not need precise mathematical model and is suitable for the nonlinear characteristic of the switched reluctance linear motor. To make predictive control more compatible with fuzzy control, a control table for fuzzy control is presented herein. And the magnetic linkage value and the electromagnetic force value adopt a table look-up method, so that the operation burden of the system is reduced.

2. The predictive control described in claim 1, starting from a mathematical model of single-step prediction, designed by the voltage balance equation:

where U is the phase voltage, i is the phase current, R is the winding resistance, ψ is the flux linkage, and v is the speed of the mover, the current derivative with time can be obtained from equation (2) according to (1):

discretizing equation (2) to obtain equation (3)

Where i (k +1) is the current for the next control period and Ts is the control period. x (k) is the relative position of the movers.

The current predicted value for the next control loop can be expressed as:

in order to avoid a large deviation between the predicted value and the output value, the output value and the predicted value need to be compared, and a second-order Runge-Kutta method is used for keeping the accuracy of prediction. i (k +1) can be expressed as:

the velocity is assumed to be a fixed value during a control period. The relative position of the movers in the next control cycle can be expressed as:

when the movers move to the end of each mover cycle, the relative position of the next cycle may overflow, so the remainder of the mover cycle is performed on x (k + 1). And inputting the predicted value of the current and the position of the motor in the next cycle into a lookup table to obtain the predicted value of the electromagnetic force in the next control cycle. Thus, when negative electromagnetic force or electromagnetic force fluctuation occurs in a certain phase of the motor, the system can predict the negative electromagnetic force or electromagnetic force fluctuation and make corresponding adjustment.

3. The fuzzy control of claim 1, wherein: a typical fuzzy control system includes a fuzzifier, a fuzzy rule base, a fuzzy inference engine, and a defuzzifier. The input to the fuzzy controller is the deviation of the reference electromagnetic force from the predicted electromagnetic force, denoted by E, and the derivative of the deviation with respect to time, denoted by EC. The output of the fuzzy governor is a fuzzy quantity, denoted by C. First, a scaling transform is performed to convert the input values into the required domain, i.e., E and EC into discrete integers between [ -6,6 ]. Another key factor in fuzzy control is the choice of membership functions. To describe the values more accurately, two different membership functions are used for E, EC and C, and the reason for using the two different membership functions is that when the predicted value of the electromagnetic force of a certain phase is a negative value or the electromagnetic force suddenly changes, the fuzzy regulator needs to change the reference current value in time to suppress the electromagnetic force ripple. In order to prevent a drastic change in the reference current value and improve the stability of the system, it is necessary to appropriately reduce the sensitivity of the fuzzy regulator to electromagnetic force deviation.

4. Fuzzy control as described in claim 3, on the basis of which specific parameters of the fuzzy control are selected, characterized in that: the optimal k1, k2, and k3 values need to be determined. First, for the selection of k3, too small a selection of k3 will result in system dynamic response process side length, and too large a selection of k3 will result in system oscillation. Since the maximum value in the control table is 5.35, after multiplying by the multiplier k3, the maximum current which is lower than the maximum current allowed by the switched reluctance linear motor is taken as k3, which is 2.5. For the selection of k1, k2, different combinations of k1, k2 are input in the Simulink fuzzy module, resulting in the combination of k1, k2 that minimizes the pulsation of the electromagnetic force. For the data obtained from the simulation, k1 was taken as 0.1 and k2 was taken as 0.1.

5. After the fuzzy control specific parameters are determined as described in claim 4, in order to make the electromagnetic force more continuous and give more adjustment opportunities to the predictive fuzzy control, appropriate system parameters should be determined so as to appropriately increase the conduction range of the MOSFET. In order to find the optimal conduction range, a switched reluctance linear motor driving system model is established in Simulink. Different switch positions are combined into the system to minimize electromagnetic force ripple, thereby determining the switch position of the system. When the pulsation of the electromagnetic force is minimum, the switch positions are 1mm and 27mm, respectively. Therefore, 1mm and 27mm were selected as the on and off positions of the system.

6. The design of a switched reluctance linear motor controller as claimed in claim 1, which is validated for its effect in suppressing torque ripple after selection of appropriate parameters. The method is characterized in that: the suppression effect of the switched reluctance linear electromagnetic force pulsation was verified from different control methods, respectively. Through experimental comparison, the pulsation (96.8%) of the electromagnetic force of the fuzzy control is predicted to be lower than that of the fuzzy control (145.7%) and the PI control (163.4%). This is because the control of the electromagnetic force tends to lag the current and flux linkage by one control cycle under conventional control methods. By predicting the electromagnetic force, the electromagnetic force can be controlled in advance, and a better control effect is achieved. Although the control method has a certain relieving effect on the longitudinal end effect of the SRLM, the influence of the longitudinal end effect still exists. Optimizing the motor body design can solve this problem.

Technical Field

The invention relates to a method for inhibiting electromagnetic force pulsation of a switched reluctance linear motor, in particular to a prediction fuzzy control algorithm of the switched reluctance linear motor.

Background

The switched reluctance linear motor is a special linear motor, and the simple structure and the excellent characteristics thereof gradually attract the attention of experts and scholars. However, the development of the switched reluctance linear motor is limited because the switched reluctance linear motor is affected by the longitudinal end effect and the low motion frequency, and generally has larger force pulsation than the rotary motor. In the last two decades, experts and scholars at home and abroad begin to carry out deep theoretical research and performance practice on the switched reluctance linear motor, and the development of related key technologies enables the switched reluctance linear motor to become one of linear motors which can be selected from linear driving systems.

Switched reluctance linear motors have a series of unique advantages: (a) the rotor part has no coil and no brush, the whole motor has no permanent magnet, and the rotor part can be suitable for application occasions with severe environment, wherein (b) the starting current is small, but the starting torque is large (c) the regenerative braking and energy feedback can be realized, the energy waste is reduced, and (d) the controllable parameters in the system are more, so that the control method is flexible, and the possibility is provided for high-performance control research. Therefore, the switched reluctance linear motor is widely applied to traction, rail transit and direct drive systems. However, due to the nonlinear characteristics, the double salient pole structure and the longitudinal end effect of the switched reluctance linear motor, the switched reluctance linear motor may generate electromagnetic force pulsation during operation, causing vibration and noise, and limiting the development of the switched reluctance linear motor. Two means are mainly adopted for reducing electromagnetic force pulsation of the linear motor, one means is to improve the motor structure, and the other means is to optimize a control algorithm. However, the current research is mainly focused on the rotating electrical machine, and the research on the electromagnetic ripple suppression control method of the switched reluctance linear motor is not sufficient. The invention aims to provide a control algorithm capable of effectively inhibiting electromagnetic force torque pulsation of a switched reluctance linear motor so as to further improve the control performance of the SRLM.

Disclosure of Invention

In view of the above problems, the present invention provides a method for suppressing electromagnetic force ripple of a switched reluctance linear motor.

In order to achieve the technical purpose, the invention adopts the following technical scheme to realize:

the invention is that the three-phase asymmetric half-bridge circuit is adopted as a topological structure of the power converter aiming at the overall composition of the control system. And a three-phase 6/4 switched reluctance linear motor was prototyped. The three-phase asymmetric half bridge can supply power to the phase winding in the switching-on stage and can feed energy back to the direct-current power supply in the afterflow stage. The structure can make each phase work independently, has good continuous flow channel, can fully utilize the rated voltage of the switch tube, and can realize flexible control of each phase. Therefore, the asymmetric half-bridge structure is widely used.

In the aspect of control algorithm, on the basis of the angle position control method, a prediction module and a fuzzy adjustment module are added. The angle position control means adjusting exciting current by changing an on angle and an off angle of a motor, and then controlling output voltage. The fuzzy adjustment module adopts fuzzy control, does not need an accurate mathematical model for the fuzzy control, and is suitable for the nonlinear characteristic of the SRLM. In terms of the design of the prediction module, the voltage balance equation:

where U is the phase voltage, i is the phase current, R is the winding resistance, ψ is the flux linkage, and v is the speed of the mover, the current derivative with time can be obtained from equation (2) according to (1):

discretizing equation (2) to obtain equation (3)

Where i (k +1) is the current for the next control period and Ts is the control period. x (k) is the relative position of the movers. The current predicted value for the next control loop can be expressed as:

in order to avoid a large deviation between the predicted value and the output value, the output value and the predicted value need to be compared, and a second-order Runge-Kutta method is used for keeping the accuracy of prediction. i (k +1) can be expressed as:

the velocity is assumed to be a fixed value during a control period. The relative position of the movers in the next control cycle can be expressed as:

when the movers move to the end of each mover cycle, the relative position of the next cycle may overflow, so the remainder of the mover cycle is performed on x (k + 1).

And inputting the predicted value of the current and the position of the motor in the next cycle into a lookup table to obtain the predicted value of the electromagnetic force in the next control cycle. Thus, when negative electromagnetic force or electromagnetic force fluctuation occurs in a certain phase of the motor, the system can predict the negative electromagnetic force or electromagnetic force fluctuation and make corresponding adjustment.

Has the advantages that:

the invention takes a three-phase double-sided SRLM as a prototype and an RT-LAB as a controller. In order to suppress the electromagnetic force pulsation of the SRLM, the electromagnetic force of the next cycle of the SRLM is first predicted, and then the predicted value is made to follow the reference value given by the speed closed loop of the PI regulator using the fuzzy regulator. By predicting the electromagnetic force, the problem of negative electromagnetic force of the motor can be effectively solved, and the method is suitable for the low-speed SRLM. Meanwhile, the prediction of the electromagnetic force also improves the commutation problem of the motor. Fuzzy control is well suited for SRLMs that are difficult to describe with mathematical models. Through experimental comparison, the pulsation (96.8%) of the electromagnetic force of the fuzzy control is predicted to be lower than that of the fuzzy control (145.7%) and the PI control (163.4%). The loading capacity of the system was then tested and it was found that the system still suppresses the electromagnetic force ripple after increasing the load.

Drawings

Fig. 1 is a system prototype and hardware platform of the present invention.

Fig. 2 is a topology diagram of the power converter of the present invention.

Fig. 3 is a system configuration diagram of the algorithms of the present invention.

FIG. 4 is a diagram of the structure of the fuzzy controller of the present invention and the relationship between K1, K2 and the pulsation of the electromagnetic force.

Fig. 5 is an illustration of the effect of the switch position of the present invention on the electromagnetic force.

Fig. 6 is a graph of predicted current and actual current waveforms of the present invention.

FIG. 7 is a three-phase electromagnetic force waveform for the open loop and prediction ambiguity case of the present invention.

Fig. 8 is a waveform diagram of the electromagnetic force under control of the algorithms of the present invention.

Fig. 9 is a waveform diagram of an electromagnetic force in the case of an added load of the present invention.

Detailed Description

The invention is further illustrated below with reference to examples.

The switched reluctance linear motor torque control experimental system comprises a DSRLM prototype, an RT-LAB digital controller, an isolator, a driving circuit, a current sensor, a position sensor, a rotating speed measuring device, a torque measuring device, a power supply system, a power converter and the like, and the structure of the switched reluctance linear motor torque control experimental system is shown in figure 1.

The power converter is mainly designed around improving the operation stability and reducing the cost, so a three-phase asymmetric half-bridge circuit is adopted as the novel power converter topology structure of the experiment. As shown in fig. 2, this structure allows each phase to operate independently, has a good follow-through channel, and moreover, can sufficiently utilize the rated voltage of the switching tube, while realizing flexible control of each phase. Therefore, the asymmetric half bridge is suitable for high voltage and high power occasions.

In the aspect of control algorithm, a prediction module and a fuzzy control module are added on the basis of an APC control method, and the structure of a control system is shown in FIG. 3. The core part of the control algorithm consists of two parts, one part is Model Predictive Control (MPC), and the MPC has the advantages of good stability, easy modeling, capability of effectively processing multivariable and constraint problems, capability of improving the stability of the system and suitability for motor control. The second is fuzzy control. Fuzzy control does not need an accurate mathematical model and is suitable for the nonlinear characteristic of the SRLM. In order to make MPC have better compatibility with fuzzy control, the method of table lookup is adopted for the control table, flux linkage value and force value of fuzzy control, so as to reduce the operation burden of system.

The structure of the fuzzy control is shown in fig. 4 (a). The two-dimensional look-up table is the control table of the fuzzy governor to obtain the clarity output roundness affecting the characteristics of the fuzzy control system. Too little selection of k3 will result in longer dynamic response process of the system, and too much selection of k3 will result in oscillation of the system. First, for the selection of k3, since the maximum value in the control table is 5.35, after multiplying by the multiplier k3, it should be lower than the maximum current allowed by SRLM, and k3 is taken as 2.5. For the selection of k1, k2, different combinations of k1, k2 are input in the fuzzy module, resulting in the combination of k1, k2 that minimizes the pulsation of the electromagnetic force. The relationship between the selection of k1 and k2 and the pulsation of the electromagnetic force is shown in fig. 4 (b). For the data obtained from the simulation, k1 was taken as 0.1 and k2 was taken as 0.1.

TABLE 1 fuzzy control table

In order to find the optimal conduction range, an SRLM driving system model based on the structure of fig. 5(a) is established in Simulink. Different switch positions are combined into the system to minimize electromagnetic force ripple, thereby determining the switch position of the system. Through simulation under Simulink, electromagnetic force pulsations at different switch positions are obtained, as shown in fig. 5 (b). When the pulsation of the electromagnetic force is minimum, the switch positions are 1mm and 27mm, respectively. Therefore, 1mm and 27mm were selected as the on and off positions of the system. In the case of the open loop, when the on position is-2 mm and the off position is 20mm, that is, when the excitation range is not large enough, the three-phase electromagnetic force is obtained as shown in fig. 5(c), and when the off range is not large enough, although the electromagnetic force of each phase does not have a negative value in the commutation region, when the electromagnetic force of one phase is decreased, the electromagnetic force of the other phase is not increased, and the total electromagnetic force is rapidly decreased, as can be seen from fig. 5 (c). In fig. 5(d), under the control method of the predictive fuzzy control, the on position is 1mm, the off position is 27mm, the electromagnetic force is more continuous, and the overlap of the electromagnetic force of the commutation region also prevents the sharp drop of the total electromagnetic force.

In order to ensure the effectiveness of the control algorithm, the accuracy of the predicted current must be ensured. Since the electromagnetic force is obtained by looking up the current table and the position table, the accuracy of the current prediction directly relates to the accuracy of the electromagnetic force prediction. Taking phase a current as an example, the reference rotation speed is set to 0.8m/s, k1 is set to 0.1, k2 is set to 0.1, and k3 is set to 2.5, and the predicted current waveform and the actual current waveform are obtained, as shown in fig. 6. In fig. 6, the black line is the actual current and the blue line is the predicted current. In FIG. 6, it can be seen that the current prediction module is able to accurately predict future current values. And inputting the predicted current data and the shifter position of the next sampling period into a lookup table to obtain the predicted electromagnetic force.

Fig. 7(a) is a waveform of the three-phase electromagnetic force in the open loop control mode. As can be seen from fig. 7(a), in the open loop state, when the conduction range of the MOSFET is expanded, the electromagnetic force of the system generates a negative value, which reduces the efficiency of the motor, increasing the ripple of the electromagnetic force. In addition, due to the influence of the longitudinal end effect, the three-phase current is unbalanced, so that the three-phase electromagnetic force is also unbalanced. Fig. 7(b) is a three-phase electromagnetic force waveform diagram of the system under the predictive fuzzy control. In the same on-off range of fig. 7(a), the three-phase electromagnetic force under the prediction fuzzy control does not have a negative value. Thereby improving the efficiency of the motor and reducing the pulsation of the electromagnetic force. Meanwhile, unbalance of the three-phase electromagnetic force is reduced to a certain degree.

And simulating the system by using Simulink to obtain the switch position combination with the minimum electromagnetic force pulsation in the open loop state. The opening position is-2 mm, and the closing position is 20 mm. As shown in fig. 8(a), the electromagnetic force of the motor causes a large starting torque at the start of starting. The commutation region of the SRLM is mainly subject to large electromagnetic force ripple. The waveform of the electromagnetic force under the PI control at the same given speed is as shown in fig. 8(b), and the suppression of the torque ripple is insignificant under the PI control compared to the open loop. The reason for this may be that the PI control method has hysteresis, the structure is relatively simple, and the current control of the PI module has a small suppression effect on electromagnetic ripple. The nonlinear and low frequency characteristics of the SRLM also affect the effect of PI regulation. To demonstrate the role of the prediction module in suppressing electromagnetic force ripple, the parameters of the fuzzy control are adjusted to the optimal state. As can be seen from fig. 8(c), the electromagnetic force ripple problem under the fuzzy control is significantly improved as compared with the PI control. The electromagnetic force waveform of the system under the prediction fuzzy control is as shown in fig. 8(d), and under the prediction fuzzy control, after the electromagnetic force is stabilized, the electromagnetic ripple is suppressed better than the open loop, PI control and fuzzy control. Since the prediction fuzzy control can suppress pulsation of the electromagnetic force of the SRLM more significantly than the simple fuzzy control, it is effective for the prediction of the electromagnetic force.

As can be seen from the former section, the predictive fuzzy control can effectively suppress the pulsation of the electromagnetic force of the motor. Compared with the PI control and the fuzzy control which act independently, the effect of restraining the electromagnetic force pulsation is more obvious. In order to test the load capacity of the predictive fuzzy control system, the parameters determined above are taken. The motor was then loaded and tested for performance. When the load is increased to 50N, other parameters are unchanged, the electromagnetic force curve of the SRLM is shown in figure 9(a), and after the load is increased, the electromagnetic force ripple can still be well suppressed by the system. Continuing to increase the load to 100N, an electromagnetic force waveform is obtained, and as shown in fig. 9(b), when the load is 100N, the suppression of electromagnetic force ripple by the system is weakened. At this time, the power supply voltage of the system may be increased to provide a greater electromagnetic force.