Research and design of self-tuning fuzzy controller for motor soft start

1 Introduction

Three-phase asynchronous motors are widely used, but the current is very large (6-8 times) when starting at full voltage directly. Traditional methods use step-down starting methods such as Y-△ conversion, autotransformer and stator circuit series reactor to reduce the starting current. The starting parameters of the starting equipment are generally not adjustable, making its load adaptability poor. The motor soft starting method has the advantages of no impact current, adjustable starting parameters, soft stop function, light load energy saving, etc. and is gradually being widely used. The schematic diagram of the impact of each starting method on the power grid is shown in Figure 1.

At present, the soft start method mainly adopts the method of thyristor AC voltage regulation. During the motor starting process, by controlling the size of the thyristor trigger angle, the stator terminal voltage and starting current of the motor can be changed according to the rules set by the working requirements. The starting mode and starting current of the motor can be adjusted and set arbitrarily to put it in the best starting process. The commonly used thyristor voltage regulation control circuit is shown in Figure 2.

This paper introduces the application of parameter self-tuning fuzzy control technology in current limiting soft starting. Fuzzy reasoning and fuzzy decision-making are used to control the current during the motor starting process, thus achieving a smooth starting of the system.

2 Fuzzy control scheme

The traditional method of motor soft starting is to use closed-loop PID control to limit the current of the motor. However, since the asynchronous motor starting process is a nonlinear time-varying system, the use of PID closed-loop control cannot solve the current impact problem during the asynchronous motor starting process. Therefore, this paper uses parameter self-tuning fuzzy control technology with strong rapid adjustment ability to control motor soft starting. As a method of intelligent control, the biggest advantage of fuzzy control is that it does not rely on the precise mathematical model of the controlled object, can overcome the influence of nonlinear factors of the system, and has strong robustness to changes in the parameters of the controlled object.

3 Structure of self-tuning fuzzy control system

Conventional fuzzy controllers have the advantages of short response time, small overshoot, good robustness, suitability for nonlinear time-varying complex systems, and relatively easy model establishment.

Fuzzy controllers have good dynamic quality, but there are also some problems: First, the control rules of conventional fuzzy controllers are fixed after they are established, making it difficult to obtain the optimal control index. Compared with complex controlled objects such as the motor soft start process, this controller cannot achieve the expected control effect, and has poor ability to adapt to system and environmental changes.

For this reason, we use parameter self-tuning fuzzy control technology. During operation, according to the actual deviation and the deviation change rate, the controller selects different Ka, Kb, and Kc to meet the different requirements of dynamic and static performance. Moreover, based on the quantization and proportional factor self-adjustment method, the algorithm is simple and efficient, and the control effect is good. It is very suitable for systems such as motor soft start that have high requirements for real-time control.
As shown in Figure 3, considering the system control situation, the current deviation e and the deviation change rate ec are selected as input variables, and the change value a of the thyristor trigger angle is selected as the output. When the quantization factors Ke, Kec and the proportional factor are taken as constants, the adjustable factors K1, K2, and K3 can be optimized by continuously updating the values ​​of K1, K2, and K3 to complete the self-adjustment of the fuzzy control rules, so that the fuzzy control system has the best dynamic performance.

4 Selection of fuzzy control parameters

4.1 Fuzzy membership function

The fuzzy controller adopts two-dimensional fuzzy control. The main fuzzy control unit and the fuzzy parameter optimization control unit both use the deviation e of the output current of the asynchronous motor from the expected value and the deviation change rate ec as input variables, where the output variable of the main fuzzy control unit is the change value of the thyristor trigger angle a. The output adjustable factors K1, K2, and K3 of the fuzzy parameter optimization control unit represent the adjustment coefficient values ​​of the quantization factors Ke, Kec and the proportional factor Ka respectively.

The domain of e, ec and the output a of the analog controller is {-4, -3, -2, -1, 0, +1, +2, +3, +4), and its size is quantized into 9 levels. In the fuzzy control area, the current deviation is divided into 7 fuzzy subsets, namely the linguistic variables of Ke {negative large, negative medium, negative small, zero, positive small, positive medium, positive large), abbreviated as {NB, NM, NS, ZO, PS, PM, PB).

For the sake of simplicity of calculation, a simple triangular membership function is used for the input and output variables. The membership function assignment table is shown in the following table:

4.2 Fuzzy Control Rules

Taking into account the two signals of current deviation Ke and the rate of change of current deviation Kec2, the fuzzy inference rules adopted are in the form of:

Ke and Kec both have a fuzzy subset, so there are 49 fuzzy rules in total. The inference rule is expressed as

The corresponding fuzzy control rules are shown in Table 4

The centroid method used for defuzzification is also called the weighted average method. The defuzzification formula is as follows:

4.3 Fuzzy controller parameter self-tuning

The auto-tuning rules followed during startup are as follows:

When e and ec are large, reduce K1 and K2, reduce the resolution of large deviations, reduce deviations, and shorten the transition process time. When e and ec are small, the system is close to steady state. At this time, K1 and K2 should be increased to improve the system's resolution of small deviations and improve the sensitivity of control; when the error e is large and has the opposite sign to the error change ec, the size of the controller k3 should be appropriately increased. When the error e is large and has the same sign as the error change ec, the system response is accelerating away from the set value. In order to reduce this unfavorable trend, K3 should also be appropriately reduced. When the system response is near the set value (the error e is small at this time), in order to prevent large overshoot or undershoot, K3 should have a wider range of variation. Appropriately reduce the proportional factor to reduce overshoot. After many simulations and experiments, with the changes in the deviation Ke and the deviation change rate Kec, K1, K2 and K3 are taken according to Tables 5, 6, and 7 respectively, and ideal control results can be obtained.

5 System Simulation and Analysis

System simulation using MATLAB's Simulink, SimpowerSystems, FuzzyLogicToolbox and other tools.

5.1 Establishment of simulation model

The simulation model of the self-tuning fuzzy controller established by combining the above methods is shown in Figure 4:

The overall soft start simulation model is obtained by combining the three-phase power supply module, synchronization link module, pulse generation link module, three-phase AC voltage regulation link module and control link as shown in Figure 5.

Comparing the response curve obtained by using the traditional PID controller and the response curve obtained by using the self-tuning fuzzy controller, it can be seen that the control effect obtained by using the self-tuning fuzzy controller is more ideal, which is specifically manifested as: small overshoot, smooth startup process, and its output almost coincides with the given amount. Through the above comparison, it can be seen that the control performance of the self-tuning fuzzy controller is better than that of the traditional PID control. At the same time, it also shows that the self-tuning fuzzy controller designed in this paper has achieved better control effects.

6 Summary

This chapter combines the characteristics of motor soft start and introduces the entire process of designing a current-limiting soft-start self-tuning fuzzy controller on the basis of absorbing the respective advantages of adaptive and fuzzy control algorithms; and dynamically models the entire soft-start control system. Finally, the simulation model function is established using MATLAB's power system module library and Simulink module library to perform system simulation. The simulation results show that self-tuning fuzzy can achieve better dynamic and static performance in soft-start control.