Model-free adaptive robust decoupling control method for two-motor speed regulation system based on neural network inversion model

Abstract

The invention discloses a model-free adaptive robust decoupling control method for a two-motor speed regulation system based on a neural network inversion model, which comprises the steps of: analyzing a mathematical model of the two-motor speed regulation system to obtain an input output relational expression related to an inverse system of the original system; acquiring dynamic and static data samples of the two-motor speed regulation system under the excitation of different inputs, and carrying out training on big data samples to approach an inverse system model and construct a pseudo-linear composite system; adding a tracking differentiator to arrange a transition signal, designing a model-free control compensator according to a dynamic linearization method to compensate nonlinear feedback outputs so as to inhibit influence caused by uncertain disturbance in a multi-motor speed regulation system and reduce tension overshoot caused by a step signal. According to the invention, aiming at the characteristics of nonlinearity and strong coupling of the multi-motor speed regulation system, a torque disturbance error caused by nonlinearity and a variable structure is obviously inhibited, a problem of tension step overshoot is improved, and robust decoupling capability of the system is improved.

Description

technical field [0001] The present invention designs a method for implementing neural network inverse model-free compensation of a multi-motor speed regulation system, which is suitable for a multi-motor speed regulation control system with a Siemens PLC as a controller. The method can also be used for other nonlinear, strong coupling, robust Electromechanical control occasions with poor rod performance. Background technique [0002] At present, there are still at least two problems in the control of multi-motor speed control systems: 1) In engineering applications, it is difficult to obtain accurate mathematical models of most nonlinear strong coupling systems, which increases the difficulty of various nonlinear controls; 2) The occasional disturbance and continuous mechanical wear coexist in the multi-motor speed control system, resulting in many operating conditions of the system and more complex control target constraints, which greatly reduces the robustness and adaptab...

AI Technical Summary

Problems solved by technology

[0002] At present, there are still at least two problems in the control of multi-motor speed control systems: 1) In engineering applications, it is difficult to obtain accurate mathematical models of most nonlinear strong coupling systems, which increases the difficulty of various nonlinear controls; 2) The occasional disturbance and continuous mechanical wear coexist in the multi-motor speed control system, resulting in many operating conditions of the system and more complex control target constraints, which greatly reduces the robustness and adaptability of the system

[0003] Due to factors such as load disturbance and motor parameter perturbation, the model mismatch is caused, which affects the control accuracy and reliability of the system control, and the robust control of the system directly is not suitable for multivariable nonlinear systems with strong coupling. Therefore, although there are many advanced control algorithms in recent years, in actual engineering applications, the ideal effect in simulation is often not achieved

Images