Model-free Adaptive Robust Decoupling Control For Multi-motor Speed Control System

As a nonlinear and strongly coupled multi-motor speed control system with many operating conditions,how to improve the control performance of such complex nonlinear system will be the focus of current research.Based on the multi-motor variable frequency speed control synchronous control experimental system,this paper proposes a model-free adaptive(MFA)control based on neural network inverse(NNI)for the strong coupling problem of multi-motor speed control system under load torque disturbance to improve system adaptive decoupling.Aiming at the problem that the neural network offline training is easy to fall into the minimum value and the existing tension step overshoot problem,the solution is given,and the proposed control strategy is simulated and analyzed.The specific contents are as follows1.Considering the interference of load torque,the physical model of the multi-motor speed synchronous system is analyzed,and its mathematical model is derived.The system reversibility is proved.The relative order of the system and the mapping relationship between the input and output of the inverse system is obtained.Then,construct the neural network inverse system to obtain the pseudo-linear subsystem2.In order to improve the neural network inverse adaptive ability,dynamic linearization is performed on the basis of pseudo-linear subsystem,and a model-free compensation control strategy is proposed.The model-free adaptive control based on neural network inverse(MFA-NNI)structure is constructed to solve the problem of insufficient neural network inverse adaptive ability under load torque interference3.The particle swarm optimization is used to optimize the weight threshold of the neural network training phase to avoid the problem that the offline training falls into a minimum value.At the same time,based on the principle of tracking differentiator(TD),the transitional links of speed and tension are arranged to suppress and reduce the step overshoot of tension,and the control quality of the multi-motor variable frequency speed regulation tension is improved.4.Set up Matlab/Simulink simulation platform to construct neural network inverse approximation model.The proportional integral derivative control(PID)control,NNI control and MFA-NNI control system are designed and constructed.And the above control strategy is simulated and compared5.Based on the experimental platform of Siemens S7-300PLC,a multi-motor variable frequency speed control system was built and experimental analysis was carried out.The simulation and experimental results show that the improved MFA-NNI control does not need accurate mathematical model,and has excellent robustness against parameter perturbation and external disturbance.It realizes the decoupling between the speed and tension of the multi-motor speed control system.It not only realizes the decoupling between the speed and tension of the multi-motor speed control system,but also improves the dynamic performance of the system,and effectively suppresses the tension overshoot,which has important engineering application value.