System error-based parameter self-setting method of MIMO full-format model-free controller

Abstract

The invention discloses a system error-based parameter self-setting method of an MIMO full-format model-free controller. According to the invention, a system error set is used as input of a BP neuralnetwork, and the BP neural network performs forward calculation and outputs a penalty factor, a step factor and other MIMO full-format no-model controller to-be-set parameters through an output layer.The calculation is performed by adopting a control algorithm of the MIMO full-format model-free controller, and then a control input vector for a controlled object is obtained through calculation. The value of a system error function is minimized as a target, and the gradient descent method is adopted. The gradient information sets of all to-be-set parameters are respectively set according to thecontrol input and the reverse propagation calculation of the system error is performed. The hidden-layer weight coefficient of the BP neural network is updated on line in real time, and then the hidden-layer weight coefficient is outputted for realizing the parameter self-setting of the controller based on the system error. According to the invention, the self-setting method of the MIMO full-format model-free controller is proposed based on the system error. The problem of the parameter online setting of the controller can be effectively solved. The good control effect of the MIMO system is achieved.

Description

technical field [0001] The invention belongs to the field of automatic control, and in particular relates to a parameter self-tuning method of a MIMO full-format model-free controller based on system errors. Background technique [0002] The control problem of MIMO (Multiple Input and Multiple Output) system has always been one of the major challenges in the field of automation control. [0003] Existing implementations of MIMO controllers include MIMO full-format model-free controllers. MIMO full-format model-free controller is a new type of data-driven control method, which does not rely on any mathematical model information of the controlled object, but only relies on the input and output data measured by the MIMO controlled object in real time for controller analysis and design, and The implementation is simple, the calculation burden is small and the robustness is strong, and the unknown nonlinear time-varying MIMO system can also be well controlled, and has a good app...

AI Technical Summary

Problems solved by technology

[0006] However, the MIMO full-format model-free controller needs to rely on empirical knowledge to pre-set the penalty factor λ and the step size factor ρ before it is actually put into use. 1 ,…,ρ Ly+Lu and other parameters, the penalty factor λ and step size factor ρ have not yet been realized in the actual commissioning process 1 ,…,ρ Ly+Lu Online self-tuning of other parameters

The lack of effective parameter setting means not only makes the debugging process of the MIMO full-format model-free controller time-consuming and laborious, but also sometimes seriously affects the control effect of the MIMO full-format model-free controller, restricting the development of the MIMO full-format model-free controller. Promote application

Images