RBF-ARX (Radial Basis Function-Autoregressive exogenous) model based quick robust predictive control method

Abstract

The invention discloses an RBF-ARX (Radial Basis Function-Autoregressive exogenous) model based quick robust predictive control method. Firstly, a nonlinear state-dependent RBF-ARX model of a controlled system is established in an off-line manner by adopting a data driving technique; secondly, polyhedrons capable of covering nonlinear dynamitic characteristics of the controlled system are constructed by utilizing the established nonlinear RBF-ARX model; afterwards, by utilizing a min-max optimization principle and an invariant set based design method, in a condition that the steady state equilibrium point information of the system is unknown, an RBF-ARX model based robust predictive control method capable of realizing an optimal output tracking through the solution of a convex optimizationproblem is designed; finally, in order to solve the problem of a heavy calculation amount existing in the on-line solution of the convex optimization problem, an off-line calculation method and an on-line comprehensive technique are combined in the quick robust predictive control method; and the RBF-ARX model based quick robust predictive control method is designed.

Description

Technical field [0001] The invention belongs to the technical field of automatic control, and relates to a fast robust predictive control method based on the RBF-ARX model design, and in particular to a fast robust predictive control method for fast systems. Background technique [0002] In the actual industry, production equipment is often very complex, with strong nonlinearity, their working range is large, the working environment is complex and changeable, and the object parameters are not accurate. For example, the magnetic levitation ball system has the characteristics of non-linearity, fast response, open loop instability, and is easily affected by the power supply and working environment. Certain parameters have strong uncertainty and cannot be accurately measured. [0003] In the past few decades, predictive control technology based on linear models has been developed quite mature, but the actual industrial systems are not all suitable for being reduced to linear systems fo...

AI Technical Summary

Problems solved by technology

However, this type of control algorithm has some obvious disadvantages: 1) The actual controlled system is usually a fast, time-varying, and constrained complex nonlinear process, and it is difficult to obtain an accurate mathematical model of these complex systems; Issues such as control system convergence, robustness, and closed-loop system stability have not yet been resolved

However, this method is a method that needs to solve optimization problems with LMIs (LinearMatrix Inequalities) constraints online, and has a heavy online calculation burden. This method is difficult to apply in actual production, especially for fast control object, the small sampling period makes it difficult to complete complex online optimization calculations

Images