Electromechanical servo system with anisotropic regulator

Abstract

A method of multiobjective synthesis for nonlinear multi-mass electromechanical servo systems with uncertain plant parameters based on feed-forward robust stochastic anisotropic control to improve the accuracy of such systems is developed. The method is based on the choice of the robust control target vector by solving the corresponding problem of multiobjective nonlinear programming in which the components of the target function vectors are direct quality indicators that are specified to the system in various modes of its operation. The calculation of the target function vector componentrs and the constraints is algorithmic and is related to the synthesis of anisotropic robust regulators and to the modelling of a synthesized nonlinear system for different operating modes of the system, with different input signals and for various values of the plant parameters. The components of the unknown vector are the required weight matrices which form the target vector of robust control. The synthesis of anisotropic regulators is reduced to the solution of a system of four related Riccati equations. The solution to the problem of multiobjective nonlinear programming is based on particle swarm optimization algorithms. The results of theoretical and experimental research into the effectiveness of a two-mass nonlinear robust electromechanical servo system with synthesized anisotropic robust regulators are presented. The comparison of the dynamic characteristics of the synthesized electromechanical servo system showed that the application of synthesized anisotropic robust regulators improves the parameters of accuracy and reduces the sensitivity of the system to changes in the plant parameters compared to the existing system.