Research on the Mixed H2/H∞ State Feedback Control Parameter Problem Based on Teaching-Learning-Based Optimization Algorithm

Due to the complex principles and design rules of robust controllers, their design has always been a rather difficult problem. To address the difficulty in selecting weight matrix parameters for the reference output in the design of a hybrid H
2
/H
∞
state feedback controller in robust control problems, a parameter optimization scheme based on teaching-learning-based optimization algorithm (TLBO) is raised in this paper. Firstly, the convex optimization solution problem equivalent to the mixed H
2
/H
∞
state feedback controller with closed-loop pole region constraint is described, which leads to the research problem of this paper, that is, the selection of weight matrix parameters. Then, the weight matrix is parameterized, and the parameters are optimized and solved based on TLBO to optimize the performance of the controller. Finally, simulation experiments are designed, and the results show that compared with the H
2
state feedback control and the H
∞
state feedback control, the hybrid H
2
/H
∞
state feedback controller designed in this paper behaves better in terms of suppressing external interference and initial disturbance. Moreover, the controller raised in this paper has good robustness to the perturbation of system parameters.