The Design Of Mixed Sensitivity H_∞ Control System Based On Multi-objective Immune Algorithm

H_∞Robust control is a control theory that optimizes certain performance indicators' infinite norm so that we can obtain a robust controller. It has developed rapidly in recent years and has been very mature in theory and in the realization. H_∞mixed sensitivity design as a robust design method, it can make up for the modern control theory to rely too much on the mathematical model. The model's uncertainty was considered in the H_∞mixed sensitivity design, the transient performance and anti-jamming capability of the system was also taking into account. In addition, mixed sensitivity H_∞control is designed by choosing weighting function matrix in the frequency domain. It is similar to the classical design method in frequency domain which is familiar with engineers, thus increasing designers' attention. However, a successful design depends on the appropriate choice of weighting functions, which in turn relies on a designer's experience and familiarity with the design approach. The design procedure of the control system can be expressed as a multi-objective optimization problem when the weighting functions are considered as design parameters and the optimal solution must satisfy a set of inequality constraints. This paper adopts an artificial immune algorithm and takes advantage of the mechanism of producing and preserving the diversity of antibodies in the immune principle to solve this problem. Then a H_∞controller which meets the time-and frequency-domain indexes was searched for. A controlled object of the rocket system attitude control as an example, design its robust controller, so that it is not only robust, but also satisfy all the given performance requirements.Simulation result shows that, weighting function and robust controller can be obtained through the optimization algorithm. Normally used trial and error method can be avoided and the satisfactory weighting matrixes can be effectively searched out by means of the proposed approach. The robust controller which using the weighting matrixes for design has some advantages, such as: low sensitivity, high tracking speed, excellent stability margin.