Autonomous design of spacecraft attitude control based on normal matrix and genetic algorithm

The problem of autonomous robust attitude control for flexible spacecraft is considered. A novel autonomous design approach is introduced based on the normal matrix design theory of the multivariable control system, and the parameters of controller are optimized by genetic algorithm (GA). The robustness of proposed approach is independent of the parameters of the spacecraft attitude dynamics, including the inertia tensor of the spacecraft and the vibration of the flexible appendages. Under a class of feedback perturbation, named inverse additive perturbation, the robust stabilization criterion in normal matrix description is obtained based on the structure and numerical properties of spacecraft attitude dynamics, and this criterion converts the robustness requirement into a constrain to the control designing parameters. Applying this criterion into GA, the chromosome is shortened, and therefore the running of GA is effectively promoted. A design example shows the efficiency of the algorithm.