Robust fault tolerant nonfragile H∞ attitude control for spacecraft via stochastically intermediate observer

The observer-based fault tolerant attitude control problem for spacecraft is addressed, wherein the constraints of H∞ performance, quadratic stability, model parameter uncertainty, measurement errors, external disturbances, controller perturbation, and actuator fault and saturation are considered simultaneously. A novel fault estimation approach is proposed, where a stochastically intermediate variable is introduced. Based on the exactly available expectation of such a variable, a new kind of intermediate observer is proposed to estimate the attitude information and fault signals simultaneously. Compared with the traditional fault estimation methods, the approach proposed in this paper is less conservative for it requires neither equation constraints nor observer matching condition and can tolerate stochastic failure. To achieve fault tolerant and nonfragile control performance, the estimation of fault signals and explicit controller perturbation are also exploited in the controller design. Based on Lyapunov stability theory, sufficient conditions for the existence of the observer-based fault tolerant nonfragile H∞ controller are given based on linear matrix inequalities (LMIs) in terms of additive perturbation and multiplicative perturbation. Numerical simulations are performed to demonstrate the effectiveness of the proposed method and also illustrate the advantage over the existing state feedback control method in terms of great reduction in energy consumption.