Optimal control law for fault tolerant control systems

In this paper, an optimal control law is designed for fault tolerant control systems with Markovian parameters (FTCSMP). The matrix maximum principle is used to minimize an equivalent deterministic cost function. Three scenarios are considered. The first assumes that both the failure process and the fault detection and isolation (FDI) process are accessible for the controller. In the second scenario, the controller is reconfigured based on the decisions of the FDI process and does not need an accessible failure process. The case where the FDI process itself is not able to give any decision due to physical malfunction or excessive computational time is the third scenario. In these scenarios, optimal control laws are developed to reduce the risk of losing system stability. A computational algorithm is constructed to calculate the optimal control law.

[1]  W. Wonham.  Random differential equations in control theory , 1970 .

[2]  Frank Kozin,et al.  A survey of stability of stochastic systems , 1969, Autom..

[3]  R. Srichander,et al.  Stochastic stability analysis for continuous-time fault tolerant control systems , 1993 .

[4]  Michael Athans,et al.  The Matrix Minimum Principle , 1967, Inf. Control..

[5]  Youmin Zhang,et al.  Stochastic stability of fault tolerant control systems with model uncertainties , 2000, Proceedings of the 2000 American Control Conference. ACC (IEEE Cat. No.00CH36334).

[6]  Youmin Zhang,et al.  Analysis of the stochastic stability for fault tolerant control systems , 1999, Proceedings of the 38th IEEE Conference on Decision and Control (Cat. No.99CH36304).

[7]  M. Mahmoud,et al.  Stochastic stability of fault tolerant control systems in the presence of noise , 2000, Proceedings of the 2000 American Control Conference. ACC (IEEE Cat. No.00CH36334).

[8]  M. Mariton.  Detection delays, false alarm rates and the reconfiguration of control systems , 1989 .