Research On Dynamic Model-based Fault Diagnosis For Fault-tolerant Control System Of SRLV

In view of the fuzzy, random and uncertain faults in sub-orbital reusable launchvehicle (SRLV), the traditional fault-tolerant control methods aren’t good enough to meetthe requirements. So in order to enhance the safety, reliability and maintainability of thiskind of vehicle, we need to design a dynamic robust fault-tolerant control method for itsnonlinear system. This paper, according to the SRLV special flight mission, makes use offull-order unknown input observers (UIOs) and eigenstructure assignment (EA) tofacilitate the diagnosis of sensor/actuator faults. And a robust fault-tolerant controlmethod is proposed to carry out the mathematical simulation on SRLV. The main workincludes the following parts:First of all, the domestic and overseas research of SRLV and the technology aboutfault diagnosis and fault tolerant control is surveyed.Secondly, define the common coordinate system and its transformation for SRLV.Also, the attitude singularity problem is solved by expand the ranges of angles. Based onthe dynamic theories, the mathematical model and its linear control model are establishedin detail, which is the foundation of the follow-up studies and simulations.Then, several kinds of common faults that SRLV often has are analyzed andmodeled. The algorithm of fault detection and diagnosis is presented by using theunknown input observers and eigenstructure assignment method.Finally, based on the theoretical analysis and the characteristics of SRLV, a series ofrobust fault-tolerant controllers are designed. The simulation platform of SRLV is built asto the above work. By setting different conditions of simulation environments and faultmodes, a comparative analysis of UIOs and EA is carried out and the results alsodemonstrate the satisfied performance and robustness of the designed fault-tolerantcontrol system.