Multi-fault diagnosis and optimization method of control system

Abstract

The invention discloses a multi-fault diagnosis and optimization method of a control system. The method comprises a step of constructing unknown input Kalman filter groups corresponding to actuators and sensors according to a discrete state space model, a step of obtaining output estimation error residuals corresponding to the actuators and the sensors according to an optimal state estimator X<^><C>(i,k) of a Kalman filter corresponding to an ith actuator in a kth step and an optimal state estimator X<^><O>(i,k) of a Kalman filter corresponding to an ith sensor in a kth step, a step of carrying out weighted quadratic sum processing on the output estimation error residuals and obtaining corresponding weighted squares residuals, a step of obtaining a preliminary diagnosis result according toa comparison result of the weighted squares residuals and a mechanism detection threshold, and a step of outputting an isolation strategy and a final diagnosis result according to the preliminary diagnosis result. According to the method, the Kalman filter groups with limited structures are used to carry out control system multi-fault isolation and optimization, a global optimal solution of stateperception can be obtained, and the accuracy of a fault diagnosis result of the control system is improved.

Description

technical field [0001] The invention belongs to the technical field of spacecraft control, in particular to a method for multiple fault diagnosis and optimization of a control system. Background technique [0002] In the field of spacecraft control, the research of fault diagnosis technology is of great significance for spacecraft control systems with high safety requirements, such as satellites, airplanes, trains, automobiles, power plants and chemical plants. Fault diagnosis usually includes fault detection and fault isolation. [0003] Usually, the location and severity of faults are random to a certain extent, and fault isolation is a prerequisite for fault recovery. As the number of components in control systems continues to increase in practical applications, fault isolation has become an urgent problem to be solved. [0004] However, there are many problems in the current fault diagnosis methods that cannot be solved: (1) The existing fault diagnosis methods regard ...

AI Technical Summary

Problems solved by technology

[0004] However, there are many problems in the current fault diagnosis methods that cannot be solved: (1) the existing fault diagnosis methods regard the fault as the system state to expand the system, and require the additive fault to change slowly, which has limitations; (2) cannot Realize multiple fault isolation of sensors and actuators at the same time; (3) cannot handle multiple faults of the system

Images