Event driving-based bullet train robust sensor intermittent fault detection method

Abstract

The invention discloses an event driving-based bullet train robust sensor intermittent fault detection method, and belongs to the field of signal processing. The method comprises the following steps of: establishing a bullet train random uncertain state space model; designing an event trigger strategy; designing an event driving-based robust residual generator: offline designing a state estimatorgain: setting a state estimator initial value, calculating a middle variable, calculating the state estimator gain and calculating a state estimation error covariance upper bound, and online calculation residual: setting a state estimation initial value, calculating a one-step state predicted value, calculating an innovation, calculating a state estimated value and calculating the residual; and designing an intermittent fault detection strategy: calculating a fault detection statistical amount, calculating a fault detection threshold value and setting a fault detection logic. According to themethod, occurrence moments and disappear moments of intermittent faults can be detected at the same time, so that the stable and safe operation of the bullet trains is ensured and the practical application requirements are effectively ensured.

Description

technical field [0001] The invention belongs to the field of signal processing, and in particular relates to an event-driven method for detecting intermittent faults of a robust sensor of an EMU. Background technique [0002] The working environment of EMUs is harsh, and intermittent faults are common, posing a huge threat to traffic safety. For intermittent failures, technical personnel will follow the vehicle to analyze and deal with them. The treatment measures are relatively conservative. If necessary, stop the vehicle for inspection, which seriously affects the operating efficiency and even affects driving safety. Intermittent faults have different characteristics from permanent faults, and the existing permanent fault diagnosis methods are difficult to apply directly. On the other hand, the operating environment of the EMU is extremely complex and changeable, the system disturbance and model uncertainty are relatively large, and it is usually impossible to obtain prio...

AI Technical Summary

Problems solved by technology

For intermittent faults, technicians will follow the car to analyze and deal with them. The countermeasures are relatively conservative. If necessary, it needs to be stopped for inspection, which seriously affects the operating efficiency and even affects driving safety.

Intermittent faults have different characteristics from permanent faults, and the existing permanent fault diagnosis methods are difficult to apply directly

On the other hand, the operating environment of the EMU is extremely complex and changeable, the system disturbance and model uncertainty are relatively large, and it is usually impossible to obtain prior information such as the corresponding uncertainty structure and structural parameters, and the existing robust fault detection methods cannot be applied.

In addition, the event-driven mechanism can effectively reduce the number of EMU signal transmissions, thereby saving energy and prolonging the service life of equipment, but the existing EMU fault detection methods do not fully consider this mechanism

Images