Unmanned vehicle brake system fault diagnosis method based on Bayesian network

Abstract

The invention relates to the technical field of unmanned driving, in particular to an unmanned vehicle brake system fault diagnosis method based on a Bayesian network, which adopts a probabilistic reasoning method of a group tree propagation algorithm, the Bayesian network is converted into a group tree by the group tree propagation reasoning algorithm, and then messages are transmitted to each node in the group tree once through message transmission. Finally, the group tree meets the global consistency, after the group tree is constructed, fault diagnosis reasoning is conducted through the group tree, and after evidence is received, the function values of all the group nodes containing the evidence nodes will change and are transmitted to all the group nodes in the group tree so as to change the function values of the nodes. According to the method, the Bayesian network is used for carrying out fault diagnosis on the unmanned vehicle brake control system to determine the fault propagation direction, the uncertainty of fault causes and results is described in a probability form, and the Bayesian network is subjected to fault diagnosis reasoning by using a group tree propagation algorithm, so that the fault diagnosis precision is improved.

Description

technical field [0001] The invention relates to the technical field of unmanned driving, in particular to a fault diagnosis method for an unmanned vehicle braking system based on a Bayesian network. Background technique [0002] The unmanned vehicle system mainly includes three parts: environment perception, behavior decision-making and action implementation. The unmanned vehicle senses the surrounding environment through the sensor, the main control computer and the behavior decision-making system analyze and calculate the path plan, and finally realize the autonomous driving by controlling the steering and speed. The braking system is one of the core subsystems of the control system. During the driving process of unmanned vehicles, if the braking system has brake failure, self-braking or braking deviation, etc., serious traffic accidents may occur. It can be seen that the safety of the braking system directly affects the safety of the driverless vehicle. Driving safety of...

AI Technical Summary

Problems solved by technology

[0003] The usual fault diagnosis method is based on the reasoning of the causal relationship between the fault symptom and the cause of the fault, but for the braking system of an unmanned vehicle with a complex structure, this causal relationship is not necessarily a one-to-one correspondence. mapping relationship, but a strong randomness and uncertainty

The main factors leading to this kind of randomness and uncertainty are the complexity of the diagnostic object, the limitation of the test method, the imprecise expression of knowledge, etc. When the control system shows a specific fault, it may be from a fault state to another fault state or the result of multiple faults alternately existing, it is difficult to determine the real cause of the fault

Images