Bayesian network-based fault detection method

Abstract

The invention belongs to the technical field of battery management systems and relates to a Bayesian network-based fault detection method. The invention mainly aims to solve the problem of the low fault detection efficiency of an existing battery management system. The Bayesian network-based fault detection method includes the following steps that: a Bayesian network topological structure having conventional fault representation and fault-cause corresponding relations is established; the Bayesian network topological structure is trained, and a conditional probability is obtained; Bayesian network topological structure parameter learning is performed, and the prior probabilities of fault representations are obtained; after a certain fault representation occurs on a battery management system, the Bayesian network topological structure calculates the posterior probabilities of corresponding fault causes according to the prior probability of the corresponding fault representation and the conditional probability; and a fault cause corresponding to the maximum posterior probability is adopted as a fault detection result. With the Bayesian network-based fault detection method of the invention adopted, frequent testing by technicians is not required; the process of fault detection is simplified; and the efficiency of the fault detection of the battery management system is improved. TheBayesian network-based fault detection method is suitable for the battery management system.

Description

technical field [0001] The invention relates to the technical field of battery management systems, in particular to a fault detection method. Background technique [0002] With the gradual development of battery technology, more and more new energy devices using batteries as energy storage devices are applied to all aspects of people's lives. Due to the problems of series and parallel connection, safety, and difficulty in estimating the battery power of secondary batteries, it is necessary to use a battery management system to intelligently manage and maintain battery cells, improve battery utilization, monitor battery status, and prevent batteries from overcharging and overcharging. Discharge and other problems, prolong the service life of the battery. The current battery management system has some relatively mature methods for monitoring battery power and battery charge and discharge control, but the structure of the battery management system is complex, and it is prone t...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to solve the problem that the fault detection efficiency of the existing battery management system is low, and propose a fault detection method based on Bayesian network

Images