A Fault Diagnosis Method of Bridge Crane Based on Bayesian Network

Abstract

The invention discloses a bridge crane fault diagnosis method based on Bayesian network. Bayesian networks applied to crane fault diagnosis have not yet been discovered. The present invention first classifies the fault events that have been sent, obtains a fault tree model, and establishes a Bayesian network structure of fault events; secondly, classifies the nodes in the Bayesian network structure to obtain a module state layer, a fault reasoning layer, and a fault bottom event Layer; collect the output signal of the module state layer again, and select the fault diagnosis mode according to whether each module of the bridge crane is operating normally; then convert the actual state performance parameters of the fault bottom event into the current fault probability according to the acquired fault bottom event layer, and reset the fault bottom event layer The prior probability makes the diagnosis model more in line with the actual situation; finally, the Bayesian network algorithm is used for fault diagnosis to realize fault performance prediction or first fault location. The present invention fully decouples fault events, and constantly updates the fault probability of the fault bottom event during the troubleshooting process until the fault is located.

Description

technical field [0001] The invention belongs to the field of fault diagnosis, and in particular relates to a fault diagnosis method for bridge cranes based on a Bayesian network. Background technique [0002] Large bridge cranes are one of the special equipment under national key supervision. With the development and reform of my country's modernization, the scale of industrial production continues to expand, and higher requirements are put forward for the normal operation of bridge cranes in production and application. According to production statistics, the production interruption time due to crane failure accounts for 15% of the total production stop time. Therefore, it is necessary to establish an effective bridge crane fault diagnosis system to improve the timely repair rate of faults and ensure the production progress of enterprises. [0003] The core of the fault diagnosis system is how to quickly locate the first fault point of the fault and carry out preventive main...

AI Technical Summary

Problems solved by technology

After searching the literature of the prior art, it was found that the public document "A Bridge Crane Fault Diagnosis Method Based on Fault Tree and Bidirectional Associative Memory Neural Network" (Automation and Information Engineering, 2014) proposed a fault diagnosis system for bridge cranes. The fault diagnosis method based on the combination of fault tree (FTA) and bidirectional associative memory neural network (BAM) has the following disadvantages: this method greatly simplifies the fault tree model of the crane, and the bidirectional neural network model is more effective in dealing with the fault coupling level. It takes a long time to reach network resonance in the network, the diagnosis efficiency is low, and the reasoning process is not based on the actual performance status of the bridge crane

Images