Ventilation Equipment and Facility Reliability Analysis Based on Evidence Theory and Neural Network

This article proposed a new solution through the combined use of evidence theory and neural network method for the evaluation of the reliability of the mine ventilation equipment and facilities. First, the outputs of multiple BP neural network serve as the basic probability distribution function of the evidence theory, after which the results are fused according to the rules of evidence combination, from which the security degree of the ventilation system can be figured out. The example shows that the evaluation results by this method are reasonable and objective, and very effective for the reliability evaluation of mine ventilation equipment and facilities. In light of the output instability of single BP neural network, we can select multiple BP networks, and through the comparison of the uncertainty degree of different lines of BP network after their fusion, we are able to analyze the impact of the selected lines of BP network on network instability.