Research on the Risk Assessment Method of Power Operation Based on Fuzzy Output Neutral Network

Abstract

In order to accurately and objectively evaluate the risk level when power industry implement a power operation, this paper proposes a risk assessment method for power operation based neutral network with fuzzy output. Firstly, based on the analysis and summary of the power operation historical accident, 19 power operation risk assessment indexes are selected from personnel expedition, mechanical devices and equipment, operating time and environment, and operation conditions. Secondly, in view of the risk level of power operation is a fuzzy phenomenon, a three-layer neural network model of power operation risk assessment is established, in which the assessment indexes are set as input, and five fuzzy risk memberships are output. The model introduces the risk membership degree as the output result, which can not only determine the final risk level of power operation, but also explore the potential risks in power operation according to the risk membership of power operation. Finally, through the comparison of three risk assessment methods, it is verified that this method is more reasonable and objective for power operation assessment.