Evaluation method of tank floor acoustic emission testing based on BP neural network training

In this paper, the acoustic emission characteristic parameters which can represent the safety state of the tank floor and other factors are used as input parameters. Through a large number of test data, the neural network expert system suitable for the acoustic emission evaluation of the tank floor is established. Finally, the accuracy of the evaluation system is verified by the actual field results of multiple tanks, and the reliability of the evaluation of the acoustic emission testing of the tank floor is improved.