The application of neural network for Sneak Circuit Analysis on the aircraft electrical system

Considering the difficulties such as the complicated process, the obtaining of clues and the complex calculation, a new method of Sneak Circuit Analysis is presented. The artificial neural network model is used to analyze the electrical system topographies and forecast the output of the system under the unknown input, and identify the sneak condition. The method is mainly based on Back-Propagation Neural Network and performed on the MATLAB Software. The neural network model is built based on the principles and information of the circuit topographies. Function according to circuit design and historical performance data generated the neural network training samples which are used to train the neural network and predict system output, and so identify the sneak condition. In the effort to improve the accuracy of forecasting, the Genetic Algorithm and other ways are used to optimize the BP network and better results are achieved. In the analysis of actual cases, the forecasting accuracy of the neural network model can reach 60 percent, high enough to achieve the aim of forecasting the sneak condition.