The research of applying Neural Network to the Vehicles Equipment performance and cost trade-off analysis

The reliability indicator of vehicles equipment has a non-linear relationship with maintenance costs, and neural network can be used in making an correlation analysis about them. Using MTBF of the control system, transmission system and run system as inputs, maintain cost as output, data of the eight vehicles as samples, we set up the neural networks, then make training and adjustment. Using test samples to validate after the training, the result shows that the model is reasonable. Then make performance and cost trade-off analysis at the equipment demonstration and development phase, through the correlation model between performance indicators and the cost.