A Time-frequency Signal-based Convolutional Neural Network Algorithm for Fault Diagnosis of Gasoline Engine Fuel Control System

This study aims to apply the convolutional neural network algorithm to diagnose the fault of the gasoline engine fuel control system. First, run the system in the Simulink environment and get its operational data, including Fuel, Fuel/air ratio. However, in order to improve the robustness of the proposed method, additive white Gaussian noise is added to the signal. Then the short-time Fourier transform is used to obtain the characteristics of time, frequency and amplitude, and become the data source of neural network modeling. The experimental results show that the 16-layer convolutional neural network architecture can completely diagnose the operating state of the system, including normal and four types of faults. At the same time, it has the advantage of computational efficiency. In the future research work, compound faults, variable speed conditions and algorithm optimization are the key points. It is expected that the operation and maintenance of the plant can be more intelligent, so as to reduce the probability of machine abnormalities and accidents, so that life and property can be better protected.