Fault Prediction of P-R-N-D Shifting Force Based on BP Neural Network

• Published in: 2023 IEEE 16th International Conference on Electronic Measurement & Instruments (ICEMI) pp. 487 - 491
• Main Authors: Wang, Shuxian, Liu, Shiyou, Su, Zuqiang, Wang, Xin, Liu, Linlin, Zhu, Hang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 09.08.2023
• Subjects: Analytical models; BP neural network; fault prediction; Force; Neural networks; offline inspection; Predictive models; shift force; Simulation; Time series analysis; Training
• DOI: 10.1109/ICEMI59194.2023.10270041

As an important reference indicator for transmissions, it is necessary to conduct offline testing of PRND shifting force. By studying the shift force of PRND(Parking, Reverse, Neutral, and Drive), a fault prediction method for PRND shift force based on BP neural network is proposed to improve the offline detection efficiency of PRND shift force. Based on the data obtained from the offline detection platform, a dataset for network training is obtained after processing; Build a PRND shift force fault prediction model, design and train a BP neural network, and after error backpropagation, the next layer gradient determines the gradient of the current layer to obtain the prediction network; Bringing the dataset into the network for simulation and comparing the simulation results, it was found that the network can achieve ideal results.