Modeling and Analysis of Shift Schedule for Automatic Transmission Vehicle Based on Fuzzy Neural Network

摘要

The established methods of shift schedule for automatic transmission vehicle coupled with its advantage and disadvantage had been analyzed, the best shift schedule based on fuzzy neural network has been advanced, structure and algorithm on the basis of the fuzzy neural network by Takagi-Sugeno (T-S) modeling was studied. The shift schedule model for automatic transmission vehicle based on fuzzy neural network with T-S modeling was established. Two parameters for fuzzy logic rules and membership functions of the shift schedule for automatic transmission vehicle according to the skilled driver's experience and expert's knowledge had been made, membership functions and fuzzy logic rules was modified through train mechanism of artificial neural network based on experiment sample, the fuzzy-BangBang control model of engine was established by utilizing difference between the real and target speed and ratio of changing speed. Shift schedule of automatic transmission vehicle was simulated based on fuzzy neural network and the results indicated that this shift schedule based on fuzzy neural network of T-S modeling has validity and feasibility.