Research on Chinese drivers' shifting quality evaluation based on simulative traffic environments

摘要

In China, most of drivers drive manual transmission vehicles, and the high number of road accidents was caused due to unreasonable driving behaviors, such as wrong, or prentice, unpracticed acceleration shifting operations. Based on the existing domestic and international research results about driving behaviors, general theory and methods of driving behavior characteristics were analyzed, correlation methods were highlighted. Driving behavior tests under simulative traffic environment scenes were carried out on the fixed driving simulator and driving behavior simulation experiment platform developed by the State Key Laboratory of Automobile Dynamic Simulation of Jilin University. Actual tests combined with impact factors were analyzed. Adaptive Neuron-Fuzzy Inference System (ANFIS model) was set up by Matlab Fuzzy Inference Systems Toolbox, appropriate membership functions and mixed-learning algorithm was selected. Finally, the driver ANFIS model applied to acceleration shifting operation was obtained.