A Research On Self-adaptive Shift Schedule Of Automatic Transmission Vehicle Based On Driver’s Characteristic

Driver’s characteristic is the driving habit that is reflected in the process of driving and it can reflect the driver’s disposition.Drivers with different characteristics generally have different requirements on vehicle performance.Thus,the self-adaptive shift schedule of automatic transmission vehicles based on driver’s characteristics was investigated.Firstly,the driving data of some drivers in starting,accelerating and braking test conditions were collected in driving simulator.Then the driver’s clustering characteristic parameters in such conditions were selected by principal component analysis.The collected driving data were clustered respectively by K-means,Self-Organizing Map neural network and Gaussian Mixture Model clustering algorithm.After that,the cluster results were compared and analyzed.The cluster results provide classification thresholds of diver’s characteristic in starting,accelerating and braking conditions which are the base of diver’s characteristic identification.By comparing and analyzing common driving intention identification models,fuzzy logic inference model was selected to identify driver’s intention.According to hierarchical identification rules,the driver’s short-term driving intentions in vehicle longitudinal control were classified.The parameters,membership functions and fuzzy inference rules of accelerating and velocity keeping intention identification model and braking intention identification model were designed respectively.The Simulink model for driver’s short-term driving intention identification was built.Driving intention identification model was added to the driving simulator,and the effectiveness of the proposed short-term driving intention recognition model was verified on the driving simulator.According to the acceleration of the vehicle,the starting accelerating and braking condition recognition program was designed.Based on the results of cluster analysis of driver’s characteristics,a driver’s characteristic identification model that based on SOM neural network was established,which can judge the driver’s characteristics according to the history driving data.The optimal dynamic and optimal economic shift schedules of vehicle transmission were designed by graphic method.A shift schedule that can be self-adaptive to driver’s characteristic was established,which is based on optimal dynamic shift schedule,optimal economic shift schedule and driver’s characteristic index obtained by driver’s characteristic identification model.A self-adaptive shift schedule that can consider the driver’s short-term intention was established,which is based on optimal dynamic shift schedule,optimal economic shift schedule and driver’s comprehensive evaluation parameter that is determined by driver’s characteristic index and driver’s short-term driving intention.The effectiveness of the system was also verified.