Vehicle Handling Inverse Dynamics Modeling And Simulation While Encountering An Emergency Collision Avoidance

Based on analyzing domestic and foreign research status about vehicle handling inversedynamics and vehicle handling stability evaluation, he dissertation combines theoretical analysis andreal vehicle test. On the one hand, according to quadratic single neuron self-adaptive PID controltheory, establish a three DOF vehicle steering model with the nonlinear tire cornering. Then simulateand research the model on high-speed emergency avoidance conditions. On the other hand, in the lightof fuzzy multi-level comprehensive evaluation theory, vehicle handling stability evaluation system isestablished. In general, detailed research contents and major achievements of this dissertation arelisted as follows:(1)Achieve two sample vehicle handling stability tests. Select some representative track routes,such as serpentine line, single-shift line and double-shift line. Record all test response of each statevariable on every condition, as well as the driver subjective evaluation.(2)Taking the tire laterodeviation characteristics into account, a3-DOF nonlinear model isestablished, with steering wheel angle as the input. Based on quadratic single neuron self-adaptivecontrol theory, establish an emergency avoidance vehicle model controlled by steering and verify it bymeans of real vehicle test. Then determine high-speed emergency avoidance route, serpentine line anddouble-shift line. Simulate and analyze the two vehicle models. Results show that, with speeddropping, driver manipulation burden is slightly heavier and yaw rate increases in varying degrees, aswell as lateral acceleration, but roll angle of the vehicle body decreases.(3)According to fuzzy multi-level comprehensive evaluation theory and driver scoring countedvia real vehicle test statistics, found vehicle handling stability evaluation system and evaluate the twovehicles’ handling stability objectively. Goodness of fit, between the results achieved by the systemand fuzzy multi-level comprehensive evaluation, is very good. Next, optimize vehicle parameters,applying multi-island genetic algorithm. Simulate optimized model and the before on the condition ofsingle-shift line, with different speed. Results show that, tracking accuracy of the optimized model ishigher significantly in different speed.