基于动态目标位置的智能车辆控制研究

Research on Intelligent Vehicle Control Based on Dynamic Target Position

为了真实地模拟驾驶员在动态环境中变换车道、避让静态和动态障碍物的行为方式,提出了动态目标位置概念,并采用三次样条曲线作为车道变换的路径拟合曲线.本文以模糊逻辑为控制策略,以T-S模糊模型为控制结构,以自适应神经网络为隶属度函数的参数调整手段,设计出一种智能车辆横向运动控制器,并通过计算机仿真实现.结果表明,基于动态目标位置概念的控制器设计具有较好的控制性能,能够较为理想地模拟实际交通环境中车辆横向运动的特性.

In order to simulate the behavior pattern of drivers during lane changing and static/moving obstacle avoiding in a dynamic environment, this paper proposes a dynamic target position concept, and employs cubic spline curves as the path fitting curves for lane-changing. Additionally, an intelligent lateral motion controller for vehicles is designed by using the fuzzy logic as its control strategy, the T-S fuzzy model as its structure and an adaptive neural network as the parameter-adjusting means for membership function, and the controller is implemented with computer simulation. The simulation results show that the designed controller based on the dynamic target position concept gives satisfactory, control performance, and can simulate ideally the lateral motion characteristics of vehicles in real traffic situation.