自适应神经模糊推理的四轮转向车辆转向控制研究

模糊规则的建立和隶属度函数的确定是设计模糊系统的难题。基于神经网络和模糊逻辑的自适应神经模糊推理系统,能够从仿真数据中自动提取出If-Then规则。并在Matlab/Simulink软件中,建立包含侧向运动、横摆运动、侧倾运动三个自由度的四轮转向车辆三自由度动力学模型。将得到的If-Then规则读取到模糊控制器中和三自由度车辆模型进行联合仿真。其中模糊控制器以方向盘转角、方向盘转角速度和车速作为输入,后轮转角作为输出。最后与前轮转向的车辆进行转向盘角阶跃仿真对比。仿真分析结果表明:基于自适应神经模糊推理系统建立的后轮转角模糊控制器能够实现理想的质心侧偏角和车辆横摆角速度响应,提高了车辆的操纵稳定性。

Research on Steering Control of Four-Wheel Steering Vehicle Based on Adaptive Neuro-Fuzzy System

The establishment of fuzzy rules and the determination of membership function are the difficult problems in designing fuzzy system.The adaptive neural fuzzy inference system based on neural network and fuzzy logic can automatically extract If-Then rules from simulation data.In the Matlab/Simulink software,a three-degree-of-freedom dynamic model of four-wheel-steering vehicle is established,which includes three degrees of freedom:lateral motion,yaw motion and roll motion.The obtained If-Then rules are read into fuzzy controller and 3 DOF vehicle model for joint simulation.The fuzzy controller uses Hand-Wheel Angle,Hand-Wheel Angle speed and speed as input,and rear wheel angle as output.Finally,the steering wheel angle step simulation is compared with the front wheel steering vehicle.The simulation results show that the fuzzy controller based on the adaptive neural fuzzy inference system can achieve the ideal sideslip angle and the yaw rate response of the vehicle and improve the vehicle handling stability.