智能汽车自动紧急转向避撞的跟踪控制方法对比研究

为了提高智能汽车的主动安全性，提出3种不同的自动紧急转向避撞跟踪控制方法。首先建立汽车避撞简化模型，对制动、转向及两者相结合的3种不同避撞方式进行对比分析。其次，为深入研究汽车避撞过程中的实际响应，建立包含转向、制动及悬架3个子系统耦合特性的底盘18自由度统一动力学模型，并进行相关试验验证。随后构建智能汽车自动紧急转向避撞控制框架，对五次多项式参考路径和七次多项式参考路径的横摆角速度和横摆角加速度进行对比分析。接着以线性2自由度转向动力学模型为参考对象，对最优控制四轮转向、最优控制前轮转向、前馈与反馈控制相结合的前轮转向3种不同的跟踪控制系统分别进行设计。最后，以汽车底盘18自由度统一动力学模型为研究对象，对上述3种避撞控制系统进行仿真试验对比分析。研究结果表明：与制动避撞相比而言，转向避撞所需的纵向距离有较大降低，随着车速的增加和路面附着系数的越低，效果越明显；七次多项式参考路径比五次多项式参考路径的避撞过渡过程更为平缓，当实际车速与控制器所用车速不一致时，前者避撞性能表现更优；最优四轮转向控制系统在高、低2种不同附着路面都具有较好的避撞效果，最优前轮转向控制系统次之，而前馈与反馈相结合的前轮转向控制系统在低附着路面上则表现出严重的失稳。

Comparative Study on Tracking Control Methods for Automatic Emergency Steering and Collision Avoidance of Intelligent Vehicles

To improve the active safety of intelligent vehicles, three different tracking control methods for automatic emergency steering collision avoidance are proposed. First, a simplified model of vehicle collision avoidance was developed, by which three different collision avoidance modes of braking, steering, and their combination were compared and analyzed. Second, an 18-degree-of-freedom-(DOF) unified dynamic model of vehicle chassis, including the coupling characteristics of the steering, braking, and suspension subsystems, was developed to study the actual response of the vehicle during collision avoidance. The correctness of the above model was verified by relevant experiments. A control framework for an automatic emergency steering collision avoidance system for intelligent vehicles was constructed. The yaw rates and yaw angular accelerations of the fifth and seventh polynomial reference paths were compared and analyzed. Using the linear two-DOF steering dynamic model as the reference object, three different tracking control systems were designed, including the optimal control of four-wheel steering, optimal control of front wheel steering, and combination of feedforward and feedback of front wheel steering. Using the 18-DOF unified dynamic model of vehicle chassis as the research object, three different types of collision avoidance control system were compared and analyzed by simulation tests. The longitudinal distance required for steering collision avoidance is largely reduced compared to that of the braking collision avoidance. The effect increases with the increase in vehicle speed and decrease in road adhesion coefficient. The lane-changing process of collision avoidance of the seventh-order polynomial reference path is smoother than that of the fifth polynomial reference path. When the actual speed is inconsistent with the speed used by the controller, the performance of the former is higher than that of the latter. The four-wheel steering control system has good collision avoidance performances on high-and low-adhesion roads, followed by the optimal front wheel steering control system, while the feedforward and feedback control system of front wheel steering exhibits a serious instability on a low-adhesion road.