【作者】 陈亮；

【导师】 秦兆博； 贺冠强；

【作者基本信息】 湖南大学 ， 车辆工程（专业学位）， 2021， 硕士

【摘要】 自动驾驶车辆横/纵向控制的主要功能是使车辆按照期望参考轨迹安全可靠地行驶,合理的车辆动力学模型和控制方法对车辆横/纵向控制至关重要。除此之外,精确的车辆状态与路面参数也是控制系统的关键输入。本文以智能网联车辆为研究对象,针对基于状态参数估计的自动驾驶车辆横/纵向动力学控制问题进行研究。主要内容如下:(1)基于自适应无迹卡尔曼滤波(AUKF)的多状态参数联合估计方法。采用一种派生的容错噪声统计特性估计器与无迹卡尔曼滤波(UKF),在只需自动驾驶车辆标配定位传感器的条件下,实现了对车辆质心位置、路面附着系数、轮胎侧偏刚度、轮胎垂向力以及路面坡度等状态参数的解耦并实现了多参数联合估计。仿真与实验结果表明,该方法可以给予横/纵向控制器精确的动力学参数输入。(2)基于最优前轮侧偏力的横向控制方法。针对大横向加速度工况下的路径跟踪问题,提出了一种基于最优前轮侧偏力的线性二次型调节器(LQR)横向控制方法。通过构建基于“前馈+反馈”的LQR控制器对前轮侧偏力进行在线求解。结合轮胎垂向力、侧偏刚度以及路面附着系数等状态参数估计,该方法在大横向加速度工况下可以降低路径跟踪误差并能有效提升车辆的操纵性与横向稳定性。(3)基于期望驱/制动力补偿的纵向控制方法。为提高纵向控制器的精度与鲁棒性,采用双层控制结构,下层控制器采用基于速度反馈的比例积分(PI)控制方法对车辆底层的油门和制动踏板进行控制以达到在平坦路面上跟踪期望速度的目的。上层控制器采用基于模糊系统的期望驱/制动力补偿算法,将坡度的影响近似转化为车辆所需额外的期望驱/制动力,进而通过模糊系统将期望驱/制动力转化为油门踏板开度与制动踏板开度的补偿量,以实现在坡路工况下精确地纵向速度跟踪控制。(4)基于智能网联车辆平台的实验研究。分别对多状态参数联合估计方法与横/纵向动力学控制方法进行实验验证。实验结果表明,所设计的AUKF估计算法能精确地估计车辆状态与路面参数,所提出的横/纵向动力学控制策略在不同的典型工况下均能精确的进行轨迹跟踪,具有较高的工程应用价值。更多还原

【Abstract】 The main function of lateral and longitudinal dynamic control is to make the vehicle track the reference trajectory safely and reliably.It is vital for lateral and longitudinal control to adopt the appropriate vehicle dynamic model and control method.In addition,the accurate state and parameter of vehicle are also the key input.The paper takes the intelligent connected vehicle platform as the research object and study the problem of lateral and longitudinal dynamic control for autonomous vehicle base on state and parameter estimation.The main research has the following four points.(1)The state and parameter estimation method based on adaptive unscented Kalman filter(AUKF).A derivative fault-tolerant noise statistic estimator and AUKF are adopted to estimate the center of mass position,coefficient of road adhesion,tire cornering stiffness,tire vertical force and road slope.Simulation and experiment results show that this method can give precise dynamic parameter input to lateral and longitudinal controllers.(2)Lateral control method based on the optimal front-tire lateral force.Focusing on the large lateral acceleration condition,a lateral control method for autonomous vehicles based on the optimal front-tire lateral force is proposed in the paper.By constructing the LQR controller based on the feedforward and feedback,the desired front-tire lateral force is solved online in real time.By combining the state and paramete such as vertical tire-force,tire cornering stiffness and coefficient of road adhesion,the method can not only reduce the path tracking error,but also ensure the steerability and the lateral stability of the vehicle.(3)Longitudinal control method based on the desire driving and break force compensation.To improve the accuracy and robustness of controller,the paper apply the hierarchical control structure.The lower controller uses PI controller to control the throttle and brake pedal to track the desired speed,and finally to achieve accurate longitudinal dynamic control.The upper controller adopts the longitudinal control method with desire driving force and break force compensation based on fuzzy system to realize the accurate longitudinal control in rampway.(4)The experiment research based on intelligent connected vehicle platform.The method of estimation and control is verified simultaneously.The experimental results show that the vehicle state and parameter estimation method can estimate the road adhesion coefficient,tire cornering stiffness,vertical tire-force,and road slope accurately.The presented lateral dynamic control and longitudinal dynamic control strategy can effectively tracking desired trajectory in the diffierent typical conditions,which has high engineering application value.更多还原