摘 要

自适应巡航控制系统是近年来兴起的驾驶辅助系统的一种。与传统巡航相比，自适应巡航能够通过车载传感器对前方车辆进行感知，根据控制指令实现巡航或跟车功能，一定程度上减轻驾驶员负担并减少交通事故的发生。

目前自适应巡航控制系统通常采用雷达作为传感器感知本车前方的障碍物等道路交通环境，依据车辆当前所处的交通环境及运行状态决策出理想的控制量，并通过对本车发动机及制动系统的干预实现巡航功能，同时防止与前车发生追尾。根据自适应巡航控制系统的功能，可以将其算法分为三个关键环节：感知、决策和车辆纵向控制。

决策算法是自适应巡航控制系统的核心。自适应巡航控制系统的巡航功能、车距保持功能主要取决于其决策算法。现有自适应巡航控制系统为了避免与前车发生追尾碰撞，在系统设计中更加关注安全性，然而对舒适性考虑不足。自适应巡航控制系统的舒适性表现会直接影响汽车乘员对该系统的满意度评价。通过对目前国内外自适应巡航控制系统研究现状调查发现，分工况控制和分层控制的应用越来越广泛，以应对日益复杂的交通环境，本文基于经典控制方法采用分工况的模式针对自适应巡航控制系统实际使用过程中涉及的典型工况设计了一套决策算法，并针对控制系统的舒适性单独进行了设计，从而使设计自适应巡航控制系统有更好的使用体验。

车辆纵向控制算法是自适应巡航控制系统与车辆的接口，其性能直接影响系统整体功能。车辆纵向控制算法依据决策算法的加速度指令，通过对车辆动力传动系统及制动系统的干预来实现对车辆加速度的跟踪控制。基于模型的方法可以减少标定匹配工作，通过精确的车辆动力学模型，将决策控制算法输出的加速度转化为节气门开度和制动主缸的压力，从而实现对车辆的纵向控制。

在测试过程中，有时不可避免地会有一些干扰信号的混入，在进行试验数据处理之前，需将干扰信号从测试信号中分离出来；在进行动态测试的数据处理时，常需对频率进行筛选。 在工程上将信号的分离与筛选称为滤波。为了滤除传感器所叠加的噪声，在子系统中添加了低通滤波器和卡尔曼滤波器实现信号的准确和对系统状态的准确估计。

为了验证决策算法的有效性和车辆动力模型的精确性，利用MATLAB/Simulink和CarSim仿真软件搭建了决策控制算法的模型，进行了联合仿真，通过实验结果的分析证明了决策控制算法的有效性。

关键词：自适应巡航控制系统，决策控制算法，车辆动力学模型，卡尔曼滤波，MATLAB/Simulink、CarSim仿真

ABSTRACT

The adaptive cruise control system is one of the rising driving assistance systems in recent years. Compared with the traditional cruise, the adaptive cruise through the vehicle sensor to the front of the vehicle perception, according to the control command to achieve function that cruise or follow the car, to some extent reduce the burden on the driver and reduce the occurrence of traffic accidents.

At present, the adaptive cruise control system usually uses the radar as the sensor to perceive the road traffic environment such as the obstacle in front of the vehicle, and make the ideal control amount according to the traffic environment and the running state of the vehicle, and through the vehicle and the braking system Of the intervention to achieve cruise function, while preventing the occurrence of rear-end with the car. According to the function of adaptive cruise control system, the algorithm can be divided into three key links: perception, decision making and vehicle longitudinal control.

Decision algorithm is the core of adaptive cruise control system. The cruise control of the adaptive cruise control system depends on its decision algorithm. The existing adaptive cruise control system in order to avoid rear-end collision with the car in the system design more attention to safety, but the lack of comfort. The comfort performance of the adaptive cruise control system directly affects the evaluation of the satisfaction of the vehicle occupant. Based on the current research situation of adaptive cruise control system at home and abroad, it is found that the application of sub-condition control and hierarchical control is more and more extensive to deal with the increasingly complex traffic environment. Based on the classical control method, a set of decision algorithms is designed for the typical conditions involved in the actual use of the adaptive cruise control system and designed for the comfort of the control system, so that the design of the adaptive cruise control system has a better experience.

The vehicle longitudinal control algorithm is the interface between the adaptive cruise control system and the vehicle, and its performance directly affects the overall function of the system. The vehicle longitudinal control algorithm is based on the acceleration command of the decision algorithm, and the vehicle acceleration control is realized by the intervention of the vehicle power transmission system and the braking system. The model-based approach can reduce the calibration matching work, through the precise vehicle dynamics model, the decision control algorithm output acceleration into the throttle opening and brake master cylinder pressure, in order to achieve the longitudinal control of the vehicle.

In the testing process, sometimes there will be some interference signal mixing, in the test data processing, the need to separate the interference signal from the test signal; in the dynamic test data processing, often need to filter the frequency The The signal separation and screening is called filtering. In order to filter out the noise superimposed on the sensor, a low-pass filter and a Kalman filter are added to the subsystem to achieve accurate signal and accurate estimation of the system state.

In order to verify the validity of the decision algorithm and the accuracy of the vehicle dynamic model, the model of the decision control algorithm is built by using MATLAB / Simulink and CarSim simulation software. The simulation results show that the decision control algorithm is effective.

Keywords: adaptive cruise control system, decision control algorithm, vehicle dynamics model, Kalman filter, MATLAB / Simulink, CarSim simulation

目 录

第1章 绪论 1

1.1 课题提出的背景和意义 1

1.2 自适应巡航控制系统的研究现状 3

1.2.1 自适应巡航控制系统国外研究现状 4

1.2.2 自适应巡航控制系统国内研究现状 8

1.3 本文主要研究内容 10

第2章 自适应巡航控制系统模式划分及决策算法开发 11

2.1 巡航控制模式 12

2.2 跟随控制模式 13

2.3 弯道控制模式 15

2.4 本章小结 20

第3章 自适应巡航控制系统舒适性功能设计 21

3.1 驾乘人员舒适性特征参数选定 21

3.2 舒适性功能算法设计 22

3.3 本章小结 24

第4章 自适应巡航控制系统仿真模型搭建 25

4.1 自适应巡航控制系统模型 25

4.2 车辆动力学模型 28

4.3 滤波算法 30

4.3.1 低通滤波 30

4.3.2 卡尔曼滤波 31

4.4 CarSim仿真环境设置 34

4.4.1 CarSim软件介绍 34

4.4.2 车辆模型设置 35

4.4.3 传感器参数设置 36

4.4.4 仿真场景设置 37

4.5 本章小结 38

第5章 典型工况仿真结果分析 39

5.1 巡航工况 39

5.2 跟随工况 41

5.3 弯道工况 43

5.4 本章小结 44

第6章 总结与展望 45

6.1研究工作总结 45

6.2 展望 45

致 谢 49

第1章 绪论

随着社会的发展和人民生活水平的提高，我国汽车保有量逐年增长。庞大的汽车保有量、滞后的交通设施建设和驾驶员技术水平的参差不齐等因素加剧了交通拥堵程度、交通事故的发生等社会问题。交通事故分析统计数据表明，每年有大量的交通事故是由于驾驶员疲劳驾驶因素造成的。自适应巡航控制系统等先进驾驶辅助系统能够很大程度上减轻驾驶员的劳动强度，提高汽车安全性，增加道路通行能力等，因此先进驾驶辅助系统成为各大车企、高校目前的研究方向和热点之一。