基于Kalman滤波器的车道保持辅助系统研究

以非线性车辆动力学模型作为系统被控对象,利用Matlab/Simulink软件设计一种基于Kalman滤波算法的车道保持驾驶辅助系统。运用Kalman算法估计车辆行驶状态信息,并利用"预瞄—跟随"驾驶员模型—车辆模型—控制器所组成的驾驶员模型在回路仿真的方式对所设计系统进行验证。结果显示所设计的车道保持辅助系统能有效提高车辆路径跟踪能力。     

无人驾驶车路径跟踪控制研究

研究被控对象无人驾驶车,基于预瞄控制思想,设计一种无人驾驶车路径跟踪控制器,将控制器分为预瞄控制和补偿控制两部分,预瞄控制模拟驾驶员在驾驶车辆过程中对前方的道路环境信息进行预瞄,根据道路曲率程度决定方向盘转向,补偿控制是对车辆遇到干扰偏离原车道的纠正。仿真实验结果表明,该控制器能够保证无人驾驶车准确跟踪各种参考路径,且具有较好的鲁棒性。     

智能四轮转向车辆稳定车道线保持串级控制

为实现四轮前后轮转向车辆的稳定车道线保持控制,提出集成直接横摆力矩和车道线保持的串级控制策略.主控制器实现车道线保持前轮转角控制.副控制器实现车辆稳定性控制.主控制器基于MPC(model predictive control)算法控制车辆前轮转角,通过调整前轮转角使得横向位置偏差和航向角偏差最小.主控制器的车辆前轮转角作为副控制器的输入,计算期望滑移角和期望横摆率.车辆后轮转角和横摆力矩作为副控制器控制输入,基于LQ(linear quadratic)算法计算补偿车辆后轮转角和横摆力矩,实际车辆滑移角和实际横摆率跟踪期望滑移角和期望横摆率.车辆的前轮转角、后轮转角和横摆力矩作为控制输入,在副控制器实现车辆稳定性控制基础上,主控制器实现准确地车道线保持控制,保证智能车辆在车道内自主安全行驶.仿真结果表明,该串级控制策略的有效性,提高了智能车辆车道线跟踪的准确性,也提高车辆的稳定性和操纵性.     

兼顾可靠性和实时性的智能车辆车道保持北大核心CSCD

智能车辆车道保持的关键在于兼顾可靠性和实时性。为此,在车道识别时,利用已识别车道分区预测临近待识别分区车道候选位置,并仅在候选位置完成后续识别以提高识别实时性;在识别各分区时,依次采用分区阈值二值化、线性滤波、基于车道宽度滤波的方法,以提高识别的可靠性。同时,在提取车道线时根据2个检测区内的车道识别结果拟合不同曲线模型的目标车道线,并进行自适应预瞄控制。采用Lab View PXI8196和数字信号处理器F2812对智能车辆车道保持系统进行了设计。道路试验结果表明,提出的车道识别及跟踪控制方法同时保证车道识别及跟踪的可靠性和实时性,且稳定性较好。     

智能汽车自动驾驶的控制方法研究

为了解决智能汽车在无人驾驶的情况下自动跟随前方车辆行驶的问题,在预瞄跟随理论基础上提出一种自动驾驶的控制方法;该方法适用于控制一列智能车队,智能汽车通过接收前车发送的行驶状态来计算出前方路况,通过模糊自适应PID控制器来控制车辆驾驶;首先基于预瞄跟随理论设计一个汽车自动跟随模型,并指明需要跟随的物理量;然后,设计了一个模糊PID控制器来实现对给定物理量的跟踪;最后在dSPACE和飞思卡尔模型小车所搭建的实验环境下去验证控制方法的可行性;仿真实验结果表明该方法能够保证智能汽车具有良好的路况计算和车辆跟踪的精度,且具有较好的鲁棒性。     

一种提高无人车在弯道时操纵稳定性的算法

针对车辆弯道的行驶控制中,由于车辆模型维数高、强耦合及非线性强等特点,常规控制算法稳定性差.在深入分析车辆动力学的基础上,简化并建立了其整车动力学及轮胎动力学模型,以行车轨迹与预瞄轨迹的偏差作为输入,提出了一种基于偏差的纵向减速与横向TS-PD相结合的控制算法,实现车辆的弯道自动变速跟随预瞄轨迹的控制目标.仿真结果表明:上述算法与常规控制算法相对比,系统跟踪误差由1m缩小至0.2m,车辆前轮输出更为平滑,具有较强的鲁棒性.     

基于无模型自适应控制的自动泊车方案

提出一种基于无模型自适应控制的自动泊车方案.首先,通过车载传感器采集车周环境信息用于规划期望路径;然后,将自动泊车跟踪问题转化为预瞄偏差角跟踪问题,通过设计相应的无模型自适应控制算法实现自动泊车.该方案设计的优点是仅使用自动泊车过程中生成的前轮转角输入数据和预瞄偏差角输出数据,没有使用任何被控车辆的信息,因此可适用于不同车型. Matlab仿真以及与PID控制方案和Fuzzy控制方案的对比仿真结果验证了所提出方案的可行性.     

基于Lyapunov函数方法的时滞车辆纵向跟随控制

应用向量Lyapunov函数方法和比较原理，基于非线性车辆动态耦合模型，研究具有时间滞后的车辆跟随系统的指数稳定性问题，得到了车辆跟随系统的指数稳定性判据．根据滑模控制策略确定了车辆跟随系统的纵向控制规律，基于稳定性准则设计了车辆纵向跟随控制器参数．仿真结果表明，基于该方法设计的车辆纵向跟随控制器能使跟踪误差具有较快的收敛率．     

人机共驾智能车辆车道偏离横向辅助控制仿真

车道偏离辅助控制对于提高人机共驾智能车辆行驶安全性具有重要的现实意义,为此进行人机共驾智能车辆车道偏离横向辅助控制仿真研究.对车道线进行识别与跟踪,并采用基于车辆当前位置和基于车辆跨越车道线时间的联合算法进行智能车辆车道偏离判断;建立车辆系统模型和驾驶员模型,并通过基于转向和制动的可拓联合方法实现车辆偏离的横向辅助控制.实验结果表明,所研究方法在转向反应时间、车道偏离持续时间、最大转向角速度与最大横向偏离量四个指标的表现上更好,人机共驾智能车辆车道偏离横向辅助控制效果更佳.     

速度测量不确定性影响下的无人车路径跟踪预瞄控制EI北大核心CSCD

预瞄距离是无人车路径跟踪预瞄控制系统设计中的一个关键参数,其选取往往依赖于车辆行驶速度.然而,车速的测量不可避免存在偏差,这必将影响系统的预瞄控制性能.本文针对这一问题,在充分考虑车速测量不确定性对预瞄距离选取影响的基础上,提出一种基于线性参数变化(LPV)系统的鲁棒H∞路径跟踪预瞄控制算法.首先,选取车速和预瞄距离为调度变量,将原系统转化为一种具有可变权重因子的LPV系统.然后,针对无人车系统状态的不完全可测性,设计一种基于观测器的鲁棒H∞控制器.最后,给出保证闭环系统鲁棒H∞性能的凸优化求解条件.仿真结果验证了本文提出控制算法的有效性.     

基于混合整数规划的智能车横纵向一体化滚动优化决策

本文针对智能车辆的行为决策问题, 设计了基于混合整数规划的智能车横纵向一体化滚动优化决策方法. 该方法首先将纵向车速表示为非整数, 将期望车道表示为整数控制量, 建立了混合整数智能车决策简化模型; 然后, 设计了横纵向一体化滚动优化决策方法, 决策出纵向车速和换道动作, 根据系统输出与非线性约束的时域关系证明 了优化问题的递归可行性并通过遗传算法求解非线性混合整数规划优化问题. 基于车辆动力学仿真软件veDYNA 和Simulink进行了联合仿真, 并在红旗E-HS3智能车上开展了实车试验, 结果表明, 本文提出的基于混合整数规划的 智能车横纵向一体化决策方法能够实现超车、避障、跟车、停车和弯道工况下的行为决策.     

基于模型预测控制的电动车辆队列控制研究

针对车辆队列建模时参数不确定导致控制存在误差的问题,以及队列中跟随车辆稳定性问题,分析车辆纵向动力学,设计一个鲁棒MPC控制器和滑移率控制器来提高队列车辆的控制精度和稳定性.首先对纵向MPC控制器进行改进,提高车辆队列控制精度；同时为防止跟随车辆的轮胎打滑,设计一个MPC滑移率控制器对跟随车辆的轮胎滑移率进行控制约束,保证了跟随车辆的纵向稳定性.最后,进行仿真实验验证其有效性.仿真实验结果表明,与传统的LQR、MPC控制器相比,改进的鲁棒MPC纵向控制器控制精度更高,同时MPC滑移率控制器可防止跟随车辆的轮胎打滑,保证了跟随车辆的纵向稳定性.     

Studies on improving vehicle handling and lane keeping performance of closed-loop driver–vehicle system with integrated chassis control

This study proposes a new integrated robust model matching chassis controller to improve vehicle handling performance and lane keep ability. The design framework of the H_∞ controller is based on linear matrix inequalities (LMIs), which integrates active rear wheel steering control, longitudinal force compensation and active yaw moment control. To comprehensively evaluate the performance of the integrated chassis control system, a closed-loop driver–vehicle system is used. The effectiveness of the integrated controller on handling performance improvement is tested by a vehicle without driver model under a crosswind disturbance. At the same time, both the handling and lane keeping improving performance of the closed-loop driver–vehicle system is evaluated by tracking an S shape winding road. The simulation results reveal that the integrated chassis controller not only achieves preferable handling performance and stability, but also improves the vehicle lane keep ability significantly, and can alleviate the working load of the driver.     

智能车辆队列横纵向有限时间滑模控制

针对智能车辆队列横纵向控制及误差快速收敛问题,本文提出一种分布式横纵向有限时间滑模控制策略.首先,考虑跟踪误差的连锁反应及横纵向耦合效应,利用投影变换建立车辆队列横纵向误差模型,提出一种车辆队列横纵向控制框架.而后,针对误差快速收敛问题,设计非奇异积分终端滑模面(NITSM)与自适应幂次积分趋近律(APIRL),通过构造Lyapunov函数分析系统的有限时间稳定性与队列稳定性.最后,基于Trucksim/Simulink联合仿真以及实车实验进一步验证了本文方法的有效性.结果表明,本文所提方法能保证队列稳定性,并实现误差快速收敛,规避跟踪误差的连锁反应及车辆横向运动对纵向车间距误差的影响.     

基于图与势场法的多车道编队控制

多车协同驾驶能显著提高交通安全和效率,是未来5G网联自动驾驶技术的重要应用场景之一.传统上,多车协同驾驶的主要形式为单一车道上的无人车队列,其队列稳定性受队列长度、通信距离及延迟的限制.本文提出一种无人车编队方法,将单车道队列扩展为多车道护航编队.针对不同场景下的需求设计多车道编队调整策略,结合基于图的分布式控制,完成任意预定义的编队结构;同时,利用势场法对行车环境建立势场模型,实现无人车的避障轨迹规划,提高编队的避障能力;最后,结合纵横向控制器,实现无人车多车道护航编队控制.仿真实验表明,本文提出的无人车多车道护航编队方法,能适应不同交通场景,如道路变化、障碍车运动等,完成自动变换编队结构,实现安全、高效通行.     

Resilient and Safe Platooning Control of Connected Automated Vehicles Against Intermittent Denial-of-Service Attacks

Connected automated vehicles (CAVs) serve as a promising enabler for future intelligent transportation systems because of their capabilities in improving traffic efficiency and driving safety, and reducing fuel consumption and vehicle emissions. A fundamental issue in CAVs is platooning control that empowers a convoy of CAVs to be cooperatively maneuvered with desired longitudinal spacings and identical velocities on roads. This paper addresses the issue of resilient and safe platooning control of CAVs subject to intermittent denial-of-service (DoS) attacks that disrupt vehicle-to-vehicle communications. First, a heterogeneous and uncertain vehicle longitudinal dynamic model is presented to accommodate a variety of uncertainties, including diverse vehicle masses and engine inertial delays, unknown and nonlinear resistance forces, and a dynamic platoon leader. Then, a resilient and safe distributed longitudinal platooning control law is constructed with an aim to preserve simultaneous individual vehicle stability, attack resilience, platoon safety and scalability. Furthermore, a numerically efficient offline design algorithm for determining the desired platoon control law is developed, under which the platoon resilience against DoS attacks can be maximized but the anticipated stability, safety and scalability requirements remain preserved. Finally, extensive numerical experiments are provided to substantiate the efficacy of the proposed platooning method.      

Integrated Dynamics Control and Energy Efficiency Optimization for Overactuated Electric Vehicles

A large number of studies have been conducted on the dynamics control of electric vehicles or on the optimization of their energy efficiency but few studies have looked at both of these together. In this study, an integrated dynamics control and energy efficiency optimization strategy is proposed for overactuated electric vehicles, where the control of both longitudinal and lateral dynamics is dealt with while the energy efficiency is optimized. First, considering the trade‐off between control performance and energy efficiency, criteria are defined to categorize the vehicle motion status as linear pure longitudinal motion and non‐linear motion or turning motion. Then different optimization targets are developed for different motion status. For the pure linear longitudinal motion and cornering motion, the energy efficiency and vehicle dynamics performance are equally important and a trade‐off control performance between them needs to be achieved. For the non‐linear turning motion, vehicle handling and stability performance are the primary concerns, and energy efficiency is a secondary target. Based on the defined targets, the desired longitudinal and lateral tyre forces and yaw moment are then optimally distributed to the wheel driving and steering torques. Finally numerical simulations are used to verify the effectiveness of the proposed strategies. The simulation results show that the proposed strategies can provide good dynamics control performance with less energy consumption.     

Coordinated Control Architecture for Motion Management in ADAS Systems

Advanced driver assistance systems (ADAS) seek to provide drivers and passengers of automotive vehicles increased safety and comfort. Original equipment manufacturers are integrating and developing systems for distance keeping, lane keeping and changing and other functionalities. The modern automobile is a complex system of systems. How the functionalities of advanced driver assistance are implemented and coordinated across the systems of the vehicle is generally not made available to the wider research community by the developers and manufactures. This paper seeks to begin filling this gap by assembling open source physics models of the vehicle dynamics and ADAS command models. Additionally, in order to facilitate ADAS development and testing without having access to the details of ADAS, a coordinated control architecture for motion management is also proposed for distributing ADAS motion control commands over vehicle systems. The architecture is demonstrated in a case study where motion is coordinated between the steering and the braking systems, which are typically used only for a single functionality. The integrated vehicle and system dynamics using the coordinated control architecture are simulated for various driving tasks. It is seen that improved trajectory following can be achieved by the proposed coordinated control architecture. The models, simulations and control architecture are made available for open access.      

Centralized and decentralized distributed control of longitudinal vehicular platoons with non‐uniform communication topology

In this paper, the distributed control of a longitudinal platoon of vehicles with non‐uniform communication topology is studied. In the case of non‐uniform communication topology, some eigenvalues of the network's matrix may be complex which complicates the stability analysis of the platoon. Most previous studies on vehicular platooning focus mainly on uniform topologies such as uni‐directional, bi‐directional, and multi predecessors following. Since all eigenvalues of these topologies are real, the stability analysis can be performed in a straightforward manner. A third‐order linear differential model is employed to describe the upper‐level dynamics of each vehicle. The 3 N‐order closed‐loop dynamics of the platoon are decoupled to individual third‐order dynamics by presenting a new approach. Two new centralized and decentralized control protocols are introduced to perform the stability analysis of the closed‐loop dynamics. A constant time headway strategy is employed to adjust the inter‐vehicle spacing. Simulation results with different scenarios are presented to illustrate the effectiveness of the proposed approaches.