MPC-BASED steering control for backward-driving vehicle using stereo vision

We propose a steering control algorithm for autonomous backward driving in a narrow corridor. Passable spaces are detected using a stereo camera, and the steering angle is controlled by a model predictive controller (MPC). For passable space detection, an UV-disparity map is calculated from the original disparity map. Information regarding passable spaces collected by the stereo camera is used in steering control. Backward driving requires the driver’s preemptive actions, which can be learned by experience because of the non-intuitive responses (the initial motion of the vehicle is opposite to the driver’s steering angle input). This occurs because a backward-driving vehicle is a non-minimum phase system. One of the most popular steering control algorithms is Stanley method, which is based on the feedback of lateral displacement error and heading angle error. The method is very intuitive and works well for forward driving, but it exhibits significant undershoot for backward driving cases. Furthermore, the method does not explicitly consider any constraints on control inputs and states. We designed a steering controller based on the MPC technique that requires future information but can handle constraints explicitly. Because we have near-future information from the stereo camera under limited passable spaces, MPC can be effectively implemented. We performed several simulations and experiments to show the performance and superiority of the suggested method over a simple feedback-based control algorithm.     

驾驶模拟器在自动驾驶系统中的应用研究

自动驾驶车辆在实际道路上行驶之前的测试阶段是一个至关重要的环节。一个低成本、高效率以及高精度测量的自动驾驶车辆的测试方式,对于自动驾驶车辆的开发具有重要意义。将驾驶模拟器运用到研究自动驾驶车辆测试已是近年来的一个研究热点。基于虚拟驾驶场景的自动驾驶车辆的检测,通过组合虚拟驾驶场景的背景车辆、行人、交通灯、建筑、指示标牌等元素,研究将驾驶模拟器与虚拟驾驶场景的联合应用来测试自动驾驶车辆。设计了典型的交通场景,通过自动驾驶车辆和背景车辆的实时交互,研究自动驾驶车辆的各项性能指标。研究结果表明:该驾驶模拟器可以高度拟合人类驾驶体验,驾驶员通过驾驶模拟器控制背景车辆能够很好的模拟现实中的驾驶行为,对自动驾驶车辆的仿真测试起到了促进作用。     

基于变权重系数的分布式驱动无人车轨迹跟踪

为了提高分布式无人车轨迹跟踪的精度，提出了基于自主与差动协调转向控制的轨迹跟踪方法。首先，在车辆三自由度模型基础上，基于模型预测控制（MPC）实时计算前轮转角以控制车辆进行自主转向轨迹跟踪。在此过程中，为了提高自主转向下车辆的轨迹跟踪精度与行驶的稳定性，考虑多种因素，利用经验公式及神经网络控制对MPC的预瞄步数和预瞄步长进行多参数调整，实现预瞄时间的自适应控制。其次，在恒转矩需求的情况下，以轨迹偏差为PID控制器的输入及左右轮毂电机转矩为输出进行差动转向控制，实现了差动转向下的轨迹跟踪控制。然后，通过设置权重系数的方法将自主与差动转向相结合。考虑到车辆横纵向动力学因素，采用模糊控制及经验公式对权重系数进行了调整，从而在提高车辆转向灵活性与轨迹跟踪效果的同时保证车辆行驶的稳定性。CarSim与Simulink联合仿真以及实车试验结果表明：与自主转向轨迹跟踪相比，采用变权重系数的协调控制可以在不同的工况下提高车辆的转向灵活性与轨迹跟踪的精度，轨迹跟踪偏差的均方根值改善率达到了11%。所提出的协调转向控制方法可为分布式驱动车辆转向灵活性的提高及轨迹跟踪精度的改善提供一种新的思路。     

人机协作系统中车辆轨迹规划与轨迹跟踪控制研究

自动驾驶系统需具备响应驾驶人意图且有效执行驾驶人意图的能力，以解决人机协作系统中存在的人机冲突、人机优势融合等问题。提出决策层“以人为主”、执行层“以机为首”的人机协作关系，构建包含驾驶人意图识别模块、基于意图识别的轨迹规划模块与轨迹跟踪控制模块的人机协作一体化控制系统框架，并重点对轨迹规划模块与轨迹跟踪控制模块开展研究。首先，结合双向长短期记忆神经网络（Bi-directional Long Short Term Memory，Bi-LSTM）与注意力机制模型建立换道轨迹规划模型；在改进人工势场算法中引入模型预测控制并建立避险轨迹规划模型。其次，通过开展驾驶模拟器试验建立换道与避险驾驶行为数据集，为拟人化模型训练和模型参数确定提供支撑。然后，综合考虑车辆状态变量、控制输入与输出以及道路结构参数等约束条件，构建基于最优转向前轮输入的线性时变模型预测轨迹跟踪控制器，实现对规划轨迹的精准跟踪。最后，基于驾驶模拟器搭建人机协作系统硬件在环测试平台，对轨迹规划模块与轨迹跟踪控制模块开展硬件在环测试与验证。结果表明：换道与避险规划轨迹光滑且平稳，轨迹跟踪控制过程中，车辆航向角与前轮转角变化平稳；所构建的轨迹规划与轨迹跟踪控制模块在确保安全性前提下可实现不同场景中的车辆运动控制需求。     

An investigation on motor-driven power steering-based crosswind disturbance compensation for the reduction of driver steering effort

This paper describes a lateral disturbance compensation algorithm for an application to a motor-driven power steering (MDPS)-based driver assistant system. The lateral disturbance including wind force and lateral load transfer by bank angle reduces the driver's steering refinement and at the same time increases the possibility of an accident. A lateral disturbance compensation algorithm is designed to determine the motor overlay torque of an MDPS system for reducing the manoeuvreing effort of a human driver under lateral disturbance. Motor overlay torque for the compensation of driver's steering torque induced by the lateral disturbance consists of human torque feedback and feedforward torque. Vehicle–driver system dynamics have been investigated using a combined dynamic model which consists of a vehicle dynamic model, driver steering dynamic model and lateral disturbance model. The human torque feedback input has been designed via the investigation of the vehicle–driver system dynamics. Feedforward input torque is calculated to compensate additional tyre self-aligning torque from an estimated lateral disturbance. The proposed compensation algorithm has been implemented on a developed driver model which represents the driver's manoeuvreing characteristics under the lateral disturbance. The developed driver model has been validated with test data via a driving simulator in a crosswind condition. Human-in-the-loop simulations with a full-scale driving simulator on a virtual test track have been conducted to investigate the real-time performance of the proposed lateral disturbance compensation algorithm. It has been shown from simulation studies and human-in-the-loop simulation results that the driver's manoeuvreing effort and a lateral deviation of the vehicle under the lateral disturbance can be significantly reduced via the lateral disturbance compensation algorithm.     

Haptic steering support for driving near the vehicle’s handling limits; skid-pad case

Current vehicle dynamic control systems from simple yaw control to high-end active steering support systems are designed to primarily actuate on the vehicle itself, rather than stimulate the driver to adapt his/her inputs for better vehicle control. The driver though dictates the vehicle’s motion, and centralizing him/her in the control loop is hypothesized to promote safety and driving pleasure. Exploring the above statement, the goal of this study is to develop and evaluate a haptic steering support when driving near the vehicle’s handling limits (Haptic Support Near the Limits; HSNL). The support aims to promote the driver’s perception of the vehicle’s behaviour and handling capacity (the vehicle’s internal model) by providing haptic (torque) cues on the steering wheel. The HSNL has been evaluated in (a) driving simulator tests and (b) tests with a vehicle (Opel Astra G/B) equipped with a variable steering feedback torque system. Drivers attempted to achieve maximum velocity while trying to retain control in a circular skid-pad. In the simulator (a) 25 subjects drove a vehicle model parameterised as the Astra on a dry skid-pad while in (b) 17 subjects drove the real Astra on a wet skid-pad. Both the driving simulator and the real vehicle tests led to the conclusion that the HSNL assisted subjects to drive closer to the designated path while achieving effectively the same speed. With the HSNL the drivers operated the tires in smaller slip angles and hence avoided saturation of the front wheels’ lateral forces and excessive understeer. Finally, the HSNL reduced their mental and physical demand.     

基于人机共驾的车道保持辅助控制系统研究（双语出版）

车道保持控制系统是汽车安全辅助驾驶的重要组成部分，可有效提高汽车主动安全性、避免车辆无意识地偏离本车道。目前，大部分车道保持控制系统在工作时将驾驶人的操作视为外界干扰，没有考虑人机共驾阶段下驾驶人与控制系统的控制权分配问题，易造成人机冲突、影响驾驶人的驾驶感受。论文兼顾驾驶人与辅助控制系统各自优势，基于人机共驾技术对车道保持控制系统进行研究。构建基于安全行驶区域与最晚预警边界相结合的车道偏离决策模型，在保证其预警精度的同时降低计算复杂性，根据车辆行驶状态和路面附着系数动态调整预警阈值；研究串级MPC-PID控制策略实现对车辆横向位置的控制，将最优问题转化为二次规划求得目标前轮转角，利用PID算法完成对目标前轮转角的跟踪；引入共驾系数对车辆的控制权进行分配，研究共驾系数分配模型，以车辆状态误差和驾驶人转向力矩作为模糊控制的输入变量、共驾系数作为输出变量，降低辅助控制系统与驾驶人之间的冲突；最后，利用CarSim与Simulink联合仿真对所研究的控制策略进行仿真验证，结果表明共驾系数能够根据驾驶人的操作和车辆运行状态的变化实现动态调整，辅助控制力矩与驾驶人输入力矩变化趋势相同，在保留驾驶人一定操作的基础下可避免车辆偏离车道、降低人机冲突。     

An adaptive lateral preview driver model

Successful modelling and simulation of driver behaviour is important for the current industrial thrust of computer-based vehicle development. The main contribution of this paper is the development of an adaptive lateral preview human driver model. This driver model template has a few parameters that can be adjusted to simulate steering actions of human drivers with different driving styles. In other words, this model template can be used in the design process of vehicles and active safety systems to assess their performance under average drivers as well as atypical drivers. We assume that the drivers, regardless of their style, have driven the vehicle long enough to establish an accurate internal model of the vehicle. The proposed driver model is developed using the adaptive predictive control (APC) framework. Three key features are included in the APC framework: use of preview information, internal model identification and weight adjustment to simulate different driving styles. The driver uses predicted vehicle information in a future window to determine the optimal steering action. A tunable parameter is defined to assign relative importance of lateral displacement and yaw error in the cost function to be optimized. The model is tuned to fit three representative drivers obtained from driving simulator data taken from 22 human drivers.     

商用车电液复合转向系统的车道保持策略

为了提高商用车的行驶安全性,避免因驾驶人的分心驾驶出现车辆偏离车道的问题,提出一种基于电液复合转向系统的商用车车道保持策略;在建立电液复合转向系统模型、二自由度车辆模型、预瞄驾驶人模型的基础上,设计基于驾驶人在环的MPC和ADRC串级的车道保持控制策略。首先,采用MPC算法将车辆横向位置控制的最优问题转化为二次规划求得目标前轮转角;然后,考虑电液复合转向系统的不确定和干扰问题,利用ADRC算法对目标转向盘转角和实际驾驶人的转向盘转角差值以转矩信号的形式进行补偿。同时研究车道保持系统对驾驶人的干预问题,引入干预系数的概念,采用模糊控制的方法,将驾驶人手力和车辆的运动状态作为输入变量,干预系数作为输出变量,保证整车行驶安全性的前提下减小车道保持辅助系统对驾驶人的干预。最后,通过MATLAB/Simulink仿真和硬件在环试验对所设计的控制策略进行验证。研究结果表明：所设计的基于商用车电液复合转向系统的车道保持策略能够及时地纠正因驾驶人的分心驾驶而导致车辆偏离所在行驶车道的行为,特别是在弯道处出现驾驶人转向不足或过度转向的情况时,能够将车辆维持在车道线之内,保证车辆的行驶安全性,同时由于干预系数的设计,使得驾驶人也有良好的人机交互体验感。     

电子差速系统轮胎附着特性模拟仿真研究

电子差速系统相对于传统的机械式差速器可以实现转矩的精准分配,根据轮胎的纵向运动特性以及侧向运动特性,结合轮胎滑移率让内外侧车轮在过弯时拥有足够的附着力,减小整车的横摆角速度,提高过弯稳定性。采用后轮双电机的驱动方案,驱动电机采用直接转矩控制的方法,由整车控制器将指定的计算转矩信号发送给电机控制器完成动力分配,所需转矩根据驾驶员的加速踏板及方向盘转角,运用阿克曼转向模型计算得到。     

Driver propensity characterization for different forward collision warning times

The forward collision warning system, which warns danger to the driver after sensing possibility of crash in advance, has been actively studied recently. Such systems developed until now give a warning, regardless of driver’s driving propensity. However, it’s not reasonable to give a warning to every driver at the same time because drivers are different in driving propensity. In this study, to give a warning to each driver differently, three metrics classifying driver’s driving propensity were developed by using the driving data on a testing ground. These three metrics are the predicted time headway, required deceleration divided by the deceleration of the leading vehicle, and the resultant acceleration divided by the deceleration of the leading vehicle. Driving propensity was divided into 3 groups by using these metrics for braking and steering cases. In addition, these metrics were verified by making sure that braking propensity could be classified on public roads as well.     

驾驶机器人车辆动态制动力矩补偿

为了减小长期自动驾驶过程中制动性能下降带来的影响,提出了一种驾驶机器人车辆动态制动力矩补偿方法。首先建立了以车速和制动踏板力为输入,制动力矩为输出的驾驶机器人车辆制动性能离线自学习模型。然后考虑到驾驶机器人车辆长期自动驾驶导致离线自学习模型可靠性下降,建立了以车速和制动踏板力为输入,制动力矩为输出的扩展自回归在线辨识模型,并采用模糊变遗忘因子递推最小二乘法进行参数辨识。模糊变遗忘因子递推最小二乘法通过引入遗忘因子的方式,对数据施加时变加权系数,以避免出现数据增长导致的数据饱和现象。模糊变遗忘因子控制器以制动力矩辨识误差为输入,经模糊规则推理实时输出合适的遗忘因子进行参数辨识,能够有效均衡驾驶机器人车辆制动性能参数辨识的稳定性与收敛速度。驾驶机器人车辆自动驾驶过程中,根据当前车速与目标车速的大小计算出所需的制动力矩,加上反馈回来的制动力矩误差,并结合当前时刻的车速,利用制动性能离线自学习模型与机械腿逆向运动学模型实时计算出制动电机输出位移量,实现对驾驶机器人车辆制动力矩的在线补偿。仿真与试验结果表明:利用所提出的方法对车辆动态制动力矩进行辨识时,通过调节遗忘因子,辨识结果能够快速收敛且辨识误差较小。在此基础上,控制驾驶机器人车辆进行纵向车速跟踪时,能够有效减小制动性能下降造成的影响,保证控制车速跟踪误差在±1km·h-1之内。     

Comparison of All-Wheel Steerings in the System Driver-Vehicle

Different load or tires and a drive on an ice-coated road can overcharge a driver to such an extend, that the result may be an accident. Therefore the aim of development is a self-acting compensation of the vehicle to different vehicle transfer behaviour (invariant vehicle behaviour).

The calculation of so called optimal characteristics shows, that only rear-wheel steering cannot realize this aim of development. Therefore an additional front-wheel angle, which is not influenced by the driver, is necessary. A transfer function can be calculated in order to get controlled steering of the rear wheels without the influence of load.

It is not possible to realize optimal characteristics, because the parameters of the vehicle are difficult to measure. Only an optimal diagnosis and control of driving condition realize a relief for the driver in every driving situation in order to avoid most of the accidents.

The often demanded sideslip angle compensation only worsens driving conditions on ice-coated roads. Therefore systems which identify the driving condition themselves have to be favoured in any case.     

A path-following driver–vehicle model with neuromuscular dynamics,including measured and simulated responses to a step in steering angle overlay

An existing driver–vehicle model with neuromuscular dynamics is improved in the areas of cognitive delay, intrinsic muscle dynamics and alpha–gamma co-activation. The model is used to investigate the influence of steering torque feedback and neuromuscular dynamics on the vehicle response to lateral force disturbances. When steering torque feedback is present, it is found that the longitudinal position of the lateral disturbance has a significant influence on whether the driver’s reflex response reinforces or attenuates the effect of the disturbance. The response to angle and torque overlay inputs to the steering system is also investigated. The presence of the steering torque feedback reduced the disturbing effect of torque overlay and angle overlay inputs. Reflex action reduced the disturbing effect of a torque overlay input, but increased the disturbing effect of an angle overlay input. Experiments on a driving simulator showed that measured handwheel angle response to an angle overlay input was consistent with the response predicted by the model with reflex action. However, there was significant intra- and inter-subject variability. The results highlight the significance of a driver’s neuromuscular dynamics in determining the vehicle response to disturbances.     

匝道合流交通场景下自动驾驶汽车安全性测试评价方法

矿用无人运输车辆作业环境恶劣,存在大曲率弯道、坡道等非结构化道路明显特征,对无人化运输控制要求高。为改善PID等传统控制算法适应性问题,提高无人驾驶轨迹跟踪的车辆横纵向控制精度,提出一种纯跟踪与PID结合的多点预瞄横向控制、考虑模糊控制表参数拟合的纵向控制方法,减少控制参数的同时提高算法效果。根据传统控制算法设计基础控制器,结合基础算法优势进行横向与纵向控制算法设计,通过硬件在环仿真和实车测试验证算法的性能。试验结果表明,横向控制算法与斯坦利算法相比,车辆路径跟踪精度有明显改善,纵向控制方面,速度跟随误差<1 km/h,保证了车辆驾驶时的平稳性与舒适性。     

磁流变半主动悬架研究及实车试验分析

针对车辆半主动悬架系统的整车协调控制,通过悬架动力学模型分析了耦合量的影响,提出了一种主从控制方法。基于自行研制的并联常通孔式磁流变减振器和控制系统开展了实车道路试验。在越野路行驶时,驾驶员坐垫处的加权加速度降低了13.8%~42.6%,车身俯仰角速度降低了21.1%~53.7%;蛇行试验中车身侧倾角速度、角度分别平均降低了65%和38.5%;变道试验中车身侧倾角速度、角度分别平均下降65%和51%。综上所述,研制的磁流变悬架系统显著地提升了车辆的乘坐舒适性、操纵稳定性和行驶安全性。     

Comparison of All-Wheel Steerings in the System Driver-Vehicle

SUMMARY

Different load or tires and a drive on an ice-coated road can overcharge a driver to such an extend, that the result may be an accident. Therefore the aim of development is a self-acting compensation of the vehicle to different vehicle transfer behaviour (invariant vehicle behaviour).

The calculation of so called optimal characteristics shows, that only rear-wheel steering cannot realize this aim of development. Therefore an additional front-wheel angle, which is not influenced by the driver, is necessary. A transfer function can be calculated in order to get controlled steering of the rear wheels without the influence of load.

It is not possible to realize optimal characteristics, because the parameters of the vehicle are difficult to measure. Only an optimal diagnosis and control of driving condition realize a relief for the driver in every driving situation in order to avoid most of the accidents.

The often demanded sideslip angle compensation only worsens driving conditions on ice-coated roads. Therefore systems which identify the driving condition themselves have to be favoured in any case.     

Vehicle dynamics and road curvature estimation for lane departure warning system using robust fuzzy observers: experimental validation

This paper describes a new approach to estimate vehicle dynamics and the road curvature in order to detect vehicle lane departures. This method has been evaluated through an experimental set-up using a real test vehicle equipped with the RT2500 inertial measurement unit. Based on a robust unknown input fuzzy observer, the road curvature is estimated and compared to the vehicle trajectory curvature. The difference between the two curvatures is used by the proposed lane departure detection algorithm as the first driving risk indicator. To reduce false alarms and take into account driver corrections, a second driving risk indicator based on the steering dynamics is considered. The vehicle nonlinear model is deduced from the vehicle lateral dynamics and road geometry and then represented by an uncertain Takagi–Sugeno fuzzy model. Taking into account the unmeasured variables, an unknown input fuzzy observer is proposed. Synthesis conditions of the proposed fuzzy observer are formulated in terms of linear matrix inequalities using the Lyapunov method.     

基于驾驶特性的超车模型研究

超车行驶作为驾驶人行车过程中重要的行为之一,与行驶安全性有着直接的联系。为建立符合驾驶人操作习惯的超车模型,本文通过实车试验采集不同驾驶人在高速公路的超车行驶数据,并以此采用多项式回归拟合建立基于驾驶人操作特性的超车模型,最后利用prescan软件对提出的超车模型进行了仿真分析,结果表明建立的超车模型能够真实地反映驾驶人超车过程中的操作习惯,为超车行为的研究提供了可靠的理论依据。     

Development of an errorable car-following driver model

An errorable car-following driver model is presented in this paper. An errorable driver model is one that emulates human driver’s functions and can generate both nominal (error-free), as well as devious (with error) behaviours. This model was developed for evaluation and design of active safety systems. The car-following data used for developing and validating the model were obtained from a large-scale naturalistic driving database. The stochastic car-following behaviour was first analysed and modelled as a random process. Three error-inducing behaviours were then introduced. First, human perceptual limitation was studied and implemented. Distraction due to non-driving tasks was then identified based on the statistical analysis of the driving data. Finally, time delay of human drivers was estimated through a recursive least-square identification process. By including these three error-inducing behaviours, rear-end collisions with the lead vehicle could occur. The simulated crash rate was found to be similar but somewhat higher than that reported in traffic statistics.