Transient Trajectory Shaping Control Strategy Development for an Autonomous Capable Ground Vehicle

车辆稳定与安全控制一直是自动驾驶车辆控制系统设计的主要难点。本研究的目的是为具有自主能力的地面车辆开发一种系统的瞬态控制策略，以提高车辆在不利条件下的整体瞬态性能和安全性。本研究采用四轮独立驱动（FWID）汽车结构。在这种结构中，地面车辆中的每个车轮可以独立驱动，例如，由轮内（或轮毂）电机驱动。FWID汽车被认为是提高车辆行驶安全性和轻量化的有效途径。在FWID车辆中，可以通过左右车轮之间的扭矩差产生额外的偏航力矩来控制车辆的偏航运动。诱导的外部偏航力矩与主动转向系统的自动快速转向动作一起，使同时调节车辆横向速度和跟踪所需的偏航率成为可能。因此，车辆的性能，安全性和弹道能力可以显著提高。此外，通过采用车轮直接驱动结构，由于简化了传动系统，可以大大减轻车辆的总重量。尽管四轮独立驱动（FWID）结构可以大大提高车辆性能，但控制复杂性、模型不确定性和车辆轻量化等挑战仍然是一个长期问题。在本文的研究中，这些问题将通过以下框架来解决：基于车辆动力学、行驶条件和安全缓冲来建模近未来水平的车辆安全边界；通过轨迹整形控制策略将飞行器的瞬态轨迹控制在边界内。为了解决模型的不确定性问题，将采用模型驱动和数据驱动相结合的方法在线重新配置面向控制的模型，并实时定义安全边界。对于路径跟踪控制，本研究将进一步研究和应用申请人先前工作中开发的一种新的瞬态轨迹成形控制策略。正如申请人研究小组的一些最新成果，车辆侧滑角可以被预先定义的安全边界所限制。通过实时估计车辆安全运行区域和控制执行器约束确定安全边界。在提出的策略中，成形概念的新颖之处在于不强迫车辆严格遵循期望的轨迹。取而代之的是，四轮驱动产生的额外偏航力矩将用于保证车辆在正常转向时处于不确定的安全边界内。

The vehicle stabilization and safety control have long been major obstacles in the vehicle control system design for autonomous capable vehicles. The objective of this proposed research is to develop a systematic transient control strategy for an autonomous capable ground vehicle to improve the overall vehicle transient performance and safety in adverse conditions. In this research, a Four-Wheel-Independent-Drive (FWID) vehicle structure is adopted. In such a structure, each of the wheels in a ground vehicle can be independently driven, for example, by an in-wheel (or hub) motor. The FWID vehicle is considered to be an effective way of enhancing vehicle driving safety and light weighting. In a FWID vehicle, an additional yaw moment can be generated with torque differences between the left and right wheels to control the vehicle yaw motion. The induced external yaw moment together with the automatic and fast steering action from an active steering system, make it possible to simultaneously regulate the vehicle lateral velocity and track the desired yaw rate. Thus, the vehicle performance, safety, and trajectory capability can be significantly improved. In addition, by using a wheel direct drive structure, the total weight of the vehicle can be substantially reduced due to the simplified transmission system. Although the Four-Wheel-Independent-Drive (FWID) structure can substantially improve vehicle capability, the challenges in control complexity, model uncertainties, and light vehicle weight are still long term issues. In the proposed research, these issues will be solved through the following frames: the vehicle safety boundary in the near future horizon will be modeled based on the vehicle dynamics, driving conditions, and safety buffer; the vehicle transient trajectory will be controlled within the bounds via a trajectory shaping control strategy. To deal with the model uncertainty issue, a combined model-driven and data-driven approach will be used to reconfigure the control-oriented model online and define the safety boundary in real time. For the path tracking control, a novel transient trajectory shaping control strategy, developed in the applicant's prior work, will be further investigated and applied in this research. As some most recent results in the applicant's research group, vehicle side slip angle can be bounded by a pre-defined safety boundary. The safety boundary are determined by real-time estimating the vehicle safe operating region and control actuator constraints. The novelty of the shaping concept in the proposed strategy is not to force the vehicle strictly to follow a desired trajectory. Instead, the additional yaw moment from four wheel actuations will be used to guarantee the vehicle within the safety boundary given the uncertainties while the normal steering.