主动铰接式拖挂机器人柔顺转向机理与动态协调避障方法

Tractor-trailer robots adopt modular structure such that they can change shape and has a greater payload, thereby being widely used in recent years. However, the traditional passively articulated tractor-trailer robots have poor steering ability, leading to likely collision with the obstacles. To this end, this project reports a new type of actively articulated tractor-trailer robot. The research contents are mainly as follows: a systematic configuration method for the actively articulated tractor-trailer robot based on steering motor and anti-vibration spring, is proposed; the mixed rigid-flexible characteristic is explored and the performance index of flexibility is put forward afterwards; the dynamic transfer laws in robot system with respect to the ground adhesion force, the hook traction force, longitudinal driving and actively rotational torques, are clarified; the full-state systematic kinematic and dynamic models to describe all information of tractor-trailer robot are established, respectively; and then the actively flexible steering mechanism for this type of robot is revealed. In the following, with the aid of the theory of polynomials, an unified obstacle avoidance path planning method, aiming to various working conditions including point parking and trajectory tracking, are presented as considering multi-level constraints; a fast robust observation theory for the robot sideslip angle is proposed based on terminal sliding mode technology. At last, after exploring hierarchical design mechanism of the entire control system, a dynamically coordinated obstacle avoidance approach for the tractor-trailer robot is proposed, in which the planning, perceiving and controlling are effectively integrated together. This project devotes to enhancing actively flexible steering and coordinated obstacle avoidance capabilities for the tractor-trailer robot, and it looks forward to providing theoretical foundation and technical support for the robot widely application.

拖挂机器人采用模块化设计，能够变形、运载量大，近年来得到了广泛的应用，但是受欠驱动多体结构的影响，传统的被动铰接式拖挂机器人转向能力差、常与障碍物发生碰撞，为此本项目发明一种新型的主动铰接式拖挂机器人。研究内容包括：提出基于转向电机和抑振弹簧混合驱动下的主动铰接式拖挂机器人构型综合方法，研究机器人的刚柔混合特性并建立柔顺性的性能指标，阐明地面附着力、挂钩牵引力、纵向驱动与主动转向力矩的动态传递规律，建立描述系统所有状态变化的全信息运动学与动力学模型，进而揭示拖挂机器人主动柔顺转向机理；采用多项式理论，提出多源约束下定点泊车-轨迹跟踪多工况统一避障路径规划方法，研究基于终端滑模技术的侧滑角快速鲁棒观测原理，探索控制系统分层设计机制，最终形成一套集规划-感控一体化的拖挂机器人动态协调避障方法。本项目致力于增强拖挂机器人的主动柔顺转向与协调避障能力，为其开发应用提供理论依据和技术支撑。

拖挂机器人采用模块化设计，具有载货量大、能够变形、机动灵活、运货成本低以及能源消耗少等优点，近年来得到了广泛的应用。然而传统的被动铰接式拖挂机器人转向能力差，尤其是遭遇狭窄路况时，挂车的位姿难以控制，常与障碍物发生碰撞。本项目针对上述问题开展研究，提出了基于转向电机和抑振弹簧混合驱动下的主动铰接式拖挂机器人新构型，揭示了拖挂机器人主动柔顺转向机理，通过采用麦格纳姆轮减小轮地间的摩擦效应；在此基础上，结合多项式优化轨迹和模型预测控制技术，提出复杂环境下的机器人协调换道策略；采用分层控制策略，将系统的状态观测器设计方法、避障路径规划理论，动态协调控制策略有机融合，依次解决拖挂机器人相关的感知、决策、规划和控制一系列问题。在资助期间内，共发表期刊论文26篇(第一标注12篇、第二标注各14篇)，其中23篇SCI、1篇SSCI，论文均发表在IEEE TII、IEEE TSMCS、IJC, ASC和Nonlinear Dyn等本领域权威期刊上；发表会议论文13篇(第一标注11篇，第二标注2篇)，论文中的学术思想和理论方法得到专家和学者的广泛引用。同时，为宣传研究成果，课题组积极参加各类学术会议，获得国际会议“最佳学术论文”和“最佳学生论文”等提名奖项；在研期间获得授权发明专利16项；在人才培养方面，资助期间毕业硕士研究生12人、毕业博士研究生2人，学生毕业论文先后获得辽宁省优秀硕士学位论文1篇，校优秀硕士学位论文6篇，2名已毕业博士生均获得国家奖学金，多人获得省级和市级优秀毕业生称号；在成果推广方面，课题组着重将拖挂机器人的协调控制、路径规划等理论和方法进行推广，目前已经与多家企业建立起合作关系。上述理论研究和学术成果，能够显著提高拖挂机器人的运动精准性和操控灵活性，促进欠驱动拖挂机器人工程应用与推广，具有理论研究和实际应用的双重价值。