【摘要】 本文面向域控制的无人矿卡提出了一种具备故障冗余机制的切换模型预测控制轨迹跟踪策略。首先建立了运动学与动力学两种不同复杂度的预测模型，提出了综合执行器性能、跟踪精度、稳定性边界的多维度控制约束与成本函数，构建了基于误差反馈与预瞄曲率的自适应权重调节器，提高轨迹跟踪策略的工况适应性和稳定性。在此基础上提出了智能驾驶域跟踪控制功能失效场景下的底盘域控制器跟踪控制冗余机制，保障故障下车辆跟踪控制的最差性能可接受。最后，通过离线仿真验证了所提轨迹跟踪策略与非线性模型预测控制策略相比，在保证跟踪效果相当的同时提高了实时性。硬件在环仿真表明，所提策略提高了车辆弯道工况的轨迹跟踪精度和平稳性，满足实际运输场景需求，面向突发的跟踪控制功能失效可实现稳定停靠。更多还原

【Abstract】 This paper proposes a switched model predictive control(SMPC) trajectory tracking strategy with a fault-tolerant redundancy mechanism for domain-controlled unmanned heavy-duty mining transport vehicles. Firstly, two predictive models of different complexity—kinematic and dynamic—are established. A multi-dimensional control constraint and cost function are proposed, taking into account of actuator performance, tracking accuracy, and stability boundaries. An adaptive weight regulator based on error feedback and preview curvature is constructed to enhance the adaptability and stability of the trajectory tracking strategy under varying working conditions. Furthermore, a redundancy mechanism is introduced for the chassis domain controller to maintain tracking control performance in the event of a failure in the intelligent driving domain′s tracking control function, which ensures that even under fault conditions, the vehicle′s tracking performance remains within an acceptable range. Finally, offline simulation demonstrates that compared to nonlinear model predictive control strategies, the proposed method achieves comparable tracking performance while improving real-time responsiveness. The hardware-in-the-loop simulation results show that the proposed strategy enhances trajectory tracking accuracy and stability under vehicle cornering conditions, meeting the vehicle tracking accuracy requirements of real transport scenarios. In the event of sudden control function failures, stable docking can be achieved.更多还原