研究了一类直流电机驱动的轮式移动机器人路径跟随控制问题。由于直流电机的结构参数时变且难于测量,这里采用RBF神经网络在线对其辨识,并将辨识结果加入到所制定的反馈线性化路径跟随控制方案中。该方案同时兼顾了车体运动学模型、电机驱动模型的动态特性,使得研究结果与实际相符合。基于李亚普诺夫稳定性的证明和仿真结果证明了该方法的正确性和有效性。 A kind of path following control problem of wheeled mobile robot driven by DC motor is studied. Because of the time-varying and difficult-to-measure structural parameters of DC motor, RBF neural network is used to identify it online and the identification results are added to the feedback linearized path following control scheme. The scheme takes into account the dynamic characteristics of the body kinematics model and the motor-driven model at the same time, which makes the research results consistent with the actual situation. The proof and simulation results based on Lyapunov stability show that the method is correct and effective.