基于RBF和模糊预测的AGV轨迹跟踪控制

AGV Tracking Control Based on RBF and Fuzzy Predictive Control

针对AGV系统存在初始位姿误差,而且给定轨迹又不连续时,应用传统的轨迹跟踪控制方法,就会使其初始速度产生较大跳变的问题,基于径向基函数(RBF)神经网络的非线性动态系统在线建模,将模糊控制技术与预测控制技术相结合,提出基于反演(Backstepping)方法的速度控制器和基于RBF的模糊预测转矩控制器,实施AGV路径跟随和轨迹跟踪控制.仿真和实验结果表明,设计的速度控制器和转矩控制器使AGV系统不仅有较好的动态性能,而且具有较强的鲁棒性.

Initial position error exists in AGV system.When the reference trajectory is not a larger initial velocity jump continuous for AGV system,the adoption of traditional fuzzy trajectory tracking control will lead to a larger initial velocity jump.Based on the radial basis function （RBF）neural network modeling of nonlinear dynamic system online,the fuzzy control technology and the predictive control technologies are combined to establish a backstepping speed controller and an RBF torque fuzzy predictive controller,which implement the control of AGV path and trajectory tracking.It is shown through simulation and experimental results that the proposed speed controller and torque controller not only have better dynamic performances,but also have strong robust-ness.