神经网络PID控制的机器人视觉反馈跟踪

摘要

探讨针对视觉空间的非完整移动机器人的跟踪控制问题.在不校准摄像机视觉参数的情况下,利用视觉反馈得到的信息,设计出非完整移动机器人轨迹跟踪的神经网络控制器.将BP网络与PID控制相结合,避免复杂的公式推导,解决参数不校准下的控制问题,并很好的实现跟踪.仿真结果证明了文中方法的有效性.%This paper discusses the visual servoing tracking problem of nonholonomic mobile robots. The neutralnetwork control approach is used to design a tracking controller for nonholonomic mobile robots with uncalibratedvisual parameters. Combining BP neutral network with PED controller is used to manipulate mobile robots firstly. Thecomplex deduce of common tracking controllers is simplified and tracking control problem with uncalibrated virualparameters is solved. The simulation results show that the method is effective.