基于FNN的轮式机器人路径跟踪研究  

Research on Wheeled Robot Path Tracking Based on Fuzzy Neural Network

作　　者：杨硕[1] 张伯泉[1] 曾安[1] 贾茂盈[2] 

机构地区：[1]广东工业大学计算机学院,广东广州510006 [2]山东理工大学计算机科学与技术学院,山东淄博255049

出　　处：《测控技术》2015年第5期65-68,共4页Measurement & Control Technology

基　　金：国家自然科学基金项目(61300107)

摘　　要：针对目前轮式机器人在路径跟踪时容易出现的偏离期望路径甚至打滑、侧翻失去控制等问题,对轮式机器人结构及其路径跟踪特点进行了分析,构建了轮式机器人运动学模型,设计了一种基于模糊神经网络(FNN)的行进路线和行驶速度分级控制的路径跟踪方法。第一级中模糊神经网络利用机器人位姿信息确定行进路线即转弯半径,第二级根据前方路径情况和转弯半径调节机器人行驶的角速度和线速度。仿真实验表明,所设计的模糊神经网络能够对所期望的路径进行快速准确地拟合,且鲁棒性强;轮式机器人路径跟踪过程稳定,不会出现失控现象。Wheeled robot often deviates from the desired path during path tracking, and even slips or rolls over. The structure of wheeled robot and its path tracking features are analyzed, and the kinematical model of wheeled mobile robot is constructed, a path tracking method with two-stage control based on fuzzy-neural network （FNN） is proposed. In the first stage, a FNN controller determines robot＇ s turning radius by processing robot＇ s pose information. In the second stage, the controller adjusts angular and linear velocities by taking advantages of the turning radius and the condition of the path ahead. The experiments show that the controlled robot can track the planned path accurately and robustly when it runs at high speed, the process of path tracking is stable and no slipping and rolling occur.

关 键 词：轮式机器人 模糊神经网络 路径跟踪 分级控制 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]