摘要
针对室外移动机器人的行驶特点 ,将车体模型划分为前轮转向模型、速度模型和位姿模型三个部分 .提出用模糊集合与神经网络相结合来建立车体前轮转向模型的方法 .首先将对前轮转向特性影响较大的行车速度模糊化 ,然后利用神经网络建立各模糊速度下的前轮转向模型 ,最后由逆模糊化过程求得模型的实际输出 .实验结果表明 。
A vehicle model of outdoor mobile robot which consists of a front wheel steering model,a speed model and a posture model is discussed based on the analysis of the driving characteristics of the robot. A method of establishing the front wheel steering model using fuzzy sets and neural networks is proposed. First the driving speed which influences the steering performance strongly is fuzzificated. Then the steering models for each fuzzy speed are constructed with neural networks. Finally the actual output of the model is calculated by a defuzzification process. It is shown by the results that the method can describe front wheel steering features exactly. The most remarkable superiority is its robustness and simplicity.
出处
《自动化学报》
EI
CSCD
北大核心
2000年第2期154-161,共8页
Acta Automatica Sinica
基金
国家 8 63计划资助项目
关键词
移动机器人
前轮转向模型
神经网络
智能机器人
Outdoor mobile robot, front wheel steering model, fuzzy sets, neural networks.
