基于T-S模糊神经网络的智能轮椅避障控制策略

Avoidance Control Strategy for Intelligent Wheelchair Based on T-S Fuzzy Neural Network

针对传统智能轮椅避障策略的路径规划效率差、功耗高等缺点,提出一种基于模糊神经网络的环境深度分区控制策略;利用红外、超声波和激光传感器的测量信息将待识别环境分为3个不同的深度区间,同时,利用T-S模糊神经网路算法融合异质传感器的测量信息,然后设计模糊控制规则,实现智能轮椅避障动作;最后建立智能轮椅的运动学模型和测量模型,并进行Simulink仿真测试;经仿真可知,该方法控制可靠,可快速无碰撞地通过障碍区,并能减少功耗,提高续航能力。

A type of depth partition strategy based on fuzzy neural network （FNN） is designed for avoidance strategy of Intelligent Wheelchair because that the traditional one has poor efficiency and high power consumption.Environment is identified in three different depth intervals by value from sensors of infrared,ultrasonic and laser,meanwhile,merging information from different sensors by T-S FNN and designing fuzzy control rules to achieve avoidance.Finally,modeling a kinematic and measurement model of intelligent wheelchair and testing by Simulink simulation.The simulation shows that the method is reliable,make Wheelchair fast through the obstacles zone without collision,can reduce power consumption and improve endurance.