【摘要】 在不稳定的环境中,最短或最优的路径规划对移动机器人的有效运行至关重要.文章扩展了一种用于移动机器人实时路径规划的神经网络方法.本文使用概率路线图(Probabilistic Roadmap, PRM)来解决移动机器人在不稳定环境中的实时无碰撞运动规划问题.神经网络拓扑中的每个神经元都具有局部连通性,并且分流方程是其神经动力学特征.因此,计算复杂度与神经网络的大小成线性关系.实时机器人的运动取决于神经网络的动态环境,其中不需要任何先前的动态环境知识甚至学习过程.仿真结果证明了该方法的有效性和高效性.更多还原

【Abstract】 In unsteady environment, the shortest or optimal path planning is critical to the effective operation of mobile robots. This paper extends a neural network method for real time path planning of mobile robots. This paper uses the Probabilistic Roadmap(PRM) to solve the real time collisionless motion planning problem of mobile robots in an unsteady environment. Each neuron in a neural network topology has local connectivity, and the shunting equation is its neurodynamic feature. Therefore, computation complexity is linear with the size of the neural network. The motion of real time robot depends on the dynamic environment of neural network, in which any prior knowledge of dynamic environment and even learning process are not needed. Simulation results show that the effectiveness and efficiency of the method.更多还原