改进型生物激励神经网络的路径规划方法

Path Planning Method Based on Improved Biologically Inspired Neural Network

针对生物激励神经网络在一些情况下存在的路径非最优问题，通过分析其产生机理，做出了相应的改进：对边界附近和障碍物之间的路径点引入了假想的非障碍物相邻点，增大了激励输入，使得这些路径点的活性值增大，解决了路径错判问题；同时，在下一个位置的决策中加入了转角最小因素，减少了路径的转折次数。仿真结果表明，改进后的生物激励神经网络方法适用于实时动态环境下的移动机器人路径规划，且全面地提升了路径质量。

Paths generated by biologically inspired neural network approach m some cases are not laeaL Through analyzing its generating mechanism, some improvements are made： imaginary non obstacle neighboring points are introduced for both trajectory points near the border and between obstacles, which leads to bigger excitatory inputs and higher activity values of these trajectory points, thus resolving the erroneous trajectory decision problem. Meanwhile, minimum turn angle factor is taken into consideration in the decision of next location, which results in less turns of generated path. Simulation results indicate that improved biologically inspired neural network can be used for real-time robot path planning under arbitrarily varying environments and improves the trajectory quality.