PATH PLANNING METHOD FOR MOVING ROBOT OF IRREGULAR SHAPE

A path planning method for a moving robot of an irregular shape, comprising the following steps: evenly dividing a 2D workspace of the robot into a topological network T of workspaces, and conducting iterative computations of biological neural dynamics to construct a biological neural network S of the workspaces; forming a kernel matrix according to shape features of the moving robot; conducting convolution computation on the kernel matrix and the biological neural network of the workspaces to obtain a moving state matrix and a rotating state matrix of the robot at network nodes; performing iterative search step by step by using the initial position and direction of the robot in the workspace as initial conditions, and constructing a node tree Tree from an initial state to a target state; and performing reverse search on the node tree Tree to obtain a planned path of the robot. The robot may be of any irregular shape, the three degrees of freedom movement of the robot in a planar space has been considered, an