Trajectory Generation of Hexapod Robots Based on Genetic Algorithm

In this paper, a novel trajectory generation algorithm for a rectangular hexapod robot in a complex environment based on a genetic algorithm (GA) when the robot is moving in a tripod gait. The main objective is to find the shortest trajectory between the initial and target locations while considering the hexapod’s physical constraints, such as the maximum step length, as well as collision avoidance with surrounding obstacles. Since rectangular hexapod robots perform better in forward and side motion compared to angular motion, it is preferable to move either horizontally or vertically. Therefore, finding the desired direction of motion (horizontal or vertical) is also considered during the trajectory planning procedure. First, a kinematic model of the hexapod is presented. Then, the planned trajectory is divided into a number of segments. Subsequently, GA is applied to find the optimum design variables that yield the shortest and collision-free trajectory. These variables include the number of steps, the length of each step, and the corresponding direction of motion (horizontal or vertical) during each segment. Finally, different simulation scenarios are conducted to demonstrate the effectiveness, convergence, and stability of the proposed algorithm.