基于智能算法的空间爬行机器人运动控制方法研究OA

The virtual simulation environment for the lunar surface roaming robot in the lunar exploration pro-gram is an important application background for the research of quadruped robots.For the inverse kinematic problem of the quadruped robot,the analytical method is used to analyze and derive the trajectory curves of the foot end based on the pendulum line,and the Walk gait and the Trot gait are designed.The results of the joint simulation of MATLAB and CoppeliaSim show that both the two kinds of gaits can realize the smooth motion.Aiming at the path planning problem,an improved fusion algorithm is proposed,the genetic algorithm(GA)is used to obtain the initial high-quality solution,which is converted to the initial pheromone value of the im-proved ant colony optimization(ACO)algorithm,to guide the search direction and improve the convergence effi-ciency.MATLAB simulation experiments show that compared with the improved GA,improved ACO algorithm and typical deep reinforcement learning methods,the fusion algorithm shows good performance advantages in path quality,stability and engineering deployability.