基于小世界算法的机械手时间最优运动规划

Time-optimal Motion Planning of Manipulator Based on Small World Algorithm

针对机械手运动规划通常存在建模复杂、参数多和优化效率低的不足,提出了基于小世界算法的机械手时间最优运动规划。首先运用末端分离法进行机械手工作姿态的简化模型建立;然后结合摆线运动方程,并以工作姿态和关节运动学为约束,建立时间最优的运动规划模型;最后引入基于精英集聚效应的小世界优化算法来提高模型求解精度。仿真测试结果表明,与禁忌遗传算法、基本小世界算法相比,精英集聚效应小世界算法使得机械手规划模型的运动时间和总均方差分别平均降低23.3%和19.7%,且表现出较好的收敛性和稳定性,进而验证了时间最优规划模型的有效性。

In order to solve problems of complex modeling,multiple parameters and low optimization efficiency in manipulator motion planning,a time-optimal motion planning based on small world algorithm was proposed.Firstly,the end separation method was used to establish a simplified model of the robot's working posture.Then,combined with the cycloidal motion equation and constrained by the working attitude and joint kinematics,a time-optimal motion planning model was established.Finally,an elite agglomeration-based small world algorithm(EASWA)was introduced to improve the accuracy of the model solution.The following numerical test results were got from the research.Compared with the tabu genetic algorithm(TGA)and the simple small world algorithm(SSWA),both the exercise time and total mean square error of the EASWA optimized manipulator planning model were decreased by an average of 23.3%and 19.7%,respectively.The optimization performance shows good convergence and stability,which further verifies the effectiveness of the proposed planning model.