基于改进麻雀搜索算法的机械臂路径规划  

Mechanical Arm in No Collision Path Planning Based on Improved Sparrow Search Algorithm

作　　者：沈孝龙 王吉芳[1] 郭子昇 SHEN Xiao-long;WANG Ji-fang;GUO Zi-sheng(School of Mechanical Electrical Engineering,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学机电工程学院,北京100192

出　　处：《组合机床与自动化加工技术》2023年第1期15-19,24,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：北京市自然科学基金:海淀原始创新联合基金项目(19L2018);国家自然科学基金(52005045)。

摘　　要：针对串联机械臂在三维空间的无碰运动规划问题,提出一种基于改进麻雀搜索的无碰运动规划方法,对比蚁群算法在此问题的应用,改进算法效果出色。首先,利用包围盒法建立机械臂与障碍物模型,使机械臂与障碍物的碰撞检测抽象为空间内判断直线与平面的相交问题;其次,把无碰运动规划问题转化为函数优化问题,利用新颖的麻雀搜索算法,融合柯西变异与反向学习机制对其控制参数改进;最后,构建三维环境模型在MATLAB中仿真验证算法的可行性。结果表明,所提算法可使机械臂在三维空间中避碰预设的障碍物且基于多项式插值法规划的轨迹连续平滑,满足各关节角度与速度的光滑连续。Aiming at the problem of touched motion planning of series robot in three-dimensional space,a non-touch motion planning method based on improved sparrow search is proposed,and the application of ant colony algorithm in this problem is excellent.First,the robot arm and obstacle model is established by the enveloping method,so that the collision detection of the robot arm and the obstacle is to determine the intersection of the straight line and plane in the space;Secondly,the problem of non-touching exercise planning is transformed into function optimization problems,using the novel sparrow search algorithm,the fusion of Kexi variation and reverse learning mechanism improves its control parameters;The results show that the algorithm can cause the robotic arm to avoid the preset obstacle in three-dimensional space and the trajectory of the polynomial interpolation method is continuous and smooth,meet the perspective of each joint and the angular velocity is smooth and continuous.

关 键 词：串联机械臂 麻雀搜索 碰撞检测 避障路径规划 

分 类 号：TH165[机械工程—机械制造及自动化] TG502[金属学及工艺—金属切削加工及机床]