一种求解冗余机械臂逆运动学的自适应动态差分进化算法

针对冗余机械臂逆运动学难以直接求解且具有多解的问题,提出一种基于改进差分进化算法的逆运动学求解方法.以KUKA LBR iiwa七自由度机械臂为研究对象,验证该方法的有效性.针对基本差分进化算法易陷入局部搜索问题,引入自适应变异操作和随机变化交叉操作,使算法可根据迭代进程调整搜索强度.基于Denavit-Hartenberg方法建立七自由度机械臂运动学模型,改进逆运动学算法适应度函数与边界处理方法,最终求得最优逆运动学解.机械臂仿真实验以求解精度和稳定性为研究目标和评价指标,结果表明,在冗余机械臂逆运动学求解方面综合性能良好,具有较强的多工况拓展度和工程实用性.

An adaptive dynamic differential evolution algorithm for solving inverse kinematics of redundant manipulators

Solving of inverse kinematics of redundant manipulators is not straightforward and undesired multiple solutions may often exist during calculation. In this work, we present an approach based on improved differential evolution algorithm for simplifying and eventually solving this issue by utilizing KUKA LBR iiwa, a 7-degree-of-freedom manipulator. To overcome the deficiency that the conventional differential evolutionary algorithm tends to fall into local searching, the adaptive mutation and random change crossover operation have been leveraged. Therefore, the algorithm is able to ad-just searching intensity automatically during the iterative process. In parallel, we established the kinematics model based on Denavit Hartenberg method, and designed the fitness function for solving the inverse kinematics and boundary treatment. Particularly, by simultaneously considering the position and posture error of the end-effector the improved adaptive dynamic differential evolution algorithm was employed to search feasible solution within the joint space. Thereafter, optimal inverse kinematics solutions were able to be acquired readily. Numerical modeling and simulation were performed to verify the utility and performance of the approach. Of these, characteristics of precision and stability were evaluated and compared with existing methods. The presented approach is capable of solving inverse kinematics of redundant manipulators effectively with methodological scalability and engineering practicality.