基于改进NSGA-Ⅱ算法的机械臂多目标轨迹规划

Multi-objective trajectory planning of manipulator based on improved NSGA-Ⅱalgorithm

针对机械臂轨迹规划时需要满足高效率、低能耗、弱冲击等性能要求,提出一种改进的快速非支配排序遗传算法(NSGA-Ⅱ)。建立了端点运动参数可任意指定的7次B样条曲线方程,构造关节空间内光滑连续插值轨迹。以最短时间、最小能耗和最弱冲击为优化目标,以机械臂关节运动速度、加速度、脉动为约束惩罚,通过加入动态选择策略和模拟二进制(SBX)与正态分布混合交叉算子,建立了机械臂多目标轨迹规划的改进NSGA-Ⅱ算法,该方法具有搜索范围广、种群多样性质量高等优势。算例结果表明,所提出的改进NSGA-Ⅱ算法可以缩短迭代时间,得到更优的Pareto前沿。

In order to meet the performance requirements of high efficiency,low energy consumption and weak impact in the trajectory planning of manipulators,an improved fast nondominated sorting genetic algorithm(NSGA-Ⅱ)was proposed.Firstly,a seven-order B-spline curve equation with endpoint configuration parameters that can be specified is established to construct a smooth continuous interpolation trajectory in the joint space;Velocity,acceleration,and pulsation are the constraint penalties.By adding dynamic selection strategy and hybrid crossover operator of simulated binary(SBX)and normal distribution,an improved NSGA-II algorithm for multi-objective trajectory planning of manipulators is established.This method has a wide search range,high quality of population diversity and other advantages;Finally,the calculation example shows that the proposed improved NSGA-II algorithm can shorten the iteration time and get a better Pareto front.