Li Ping,Peng Siyang,Che Linxian,et al.Artificial Bee Colony and Newton Algorithm for Forward Position Solution of Parallel Mechanism[J].Journal of Mechanical Transmission,2019,43(04):44-50.
Li Ping,Peng Siyang,Che Linxian,et al.Artificial Bee Colony and Newton Algorithm for Forward Position Solution of Parallel Mechanism[J].Journal of Mechanical Transmission,2019,43(04):44-50. DOI: 10.16578/j.issn.1004.2539.2019.04.009.
Artificial Bee Colony and Newton Algorithm for Forward Position Solution of Parallel Mechanism
将智能优化算法与数值迭代方法有机组合,构造一种并联机构位置正解求解的通用算法——混合人工蜂群和Newton迭代(Hybrid artificial bee colony and Newton iteration,HABC-Newton)算法。将差分进化(Differential evolution,DE)算法融入人工蜂群(Artificial bee colony,ABC)算法,形成一种能快速收敛到问题近优解的混合人工蜂群(Hybrid ABC,HABC)算法,再以该近优解为初值,应用Newton-Шамарский迭代法求出高精度位置正解。以4自由度4-SPS-CU并联机构正运动学分析为例,阐述基于HABC-Newton算法的并联机构正运动学分析方法。为了验证算法的有效性和普适性,给出4-SPS-CU、3-RRR两种耦合并联机构位置正解的数值算例。结果表明,HABC-Newton算法能以较少计算开销求得并联机构的全部高精度位置正解。还比较了HABC-Newton、ABC、DE和粒子群算法求并联机构位置正解的性能,数值实验显示,HABC-Newton算法的精度、稳健性和计算效率高于对比算法。
Abstract
By an organic combination of the intelligent optimization algorithm and the numerical iteration method, a general algorithm called hybrid artificial bee colony and Newton iteration (HABC-Newton) algorithm for solving the forward positions of parallel mechanism is constructed. The differential evolution (DE) algorithm is incorporated into the artificial bee colony (ABC) algorithm to form a hybrid ABC (HABC) algorithm which can converge quickly to the near optimal solution of the problem. Then the optimal solution is used as initial value and Newton-Шамарский iteration method is employed to find high precision solutions. Taking 4-SPS-CU parallel mechanism kinematics analysis as an example, the forward kinematics analysis method of parallel mechanism based on HABC-Newton algorithm is stated. In order to verify the effectiveness and universality of the algorithm, two numerical examples of forward kinematics such as 4-SPS-CU and 3-RRR coupled parallel mechanisms are given. The results show that the HABC-Newton algorithm can obtain all the high accurate solutions of the parallel mechanism with less computational cost. Furthermore, comparative tests to solve these examples are carried out with HABC-Newton, ABC, DE and particle swarm optimization algorithms, and the numerical experiments indicate that HABC-Newton algorithm has better performance than compared algorithms in terms of the accuracy, robustness and computational efficiency.
GALLARDO A J. A simple method to solve forward displacement analysis of the parallel manipulator[J]. Robotics and Computer Integrated Manufacturing, 2013, 30(1):55-61.
KARABOGA D, BASTURK B. A powerful and efficient algorithm for numerical function optimization: artificial bee colony (ABC) algorithm[J]. Journal of Global Optimization, 2007, 39:459-471.