基于人工鱼群与几何混合优化的圆度误差评定算法

摘要

In order to evaluate circularity error accurately and efficiently,a hybrid improved artificial fish swarm and geometric algorithm (HI-AFSA) is presented.Combining optimization algorithm with geometric method,the method first obtains the reference centre of mimmum zone circle through improved AFSA algorithm,then uses the geometric method to get the unique globe optimal solution by reference centre.A mutation mechanism is applied in classic AFSA algorithm.The convergence speed and precision of the artificial fish swarm algorithm are increased through improving the behavior of artificial fish.Finally,the experiment results show that compared with the genetic algorithm and the classic artificial fish swarm algorithm,the proposed hybrid optimization algorithm is feasible and efficient for the evaluating circularity error.%为快速、精确地评定圆度误差,提出了一种改进的人工鱼群与几何混合优化算法求解圆度误差评定问题.该方法将优化算法与几何方法相结合,首先利用改进人工鱼群算法获得最小包容区域的参考圆心,再根据参考圆心结合几何方法获得全局唯一的最小包容区域圆心,实现圆度误差的精确评定.通过在经典人工鱼群算法中引入变异机制,对算法中的聚群、觅食和追尾行为加以改进,有效提高了鱼群算法的优化效率和稳定性.实验及仿真结果表明,改进的人工鱼群混合优化算法与遗传算法、经典鱼群等算法相比,在进行圆度误差评定时,收敛速度快,且计算结果稳定性好.