论文部分内容阅读
针对基本遗传算法在解决实际最优化问题时可能存在的收敛速度慢、易于陷入局部最优解等问题,提出从适应度排序标定和混沌算子嵌入两方面进行算法性能的改进。前者避免了传统适应度值计算方法中较小目标函数值导致的具有过大适应度值的个体对种群进化方向的影响,使得种群始终能够保持恒定的进化压差促进最优解的搜索;后者则增强了遗传算法局部搜索的能力,从而提高了近似最优解向全局最优解转化的可能性。在此基础上,建立了一种基于混沌遗传算法的光纤布拉格光栅(FBG)轴向非均匀应变分布重构方法,仿真算例表明,混沌遗传算法有效改善了非均匀应变分布重构算法的收敛性能,提高了重构的精度。讨论了算法中相关参数的设置对非均匀应变分布重构精度的影响。
Aiming at the problems that the basic genetic algorithm may solve the practical optimization problems, such as slow convergence rate and easy to fall into the local optimal solution, the improvement of the algorithm performance is proposed from the aspects of fitness ranking and chaos operator embedding. The former avoids the influence of the individuals with excessively large fitness value on the population evolution direction caused by the smaller objective function value in the traditional fitness value calculation method so that the population can always maintain a constant evolutionary pressure difference and promote the search of the optimal solution. Enhances the ability of genetic algorithm to search locally, so as to improve the possibility of converting the approximate optimal solution to the global optimal solution. On this basis, a method of reconstruction of axial non-uniform strain distribution of fiber Bragg grating (FBG) is established based on chaos genetic algorithm. Simulation results show that chaos genetic algorithm can effectively improve the convergence of non-uniform strain distribution reconstruction algorithm Performance, improve the accuracy of reconstruction. The influence of the setting of relevant parameters in the algorithm on the reconstruction accuracy of the non-uniform strain distribution is discussed.