基于LM和SA的混合优化算法  

A Hybrid Optimization Algorithm Based on LM Algorithm and Simulated Annealing Algorithm

作　　者：武菊[1] 任鹏[2] 

机构地区：[1]内江师范学院数学与信息科学学院,四川内江641100 [2]西南技术物理研究所,四川成都640041

出　　处：《内江师范学院学报》2010年第8期32-34,共3页Journal of Neijiang Normal University

摘　　要：针对LM算法及模拟退火算法各自存在的不足,提出一种交替使用二者的混合优化算法（SALM算法）.该算法先通过模拟退火算法粗调得到一组全局最优近似解,再以该近似解为初值,交替使用模拟退火算法和LM算法,直至发现满意的最优解.实验结果表明,新算法不但具有模拟退火算法的全局收敛特性而且保留了LM算法的局部收敛速度,克服了单纯模拟退火算法所产生的随机性和概率性问题,也弱化LM算法对初始条件的依赖程度,保证了求解的速度和精度。In an attempt to rid the LM algorithm and simulated annealing algorithm of their own deficiencies,a hybrid optimization algorithm（SALM）of using alternatively the simulated annealing algorithm and LM algorithm has been put forth.First,agroup of globally optimal approximate solutions were obtained by way of the simulated annealing algorithm.Then take these approximate solutions as the initial values,the SA and LM algorithms were alternatively put in use until at last the satisfactory optimal solutions were found.Our findings indicate that the algorithm,in possession of both the global convergence feature of simulated annealing algorithm and the local convergence velocity of LM algorithm,successfully rids itself of the problem of randomness and probability produced by merely use of the simulated annealing algorithm and at the same time weakens its dependence on the initial conditions in the use of LM algorithm,and thus guarantees the speed and precision of the solving process.

关 键 词：LM算法 混合优化算法 模拟退火算法 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]