系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2014年
12期
3268~3273
,共null页
差分进化算法 马尔可夫链蒙特卡罗方法 参数优选 适应度 多目标differential evolutionadaptive metropolis算法
差分進化算法 馬爾可伕鏈矇特卡囉方法 參數優選 適應度 多目標differential evolutionadaptive metropolis算法
차분진화산법 마이가부련몽특잡라방법 삼수우선 괄응도 다목표differential evolutionadaptive metropolis산법
differential evolution algorithm; Markov chain Monte Carlo method; parameter optimization;fitness; multi-objective differential evolution adaptivc metropolis
在差分进化算法的基础上,受马尔可夫链蒙特卡罗方法的启发,建立了differential evolution adaptive metropolis（DREAM）算法.DREAM算法融合了马尔可夫链蒙特卡罗方法和差分进化算法的优势,较好地解决了马尔可夫链蒙特卡罗方法中搜索步长的恰当取值以及搜索方向的准确定位问题,并能有效解决差分进化算法的群体多样性和收敛速度问题.在DREAM算法基础上,引入多目标优化思想,提出了一种基于改进适应度分配策略和外部存档方案的多目标DREAM算法,并应用于岷江流域CMD-3PAR降雨-径流模型参数优选研究.结果表明：多目标DREAM算法能够找到一组范围宽广、分布均匀且数量充足的Pareto最优解供决策者评价优选.
在差分進化算法的基礎上,受馬爾可伕鏈矇特卡囉方法的啟髮,建立瞭differential evolution adaptive metropolis（DREAM）算法.DREAM算法融閤瞭馬爾可伕鏈矇特卡囉方法和差分進化算法的優勢,較好地解決瞭馬爾可伕鏈矇特卡囉方法中搜索步長的恰噹取值以及搜索方嚮的準確定位問題,併能有效解決差分進化算法的群體多樣性和收斂速度問題.在DREAM算法基礎上,引入多目標優化思想,提齣瞭一種基于改進適應度分配策略和外部存檔方案的多目標DREAM算法,併應用于岷江流域CMD-3PAR降雨-徑流模型參數優選研究.結果錶明：多目標DREAM算法能夠找到一組範圍寬廣、分佈均勻且數量充足的Pareto最優解供決策者評價優選.
재차분진화산법적기출상,수마이가부련몽특잡라방법적계발,건립료differential evolution adaptive metropolis（DREAM）산법.DREAM산법융합료마이가부련몽특잡라방법화차분진화산법적우세,교호지해결료마이가부련몽특잡라방법중수색보장적흡당취치이급수색방향적준학정위문제,병능유효해결차분진화산법적군체다양성화수렴속도문제.재DREAM산법기출상,인입다목표우화사상,제출료일충기우개진괄응도분배책략화외부존당방안적다목표DREAM산법,병응용우민강류역CMD-3PAR강우-경류모형삼수우선연구.결과표명：다목표DREAM산법능구조도일조범위관엄、분포균균차수량충족적Pareto최우해공결책자평개우선.
A novel differential evolution adaptive metropolis algorithm （DREAM） is presented, which combines the advantages of differential evolution algorithm and Markov chain Monte Carlo （MCMC） sam- pler. DREAM solves an important problem in MCMC, namely that of choosing an appropriate scale and orientation for the jumping distribution. Meanwhile, it can make a good trade-off between population diversity and convergence for differential evolution algorithm. Moreover, multi-objective DREAM is pro- posed based on the modified fitness assignment and external archive strategy, which is applied in parameter optimizaVion of CMD-3PAR hydrologic model in the Min River Basin. The results show that DREAM is capable to infer the posterior distribution of model parameters, and multi-objective differential evolution adaptive metropolis （MODREAM） is capable to generate a lot of non-dominated solutions with wide and uniform distribution for decision-makers.
