謝海華 孫輝 龔文引



摘要:梯級水電站優化調度問題的準確、快速求解,是水利學科領域需解決的基本問題。針對該問題,提出了一種新的多策略人工蜂群算法。為更好地平衡算法的全局搜索與局部搜索能力,新算法在兩個具有代表性的解搜索策略基礎上,對其融合構成新的搜索策略,同時保留了原有的兩個解搜索策略。新算法的三個候選解搜索策略,增強了對各類優化問題求解的適應性。為驗證新算法的適應性及可行性,不僅在經典的基準測試函數中對其進行測試,并且將其應用于梯級水電站優化調度問題。實驗結果表明,新算法具有適應性強、收斂速度快等優點。
關鍵詞:梯級水電站;優化調度;人工蜂群算法;收斂速度;多策略
中圖分類號:TV11文獻標志碼:A
Abstract:To accurately and quickly solve the optimal operation problem of cascade hydro-power stations is a challenge in the field of water conservancy.A new multi-strategy artificial bee colony algorithm was proposed in this study.In order to better balance the global search and local search capabilities of the algorithms,two representative solution search strategies were used in this new algorithm,and they were combined to form a new search strategy while retaining the original two solution search strategies.Therefore,the new algorithm contained three candidate solution search strategies in the process of searching new solutions,which was convenient to strengthen the adaptability to various optimization problems.The adaptability and feasibility of the new algorithm were tested in the classic benchmark function and applied to the optimal operation of cascade hydro-power stations.Experimental results showed that the new algorithm had the advantages of stronger adaptability and faster convergence speed.
Key words:cascade hydro-power stations;optimal dispatch;artificial bee colony algorithm;rate of convergence;multi-strategy
梯級水電站的優化調度,是一個高維、多約束、非線性問題。解決該問題的核心是建立準確反應實際優化調度問題的模型和采用適當的求解方法[1]。目前,優化調度的數學模型相對成熟,但對于多約束條件下,快速及準確求解是該問題的難點所在。傳統方法和群智能方法是解決優化調度問題的主要方法[2-3],其中傳統方法包括:線性規劃(Linear Programming,LP)[4]、非線性規劃(Nonlinear Programming,NLP)[5]、動態規劃(Dynamic Programming,DP)[6]和大系統法(Large-scale System,LS)[7];群智能方法包括:人工蜂群(Artificial Bee Colony,ABC)算法[8]、蟻群算法(Ant Colony Optimization,ACO)[9]、遺傳算法(Genetic Algorithm,GA)[10]、粒子群算法(Particle Swarm Optimization,PSO)[11]等。傳統方法能有效解決單庫水電站調度問題,但對于梯級水電站的優化調度問題,不僅方法復雜且存在“維數災”、易陷入局部最優等缺點。相比傳統方法,群智能算法具有實現簡單、求解速度快等優點[12]。
2005年,土耳其學者karaboga為解決多變量函數問題,提出了ABC算法,其具有收斂速度快、參數少、魯棒性強等優點,并廣泛應用至各行業,如機器人路徑優化[13-14]和圖像處理[15]等。相比其他群智能算法,ABC算法對維度不敏感(問題維度的高低不影響ABC算法性能)是它的一個顯著特點。故本文采用ABC算法求解高維的梯級水庫優化調度問題。遵循著“算法沒有最好”的理念,ABC算法亦存在缺點,如全局搜索與局部搜索之間的平衡性較差。針對該問題,眾多的研究者提出了許多改進方案。較經典的有Zhu[16]等人提出的GABC、Gao[17]等人提出的MABC、Kiran[18]等人提出的ABCVSS,其中,Zhu等人針對ABC算法局部搜索能力弱的缺點,將全局最優引入到解搜索策略中;Gao等人針對ABC算法全局搜索與局部搜索能力平衡性差的缺點,通過引入控制參數,以達到目的;Kiran等人為豐富解搜索策略,構成了解搜索策略池,以適應多種類型優化問題。
目前的研究表明,更好地平衡ABC算法的全局搜索與局部搜索能力,可有效改善算法的總體性能。為此本文提出了一種新的多策略人工蜂群算法(Multi-strategy Artificial bee colony,MsABC)算法。
[6] YOUNG G K.Finding reservoir operating rules[J].Journal of the Hydraulics Division,1967,93(6):297-322.
[7] ARVANITIDIS N V,ROSING J.Optimal operation of multireservoir systems using a composite representation[J].IEEE Transactions on Power Apparatus & Systems,1970,PAS-89(2):327 -335.
[8] KARABOGA D.An idea based on honey bee swarm for numerical optimization,Technical Report-TR06[M].Erciyes University:Kayseri,Turkey,2005.
[9] DORIGO M,CARO G D.Ant colony optimization:a new meta-heuristic[J].Proceedings of the Congress on Evolutionary Computation,1999,2(4):1470-1477.
[10] GOLDBERG D E.Genetic algorithm in search[Z].Addison-Wesley,Reading,1989.
[11] KENNEDY J,EBERHART R.Particle swarm optimization[C].IEEE International Conference on Neural Networks,1995.Proceedings.IEEE,2002,4:1942-1948.
[12] 焦鈺,王建群,賈洋洋.基于狼群算法的水電站優化調度模型參數優選[J].南水北調與水利科技,2017,15(2):58-64.(JIAO Y,WANG J Q,JIA Y Y.Parameter analysis of wolf pack search algorithm applied to optimal operation of hydropower station[J].South-to-North Water Transfers and Water Science & Technology,2017,15(2):58-64.(in Chinese)) DOI:10.13476/j.cnki.nsbdqk 2017.02.009.
[13] 黎竹娟.人工蜂群算法在移動機器人路徑規劃中的應用[J].計算機仿真,2012,29(12):247-250.(LI Z J.Application of artificial bee colony algorithm in path planning of mobile robot[J].Computer Simulation,2012,29(12):247-250.(in Chinese))
[14] 譚玉新,楊維,徐子睿.面向煤礦井下局部復雜空間的機器人三維路徑規劃方法[J].煤炭學報,2017,42(6):1634-1642.(TAN Y X,YANG W,XU Z R.Three-dimensional path planning method for robot in underground local complex space[J].Journal of China Coal Society,2017,42(6):1634-1642.(in Chinese)) DOI:10.13225/j.cnki.jccs.2016.1047
[15] 肖永豪.蜂群算法及在圖像處理中的應用研究[D].廣州:華南理工大學,2011.(XIAO Y H.Study on artificial bee colony algorithm and its application in image processing[D].Guangzhou:South China University of Technology,2011.(in Chinese))
[16] ZHU G,KWONG S.Gbest-guided artificial bee colony algorithm for numerical function optimization[J].Applied Mathematics & Computation,2010,217(7):3166-3173.
[17] GAO W,LIU S.A modified artificial bee colony algorithm[J].Computers & Operations Research,2012,39(3):687-697.
[18] KIRAN M S,HAKLI H,GUNDUZ M,et al.Artificial bee colony algorithm with variable search strategy for continuous optimization[J].Information Sciences,2015,300:140-157.DOI:10.1016/j.ins.2014.12.043.
[19] GAO W,LIU S,HUANG L.A novel artificial bee colony algorithm based on modified search equation and orthogonal learning[J].IEEE Transactions on Cybernetics,2013,43(3):1011.
[20] CUI L,ZHANG K,LI G.et al.Modified Gbest-guided artificial bee colony algorithm with new probability model[J].Soft Compute.2018(22):2217-2243.DOI:10.1109/TSMCB.2012.2222373.
[21] 王坤.改進人工蜂群算法在梯級水庫群優化調度中的應用[D].南昌:南昌工程學院,2017.(WANG K.The application of artificial bee colony algorithm in optimal operation of cascade reservoirs is improved[D].Nanchang:Nanchang Institute of Technology,2017.(in Chinese))
[22] KARABOGA D,GORKEMLI B.A quick artificial bee colony (qABC) algorithm and its performance on optimization problems[J].Applied Soft Computing,2014,23(5):227-238.DOI:10.1016/j.asoc.2014.06.035.
[23] 成鵬飛,方國華,黃顯峰.基于改進人工蜂群算法的水電站水庫優化調度研究[J].中國農村水利水電,2013(4):109-112.(CHENG P F,FANG G H,HUANG X F.Optimal operation of hydropower station reservoir based on improved bee colony algorithm[J].China Rural Water and Hydropower,2013(4):109-112.(in Chinese))
[24] 李文莉,李郁俠,任平安.基于云變異人工蜂群算法的梯級水庫群優化調度[J].水力發電學報,2014,33(1):37-42.(LI W L,LI Y X,REN P A.Optimal operation of cascade reservoirs based on cloud variation-artificial bee colony algorithm[J].Journal of Hydroelectric Engineering,2014,33(1):37-42.(in Chinese))