用户名: 密码: 验证码:
遗传算法及其在模糊辨识中的应用研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
遗传算法是一种模拟自然界生物进化的搜索算法,由于它的简单易行、鲁棒性强、尤其是其不需要专门的领域知识而仅用适应度函数作评价来指导搜索过程,从而使它的应用极为广泛。本文在参考大量国内外文献的基础上,针对基本遗传算法的缺点,提出了两种改进的遗传算法;研究了遗传算法在非线性方程组求解中的应用,并将遗传算法应用于模糊辨识中。取得的主要成果如下:
    针对基本遗传算法收敛速度慢、解的分辨率过低等不足,提出了改进变焦遗传算法,即在保持串长不变的条件下,不断把经多次进化的信息从个体串中存放到解码公式上,所空出的基因起着类似“内存条”的作用,为提高解的精度随时纳入新的基因。同时为了防止早熟现象,交叉算子中的交叉位置按非等概率选取的方法进行;在纳入新的基因时,加入与最优个体群等量互补的二进制码串,解决了复制操作导致基因缺失的问题。
    提出一种基于禁忌搜索和混沌优化的混合遗传算法,该算法利用禁忌搜索的记忆功能和混沌优化所具有的遍历性、规律性和随机性,对经过一次遗传操作的种群进行混沌寻优,引导种群快速进化。这种算法容易跳出局部最优解,搜索效率高,而且结构简单,使用方便。
    针对迭代法、最优法、MATLAB最优化工具箱求解非线性方程组中存在求解精度不高及初始矢量难选等问题,将方程组求解问题转化为遗传算法函数优化问题,建立了非线性方程组通用的遗传算法解法,并将其用于汽车滑行试验数据处理中。
    将遗传算法用于非线性系统的模糊辨识中,基于输入空间的模糊划分,采用具有自适应性的广义高斯函数为隶属函数,用改进的遗传算法来优化其形状,再根据前提参数采用递推最小二乘算法辨识结论参数,该算法综合了遗传算法的全局搜索能力和递推最小二乘算法的快速局部搜索能力。最后通过仿真实例验证了该方法的有效性与实用性。
Genetic algorithm is a searching algorithm that simulates biology evolvement in nature. Because of its briefness, strong robustness, especially it need not expert domain knowledge but fitness function as the value guide searching process, its applications are very abroad. Based on various domestic and overseas references, aiming at simple genetic algorithm’s shortcomings, the dissertation brings forward two improved genetic algorithms, studies genetic algorithm’s application on acquiring nonlinear equations’ solutions, and applies genetic algorithm to fuzzy identification. The main results are as follows:
    Aiming at simple genetic algorithm’s slow convergence and slow resolving power and so on, extended zooming genetic algorithms are brought forward, i.e. it keeps invariable string length, unceasingly leaves evolving information in decoding formula from individual string, and spare genes have function such as memory bank which is supplied with new genes for improving solution precision. Simultaneously, for preventing prematurity, crossover positions in crossover operator choose according to not equal probability. When bringing into new genes, reverse bit binary strings those have the same number as the optimal individuals are supplied, which settles gene absence problem reproduction results in.
    Using memory function of tabu search and searching, stochastic and disciplinary characteristics chaos optimization has, a hybrid genetic algorithm based on tabu search and chaos optimization is implemented to the population which passed through the genetic algorithm one time, which can induce the evolution of the population rapidly. This algorithm easily escapes from local optimal solution, have high searching efficiency, simple structure, convenient use.
    Aiming at iteration, optimization and MATLAB optimization toolbox having low precision and difficulty to choose initial vector on acquiring nonlinear equations’ solutions, equations’ solution problem is translated into genetic algorithm optimization problem. Nonlinear equations’ usual genetic
    
    
    algorithm solution has been set up, and it has been used on automobile coasting test data processing.
    Genetic algorithm is used on fuzzy identification on nonlinear systems, based on fuzzy partition of input space, it adopts adaptive extended gauss function as member function, uses improved genetic algorithm to optimize its figure, uses the recursive least square method to identify the conclusion parameters of the fuzzy model. This method synthetizes genetic algorithm global searching ability and fast local searching ability of the recursive least square. Finally the effectiveness and practicability of this method is demonstrated by the simulation results example.
引文
l J.H.Holland.Adaptation in Natural and Artificial System.Michigan:The Univ.of
     Michigan Press,1975
    2 J.H.Holland.Genetic Algorithms.Scientific American.1992:44—50
    3 D.E.Goldberg.Genetic Algorithms in Search,Optimization,and Machine Learning. MA:Addison Wesley,1989
    4 李大卫.可重复自然数编码遗传算法的最优群体规模.鞍山钢铁学院学报,2000, 23(6):419—423
    5 Ignacio Rojas,Jesús González,Héctor Pomares et.a1.Statistical Analysis of The Main Parameters Involved in The Design of A Genetic Algorithm.IEEE Trans.on Systems,Man,and Cybernetic-Part C:Applications and Reviews,2002,32(1):
    31—37
    6 N.N.Schraudolp.Dynamic Parameter Encoding for Genetic Algorithms.Machine Learning,l992,9(1):9—2l
    7 Young-Doo Kwon,Soon—Bum Kwon,Seung-Bo Jin et.a1.Convergence Enhanced Genetic Algorithm with Successive Zooming Method for Solving Continuous Optimization Problems.Computers and Structures,2003,8l(17):l715—1725
    8 L.Davis.Adapting Operator Probabilities in GA.Proc.3rd.Conf.Genetic Algorithms.San Mateco.1989:6l一69
    M.Srinivas.L.M.Patnaik.Adaptive Probabilities of Crossover and Mutation in Genetic Algorithm.IEEE Trans.on Systems,Man and Cybemetics,1994,
     24(4):656—667
    10王蕾,沈庭芝.招扬.一种改进的白适应遗传算法.系统工程与电子技术,2002, 24(5):75--78
    11 A.Blanco,M.Delgado,M.C.Pegalajar.A Real—coded Genetic Algorithm for Training Recurrent Neural Networks.Neural Networks,200l,14(1):93—1 05
    12 A.H.Wright.Genetic Algorithms for Real Parameter Optimization.Foundation of Genetic Algorithms,San Mateo,CA:Morgan Kaufmann,1991:205—218
    l3张晓缋,方浩,戴冠中.遗传算法的编码机制研究.信息与控制,1997,26(2):
    
    
    
     l34一139
    l4 Z.Michalewicz.A Modified Genetic Algorithm for Optimal Control Problems. Computers Math.Application,1992,23(12):83—94
    15 W.Atmar.Notes on Simulation of Evolmion.IEEE Trans.on Neural Network,1994, 5(1):130-143
    16田盛丰.人工智能原理与应用.北京:北京理工大学出版社,1993
    17任立红,丁永生,邵世煌。采用DNA遗传算法优化设计的TS模糊控制系统.控 制与决策,2001,16(1):16—24
    18郑朝晖,张焱,裘聿皇.一种基于复数编码的遗传算法.控制理论与应用,2003, 20(1):97一lOO
    19恽为民,席裕庚.遗传算法的运行机理分析.控制理论与应用,1996,
     l3(3):297-304
    20 K.A.De Jong.An Analysis of The Behavior of A Class of Genetic Adaptive Systems.[Ph.D Dissertation,University of Michigan].1975
    2 l A.Brindle.Genetic Algorithm for Function Optimization.[Doctoral Dissertation, Univ.ofAlberta].198l
    22 T.Back.The Interaction of Mutation Rate.Selection and Self-Adaptation within A GA.In Parallel Problem Solving from Nature,2,Amsterdan:North Holland,l992:
    84-94
    23张讲社,徐宗本,梁怡.整体退火遗传算法及其收敛充要条件.中国科学(E 辑), 1997,27(2):154—164
    24 徐宗本,高勇.遗传算法过早收敛现象的特征分析及预防.中国科学(E 辑),1996, 26(4):364-375
    25 W.M.Spears.K.A.De Jong.An Analysis of Multi—Point Crossover Foundations of GA,San Mateo.CA:Morgan Kaufmann.1991:30l一315
    26陈小平,于盛林.实数遗传算法交叉策略的改进.电子学报,2003,3l(1):71-74
    27 Z.Michalewiz.GA and Optimal Control Problems.Proc.29th.IEEE Conf. Decision and Control,Honolu,I990:1664—1666
    28 x.Qi.F.Palmieri.Theoretical Analysis of Evolutionary Algorithms with An Infinite Population Size in Continuous Space Part I:Basic Properties of Selection and Mutation.IEEE Trans.on Neural Networks,1994,5(1):102一119
    
    
    
    29草珂,刘贵思.杂交位置非等概率选取的遗传算法,信息与控制,1997,
     26(1):53—60.
    30任庆生,曾进,戚飞虎.交叉算子的极限一致性.计算机学报,2002,25(12):
     1405.1410
    3l黄晓峰,潘立登,陈标华,等.实数编码遗传算法中交叉操作的效率分析.控制
     与决策,1998,(13):496—499
    32巩敦卫,孙晓燕,郭西进.一种新的优胜劣汰遗传算法.控制与决策,2002,
     17(6):908—91l
    33袁惠梅.具有自适应交换率和变异率的遗传算法。首都师范大学学报(自然科学
     版),2000,2l(3):14—19
    34张明辉,王尚锦,具有自适应交叉算子的遗传算法及其应用.机械工程学报,
     2002,38(1):5l-54
    35 Johnjoe McFadden,Jim A1-Khalili.A Quantum Mechanical Model of Adaptive
     Mutation.Biosystems,1999,50(3):203—211
    36 DW.Boeringer.DH.Werner.Adaptive Mutation Parameter Toggling Genetic
     Algorithm for Phase—Only Array Synthesis. Electronics Letters. 2002,
     38(25):1618—1619
    37 Z.Michalewicz.Genetic Algorithms+Data Struction=Evolution Program.
     Springer,l992
    38 M.Scrinvas.Adaptive Probablity of Crossover and Mutation in Genetic Algorithms.
    IEEE Trans.on SMC,1994,24(4):656—666
    39 D.Whitley.Genitor II:A Distributed Genetic Algorithm.J.Expt.Ther.Intell.,1990.
     (2):189—214
    40方咸云,方千山,王永初.双变异率自适应遗传算法研究及其应用.南昌航空工业学院学报(自然科学版),2002,16(2):17-20
    
    4l 吴志远,邵惠鹤,吴新余.一种新的自适应遗传算法及其在多峰值函数优化中的
     应用.控制理论与应用,1999,16(1):127-129
    42 L.B.Booker.D.E.Gofdberg.J.H.Holland.Classifier Systems and Genetic
     Algorithms.Artificial Intelligence,1989,40(2):235—282
    43 A.Bertoni.M.Dorigo.Implicit Parallelism in Genetic Algorithms.Artificial
     Intelligence.1993.61(3):307—314
    
    
    
    44徐宗本,李国.解全局优化问题的仿生类算法(I)一模拟演化算法.运筹学杂志,
     1995,14(2):1-13
    45 D.B.Fogel.An Introduction to Simulated Evolutionary Optimization.IEEE Trans.
     on Neural Networks,1994,5(1):3一14
    46 D.E.Goldberg,P.Segreat.Finite Markov Chain Analysis of Genetic A lgorithms.
     Proceedings of The 2nd International COnference on Gas.Massachuwette Institute
     ofTechnology Cambridge,MA,1987:l一8
    47 A.E.Eiben. Global Convergence of Genetic Algorithms:An Infinite Markov
     Chain Analysis,Parallel Problem Solving from Nature.Berlin:Spring-Verlag.
     1991:4一12
    48 G.Rudolph.Convergence Analysis of Canonical Genetic Algorithms.IEEE Trans.
     on Neural Network,1994,5(1):96-101
    49 J.Suzuki.A Markov Chain Analysis on Simple Genetic Algorithms.IEEE Trans.on
     Systems,Man and Cybernetics,1995,25(4):655—659
    50彭宏,王兴华.具有 Elitist 选择的遗传算法的收敛速度估计.科学通报,1997,
     42(2):144-147
    5l周克民,胡云昌.遗传算法计算效率的改进.控制理论与应用,2002,19(5):
     812—814
    52 M.Li,X.Q.Yang,L.X.Zhou.Genetic Algorithm with Dynamic Regional
     Multi—Species,2003,29(2):212—2l8
    53 J.Lienig.A Parallel Genetic A Lgorithm for Performance—driven VLSI Routing.IEEE
     Transactions on Evolutionary Computation,1997,1(1):29—39
    54 Y.Fukuyama.Hsaio—Dong Chiang.A Parallel Genetic Algorithm for Generation
     Expansion Planning.IEEE Transactions on Power Systems,1996,l1(2):955—96l
    55 C.Laquerbe.P.Floquet.L.Pibouleau.et.a1.Synthesis of RTD Models via
     Stochastfc Procedures:Simulated Annealing and Genetic Algorithm.Computers and
     Chemical Engineering,2001,25(9):1169一1183
    56 Y.T.Liu.L.Ma.J.J.Zhang.Reactive Power Optimization by GA/SA/TS Combined
     Algorithms.Electrical Power and Energy Systems,2002,24(9):765—769
    57姚俊峰,梅炽,彭小奇,等.混沌遗传算法及其应用.系统一工程,2001,
     19(1):70一74
    
    
    
    58王焱,刘景录,孙一康.基于变尺度混沌优化策略的混和遗传算法.控制与决策,
     2002,17(6):958-970
    59 樊重俊,韩崇昭,胡保生,等.一类约束优化问题的改进遗传算法.控制与决策,
     1996.1l(5):609-612
    60 Alexandre H.F.Dias,Joao A.de Vasconcelos.Multiobjective Genetic Algorithms
     Applied to Solve Optimizfition Problems.IEEE Trans.on Magnetics,2002,
     38(2):l133—1136
    6 1 Y.Collette,P.Siarry,H.I.Wong.A Systematic Comparison of Performance of
     Various Multiple Objective Metaheuristics Using A Common Set of Analytical Test
     Functions.Foundations of Computing and Decision Sciences,2000,25(4):249--271
    62 Byoung Jun Park.Sung Kwun Oh and Witold Pedrycz.Fuzzy Identification by
     Means of Partition of Fuzzy Input Space and An Aggregate Objective Function.
     IEEE International Fuzzy Systems Conference Proceedings,1999:22—25,480-485.
    63 A.F.Gó mez—Skarmeta,F.Jim6nez.Fuzzy Modeling with Hybrid Systems.Fuzzy
     Sets and Systems,1999,104(2):199-208
    64 Yaochu Jin.Fuzzy Modeling of High—Dimensional Systems:Complexity Reduction
     and Interpretability Improvement. IEEE Trans. on Fuzzy Systems,
     2000,8(2):2l2—22l
    65周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社,1999
    66 陈国良,王煦法.庄镇泉,等.遗传算法及其应用.北京:人民邮电出版社,1996
    67刘守生,于盛林,丁勇,等.一种变焦遗传算法.控制与决策(增刊),
     2002:731-734
    68王凌.智能优化算法及其应用.北京:清华人学出版社,施普林格出版社,
     200l:62-82
    69雷德明.利用混沌搜索全局最优解的一种混合遗传算法.系统工程与电子技术.
    
     1999,2l(12):81—82.
    70 徐宁.周尚波,张红民,等.一种混合混沌优化方法及其应用.系统工程与电子
     技术,2003,25(2):226-227.
    71李人卫,王梦光.一种改进的混合遗传算法.信息与控制,1997,26(6):449—454.
    72 韩宗奇,李亮.测定汽车滑行阻力系数的方法.汽车工程,2001,24(4): 364—366
    73 李合生,毛剑琴,代冀阳.基丁遗传算法的广义Takagi-Sugeno 模糊逻辑系统最
    
    
    
     优参数辨识.自动化学报,2002,28(4):581-586
    74 刘副才,关新平,裴润.一种基于模糊规则的非线性系统快速模糊辨识方法.系
     统仿真学报,2002,14(5):547—550
    75 Ji—Chang Lo.Chien—Hsing Yang.A Heuristic Error—Feedback Learning Algorithm
     for Fuzzy Modeling.IEEE Trans.on Systems,Man and Cybernetics,l999,
     29(6):686—69l
    76 Y. Yoshinari.Construction of Fuzzy Models through Clustering Techniques.Fuzzy
     Sets and Systems,1993,54(1):157—165
    77王宏伟,马广富,王子才.一种基于模糊规则的模糊辨识方法.系统仿真学报,
     1998,10(6):61—64
    78 张平安,李仁厚.基于模糊聚类和卡尔曼滤波方法的模糊辨识.控制理论与应用,
     1996,13(5):639—642

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700