粒子群-神经网络集成学习算法气象预报建模研究

吴建生; 刘丽萍; 金龙

粒子群-神经网络集成学习算法气象预报建模研究

1.
柳州师范高等专科学校数学与计算机科学系, 广西, 柳州, 545004
2.
毕节地区气象局, 贵州, 毕节, 551700
3.
广西区气象减灾研究所, 广西, 南宁, 530022

基金项目: 国家自然科学基金资助项目(40675023);国家科技部社会公益性研究专项(2004DIB3J122)共同资助

计量
- 文章访问数: 949
- HTML全文浏览量: 10
- PDF下载量: 1548
- 被引次数: 0
出版历程
- 收稿日期: 2007-06-04
- 修回日期: 2007-11-26

STUDY ON THE METEOROLOGICAL PREDICTION MODEL USING THE LEARNING ALGORITHM OF NEURAL ENSEMBLE BASED ON PSO ALGORITHMS

1.
Department of Mathematic and Computer, Liuzhou Teacher Colleague, Liuzhou 545004, China
2.
Meteorological Bureau of Bijie Prefecture, Bijie 551700, China
3.
Guangxi Research Institute of Meteorological Disasters Mitigation, Naning 530022, China

摘要

摘要: 针对BP神经网络在实际气象预报应用中,网络结构难以确定以及网络极易陷入局部解问题,提出一种基于神经网络的粒子群集成学习算法的气象预报模型,以BP算法为基本框架,在学习过程中引入粒子群算法,优化设计神经网络的网络结构和初始连接权,获得一组合适网络结构和初始连接权,再进行新一轮BP神经网络训练,获得一批独立的神经网络个体,以"误差绝对值和最小"为最优准则,采用线性规划方法计算各集成个体的权系数,生成神经网络的输出结论,以此建立短期气候预测模型。以广西的月降水量进行实例分析,计算结果表明该方法学习能力强、泛化性能高,能够有效提高系统预测的准确率。
- 神经网络集成 /
- 粒子群优化 /
- 最优组合
Abstract: For the difficulty in deciding on the structure of BP network in real meteorological application and the tendency for the network to transform to an issue of local solution,a hybrid Particle Swarm Optimization Algorithm based on Artificial Neural Network(PSO-BP) model is proposed for monthly mean rainfall of the whole area of Guangxi.It combines Particle Swarm Optimization(PSO) with BP,that is,the number of hidden nodes and connection weights are optimized by the implementation of PSO operation.The method produces a better network architecture and initial connection weights,trains the traditional backward propagation again by training samples.The ensemble strategy is carried out for the linear programming to calculate the best weights based on the "east sum of the error absolute value" as the optimal rule.The weighted coefficient of each ensemble individual is obtained.The results show that the method can effectively improve learning and generalization ability of the neural network.
- neural network ensemble /
- particle swarm optimization /
- optimal combination

HTML全文

参考文献(28)

[1]	胡江林,涂松柏,冯光柳.基于人工神经网络的暴雨预报方法探讨[J].热带气象学报,2003,19(4):422-428.
[2]	HSIEH W W.Nonlinear canonical correlation analysis of the tropical Pacific climate variability using Neural Network Approach[J].Journal ofClimate,2001,14(12):2 528-2 539.
[3]	GRIORGIO C,GIOGIO G.Coupling Fuzzy Modeling and Neural Networks for River Flood Prediction[J].IEEE Transactions on Systems,Man,andCybernetic-Part C:Applications and Reviews,2005,25(3):382-388.
[4]	吴建生,金龙,汪灵枝.遗传算法进化设计BP神经网络气象预报建模研究[J].热带气象学报,2006,22(4):411-416.
[5]	何慧,金龙,覃志年,等.基于BP神经网络模型的广西月降水量降尺度预报[J].热带气象学报,2007,23(1):72-77.
[6]	金龙,况雪源,等.人工神经网络预报模型过拟和研究[J].气象学报,2004,62(1):62-69.
[7]	HANSEN L K,SALAMON P.Neural network ensembles[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,12(10):993-1001.
[8]	SOLLICH P,KROGH A.Learning with Ensembles:How Over-fitting can be useful[C] //Advances in Neural Information Processing Systems 8,Cambridge:MIT Press,1996:190-196.
[9]	周志华,陈世福.神经网络集成[J].计算机学报,2002,25(1):1-8.
[10]	MAO J.A case study on bagging boosting and basic ensembles of neural networks for OCR[C] //Processing International Joint Conference onNeural Networks 1998.Anchorage:International Joint Conference on Neural Networks,1998:1 828-1 833.
[11]	GUTTA S,WECHSLER H.Face recognition using hybrid classifier systems[C] //Proceeding International Joint Conference on Neural Networks 1996.Washington DC:Proceeding International Joint Conference on Neural Networks,1996:1 017-1 022.
[12]	SOLLICH P,INTRATOR N.Classification of seismic signals by integrating ensembles of neural networks[J].IEEE Transactions SignalProcessing,1998,46(5):1 194-1 021.
[13]	NING L,HUAJIE Z,JINJIANG L,et al.Speculated Lesion Detection in digital mammogram based on Artificial Neural Network Ensemble[C].Advances in Neural Networks ISNN,Springer Press,2005,3:790-795.
[14]	SCHAPIRE R E.The strength of weak learn-ability[J].Machine Learning,1990,5(2):197-227.
[15]	BREIMAN L.Bagging prediction[J].Machine Learning,1996,24(2):123-140.
[16]	PERRONE M P,COOPER L N.When network disagree:Ensemble method for Hybrid Neural Networks[R] //Artificial Neural Networks for Speechand Image processing.New York:Chapm & Hall,1993:126-142.
[17]	MERZ C J,PAZZANI M J.A principal components approach to combining regression estimates[J].Machine Learning,1999,36(1-2):9-32.
[18]	ZHIHUA Z,JIANXIN W,WEI T.Ensembling neural networks:Many could be better than all.Artificial Intelligence[J].2002,137(2):239-263.
[19]	BONABEAU E,DORIGO M,THERAULAZ G..Inspiration for optimization from social insect behavior[J].Nature,2000,406(6):39-42.
[20]	XIAOHUI H,EBERHART R.Multi-objective optimization using dynamic neighborhood particle swarm optimization[C] //Proceeding ofcongress on Evolutionary Computation.Hawaii:Ccongress on Evolutionary Computation,2002:1 677-1 681.
[21]	高海兵,高亮,周驰,等.基于粒子群优化的神经网络训练算法研究[J].电子学报,2004,32(9):1 572-1 574.
[22]	RUMLHART D E,HINTON G E,WILLIAMS R J.Learning representations by back propagating errors[J].Nature,1986,323(9):533-536.
[23]	REED R.Pruning Algorithms-A Survey[J].IEEE Transactions on Neural Networks,1993,4(5):740-747.
[24]	RIGET J,VESTERSTROM J S.A diversity-guided particle swarm optimizer-the ARPSO[R].Technical Report 2002-02,Department of ComputerScience,University of Aarhus,2002:345-350.
[25]	马振华.运筹学与最优化理论[M].北京:清华大学出版社,1998:235-425.
[26]	VAUTARD.SSA:a toolkit for noisy chaotic signals[J].Physical D,1992,58:95-126.
[27]	魏凤英,曹鸿兴.长期预测的数学模型及应用[M].北京:气象出版社,1990:142-146.
[28]	王惠文.偏最小二乘回归方法及其应用[M].北京:国防工业出版社,1999:258-365.