[1]石波,张冬青,马开平,等.改进RBF神经网络在我国大豆价格预测中的应用研究[J].大豆科学,2016,35(02):310-314.[doi:10.11861/j.issn.1000-9841.2016.02.0310]
　SHI Bo,ZHANG Dong-qing,MA Kai-ping,et al.Soybean Price Prediction in China Based on Improved RBF Neural Network[J].Soybean Science,2016,35(02):310-314.[doi:10.11861/j.issn.1000-9841.2016.02.0310]
点击复制

改进RBF神经网络在我国大豆价格预测中的应用研究

《大豆科学》[ISSN:1000-9841/CN:23-1227/S] 卷: 第35卷 期数: 2016年02期 页码: 310-314 栏目: 出版日期: 2016-03-20
Title:
Soybean Price Prediction in China Based on Improved RBF Neural Network
作者:
石波; 张冬青; 马开平; 刘欢
南京农业大学 工学院，江苏 南京 210031
Author(s):
SHI Bo;  ZHANG Dong-qing;  MA Kai-ping;  LIU Huan
College of Engineering, Nanjing Agricultural University, Nanjing 210031, China
关键词:
大豆; 价格; RBF; 遗传算法
Keywords:
Soybean;  Price;  RBF Neural Network;  Genetic algorithm
DOI:
10.11861/j.issn.1000-9841.2016.02.0310
文献标志码:
A
摘要:
我国大豆价格受国内外多种因素共同影响，具有非线性、随机性和高噪音等特点，采用传统数学模型进行预测，不仅分析难度大，预测误差也很大。RBF神经网络以其优良的逼近性能而被广泛应用于非线性时间序列预测之中。本文提出一种基于遗传算法优化RBF神经网络的我国大豆价格预测模型，该模型为多维输入单维输出的多变量预测模型，模型的初始输入由大豆价格的历史数据和相关影响因素数据组成。采用遗传算法对RBF神经网络输入层节点数、基函数中心、扩展常数和输出层权值进行优化，模型可以从初始输入变量中自主选择最合适的输入变量组合作为模型的输入。采用2009-2014年的大豆价格数据进行预测研究，用2009-2013年的数据作为训练集，2014年的数据作为测试集，改进RBF神经网络通过自主识别和选取中国大豆进口量、中国消费者信心指数和进口大豆到港分销价格3个因素作为相关影响因素的输入。结果表明：模型预测精度较高、泛化能力较强，能够很好捕捉大豆价格变化规律，可为大豆市场价格的准确预报提供参考借鉴。
Abstract:
Soybean price in China is influenced by various factors at home and abroad.It is difficult to analyze and predict soybean price using the traditional mathematical model for which has the characteristics of nonlinearity, randomness and high noise. RBF Neural Network is widely used in nonlinear time series prediction for its excellent approximation performance.In this paper, one price prediction model of soybean is proposed based on improved RBF Neural Network model, which is multivariable predictive model with multidimensional inputs and one-dimensional output.The initial inputs of model consist of historical data and the influencing factors of soybean.The node numbers of input layer, the centers and widths of Gaussian kernel and output layer weights are optimized by genetic algorithm.The improved model can select the most appropriate variables as inputs of the model automatically from the initial inputs.By using the data of soybean price between 2009 and 2014 to train and forecast. The improved RBF Neural Network select soybean imports of China, consumer confidence index of China and the distribution price in port of imported soybeans as the inputs of the related influence factors through automatically identification.The model’s prediction error is 3.64%, prediction results showed that the model prediction accuracy was high and generalization ability was strong.The model can capture the change rule of soybean price accurately.The model could provide reference for the accurate prediction of soybean market price.

参考文献/References:

［1］刘家富, 周慧秋, 李孝忠. 国内大豆市场价格波动及其影响因素分析［J］. 东北农业大学学报(社会科学版), 2010,8(4):10-13.(Liu J F, Zhou H Q, Li X Z.Analysis on price fluctuation and affecting factors of soybean in China［J］.Journal of Northeast Agricultural University(Social Science Edition), 2010, 8(4):10-13.)

［2］Fernández C, Soria E, SánchezSeiquer P, et al.Weekly milk prediction on dairy goats using neural networks［J］. Neural Computing and Applications, 2007, 16(4-5): 373-381.
［3］Lee C C, Chiang Y C, Shih C Y, et al. Noisy Time series prediction using M.estimator based robust radial basis function neural networks with growing and pruning techniques ［J］. Expert System with Applications, 2009, 36: 4717-4724.
［4］孙素芬, 罗长寿. 基于RBF神经网络的蔬菜价格预报研究［J］. 中国农学通报, 2011, 27(28):269-273.(Sun S F, Luo C S. Research on the prediction of vegetable price based on the RBF neural network model［J］. Chinese Agricultural Science Bulletin, 2011, 27(28):269-273.)
［5］屠星月,薛佳妮,郭承坤, 等. 基于时间序列与RBF的农产品市场价格短期预测模型［J］.广东农业科学, 2014(23):168-173.(Tu X Y, XUE J N, Guo C K, et al. Short-term forecast of agricultural products price based on time series and RBF［J］.Guangdong Agricultural Science, 2014(23):168-173.)
［6］罗长寿. 基于神经网络与遗传算法的蔬菜市场价格预测方法研究［J］.科技通报, 2011, 27(6):881-885.(Luo C S. Research on the prediction method of vegetable price based on neural network and genetic algorithm［J］. Bulletin of Science and Technology, 2011, 27(6):881-885.)
［7］Holland J H.Adaptation in natural and artificial system ［M］ Cambridge: MIT Press, 1975: 100-120.
［8］Niyogo P, Girosi F.On the relationship between generalization error, hypothesis complexity, and sample complexity for radial basis function［J］. Neural Computation, 1996, 8: 819-842.
［9］Segal R, Madnani M L, Madnani S. Radial basis function (RBF) network adaptive power system stabilizer［J］. IEEE Transactions on Power Systems, 2000,5(2):722-727.
［10］武方方, 赵银亮.一种基于蚁群聚类的径向基神经网络［J］.西安交通大学学报, 2006, 40(4): 386-389.(Wu F F, Zhao Y L.Radial basis function neural network based on ant colony clustering［J］.Journal of Xi’an Jiaotong University, 2006, 40(4): 386-389.)
［11］吴成茂, 范九伦. 确定RBF神经网络隐层节点数的最大矩阵元法［J］.计算机工程与应用, 2004, 20: 77-79. (Wu C M, Fan J L. Maximal mateix element method for determining the number of hidden nodes of RBF neural networks［J］.Computer Engineering and Applications, 2004, 20: 77-79.)
［12］刘欢, 张冬青.基于分位数回归的国产大豆价格影响因素分析［J］.大豆科学, 2014, 33(5):759-763.(Liu H, Zhang D Q. Analysis on influencing factors of domestic soybean price based on quantile regression［J］.Soybean Science, 2014, 33(5): 759-763.)
［13］王雪尽.我国大豆产业的危机及对策［J］. 经济问题, 2010(1):81-86(Wang X J.Crisis and counter measures of soybean industry in China［J］. Journal of Economic Issues, 2010(1):81-86.)

相似文献/References:

[1]刘章雄,李卫东,孙石,等.1983～2010年北京大豆育成品种的亲本地理来源及其遗传贡献[J].大豆科学,2013,32(01):1.[doi:10.3969/j.issn.1000-9841.2013.01.002]
　LIU Zhang-xiong,LI Wei-dong,SUN Shi,et al.Geographical Sources of Germplasm and Their Nuclear Contribution to Soybean Cultivars Released during 1983 to 2010 in Beijing[J].Soybean Science,2013,32(02):1.[doi:10.3969/j.issn.1000-9841.2013.01.002]
[2]李彩云,余永亮,杨红旗,等.大豆脂质转运蛋白基因GmLTP3的特征分析[J].大豆科学,2013,32(01):8.[doi:10.3969/j.issn.1000-9841.2013.01.003]
　LI Cai-yun,YU Yong-liang,YANG Hong-qi,et al.Characteristics of a Lipid-transfer Protein Gene GmLTP3 in Glycine max[J].Soybean Science,2013,32(02):8.[doi:10.3969/j.issn.1000-9841.2013.01.003]
[3]王明霞,崔晓霞,薛晨晨,等.大豆耐盐基因GmHAL3a的克隆及RNAi载体的构建[J].大豆科学,2013,32(01):12.[doi:10.3969/j.issn.1000-9841.2013.01.004]
　WANG Ming-xia,CUI Xiao-xia,XUE Chen-chen,et al.Cloning of Halotolerance 3 Gene and Construction of Its RNAi Vector in Soybean (Glycine max)[J].Soybean Science,2013,32(02):12.[doi:10.3969/j.issn.1000-9841.2013.01.004]
[4]张春宝,李玉秋,彭宝,等.线粒体ISSR与SCAR标记鉴定大豆细胞质雄性不育系与保持系[J].大豆科学,2013,32(01):19.[doi:10.3969/j.issn.1000-9841.2013.01.005]
　ZHANG Chun-bao,LI Yu-qiu,PENG Bao,et al.Identification of Soybean Cytoplasmic Male Sterile Line and Maintainer Line with Mitochondrial ISSR and SCAR Markers[J].Soybean Science,2013,32(02):19.[doi:10.3969/j.issn.1000-9841.2013.01.005]
[5]卢清瑶,赵琳,李冬梅,等.RAV基因对拟南芥和大豆不定芽再生的影响[J].大豆科学,2013,32(01):23.[doi:10.3969/j.issn.1000-9841.2013.01.006]
　LU Qing-yao,ZHAO Lin,LI Dong-mei,et al.Effects of RAV gene on Shoot Regeneration of Arabidopsis and Soybean[J].Soybean Science,2013,32(02):23.[doi:10.3969/j.issn.1000-9841.2013.01.006]
[6]杜景红,刘丽君.大豆fad3c基因沉默载体的构建[J].大豆科学,2013,32(01):28.[doi:10.3969/j.issn.1000-9841.2013.01.007]
　DU Jing-hong,LIU Li-jun.Construction of fad3c Gene Silencing Vector in Soybean[J].Soybean Science,2013,32(02):28.[doi:10.3969/j.issn.1000-9841.2013.01.007]
[7]张力伟,樊颖伦,牛腾飞,等.大豆“冀黄13”突变体筛选及突变体库的建立[J].大豆科学,2013,32(01):33.[doi:10.3969/j.issn.1000-9841.2013.01.008]
　ZHANG Li-wei,FAN Ying-lun,NIU Teng-fei?,et al.Screening of Mutants and Construction of Mutant Population for Soybean Cultivar "Jihuang13”[J].Soybean Science,2013,32(02):33.[doi:10.3969/j.issn.1000-9841.2013.01.008]
[8]盖江南,张彬彬,吴瑶,等.大豆不定胚悬浮培养基因型筛选及基因枪遗传转化的研究[J].大豆科学,2013,32(01):38.[doi:10.3969/j.issn.1000-9841.2013.01.009]
　GAI Jiang-nan,ZHANG Bin-bin,WU Yao,et al.Screening of Soybean Genotypes Suitable for Suspension Culture with Adventitious Embryos and Genetic Transformation by Particle Bombardment[J].Soybean Science,2013,32(02):38.[doi:10.3969/j.issn.1000-9841.2013.01.009]
[9]王鹏飞,刘丽君,唐晓飞,等.适于体细胞胚发生的大豆基因型筛选[J].大豆科学,2013,32(01):43.[doi:10.3969/j.issn.1000-9841.2013.01.010]
　WANG Peng-fei,LIU Li-jun,TANG Xiao-fei,et al.Screening of Soybean Genotypes Suitable for Somatic Embryogenesis[J].Soybean Science,2013,32(02):43.[doi:10.3969/j.issn.1000-9841.2013.01.010]
[10]刘德兴,年海,杨存义,等.耐酸铝大豆品种资源的筛选与鉴定[J].大豆科学,2013,32(01):46.[doi:10.3969/j.issn.1000-9841.2013.01.011]
　LIU De-xing,NIAN Hai,YANG Cun-yi,et al.Screening and Identifying Soybean Germplasm Tolerant to Acid Aluminum[J].Soybean Science,2013,32(02):46.[doi:10.3969/j.issn.1000-9841.2013.01.011]

备注/Memo

基金项目：国家自然科学基金（71101072，71301077，71401076）；南京农业大学中央高校基本科研业务费人文社会科学基金（SK2014011）。第一作者简介：石波（1989-），男，硕士，主要从事生产运作管理、产品市场扩散等研究。E-mail：shibodengge@163.com。通讯作者：马开平（1976-），女，博士，副教授，主要从事生产运作管理、产品市场扩散等研究。E-mail：makaiping@njau.edu.cn。

更新日期/Last Update: 2016-04-04