基于IOWHA算子的物流需求组合预测模型

发布时间：2018-01-16 23:33

本文关键词：基于IOWHA算子的物流需求组合预测模型　出处：《河北大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着经济全球化和一体化进程的加快，物流业作为前景广阔的新兴服务业已在世界各国发展起来，对经济的影响越来越大，发展现代物流业显然已经成为各国、各区域经济快速发展的必然趋势。为了建设高效运作的物流系统，制定出合理、有效的物流发展政策以适应经济的发展，，就要进行物流需求预测。在此背景下，分析社会经济活动对物流需求的影响，建立适当的定量预测模型，对物流需求进行科学预测，可以为物流规划和物流需求态势分析提供重要依据。论文旨在结合物流需求预测相关理论和研究现状，建立物流需求预测指标体系，选取合适的单项预测方法，并在此基础上构建组合预测模型，分别对实例进行预测和分析，以寻找能够提高预测精度的方法。本文首先对物流需求的基本理论进行概述，包括物流需求的定义、特点及主要经济影响因素，并在考虑物流需求预测指标选取原则的基础上，结合我国物流历史统计数据不足的现状，重点阐述了选取各经济指标、货运量及货物周转量作为物流需求预测所需量化指标的合理性，据此构建了物流需求预测指标体系。然后根据物流需求预测特点和预测方法特点的分析，论文选取灰色预测法和RBF神经网络进行预测模型的构建。基于此，针对单项预测方法提供信息有限、预测误差大及传统组合预测方法权系数不变、目标准则单一的问题，引用最优加权组合建模理论，将灰色关联度、向量夹角余弦和相关系数分别与IOWHA算子相结合，提出3种新的组合预测模型权重确定方法，并应用新的权重确定方法，构建了3种基于灰色模型和RBF神经网络模型的最优组合预测模型。最后，应用灰色预测模型、RBF神经网络预测模型和3种组合预测模型分别对北京市物流需求进行预测，通过比较分析说明基于相关性指标的IOWHA算子组合预测模型能够有效提高预测精度。
[Abstract]:With the acceleration of economic globalization and integration, the logistics industry as a promising emerging service industry has been developed in the world, the impact on the economy is growing, the development of modern logistics industry has obviously become a country. The inevitable trend of the rapid development of regional economy. In order to build an efficient logistics system and formulate reasonable and effective logistics development policies to adapt to the economic development, we must forecast the logistics demand. This paper analyzes the influence of social economic activities on logistics demand, establishes appropriate quantitative forecasting model, and makes scientific prediction of logistics demand. It can provide important basis for logistics planning and logistics demand situation analysis. The purpose of this paper is to establish a logistics demand forecasting index system and select a suitable single forecasting method combined with logistics demand forecasting theory and research status. On this basis, the combined prediction model is constructed, and the examples are forecasted and analyzed to find the method to improve the prediction accuracy. This paper first summarizes the basic theory of logistics demand, including the definition of logistics demand, characteristics and main economic factors, and on the basis of considering the principle of logistics demand prediction index selection. Combined with the current situation of insufficient historical statistical data of logistics in China, the rationality of selecting various economic indicators, freight volume and freight turnover as the quantitative indicators needed for logistics demand prediction is emphasized. According to the analysis of the characteristics of logistics demand forecasting and forecasting methods, the paper selects the grey forecasting method and RBF neural network to build the forecasting model. Aiming at the problems of limited information provided by single prediction method, large prediction error, constant weight coefficient of traditional combination forecasting method and single objective criterion, the grey correlation degree is introduced by using the optimal weighted combination modeling theory. Based on the combination of vector angle cosine and correlation coefficient with IOWHA operator, three new methods for determining the weight of combined prediction model are proposed, and a new weight determination method is applied. Three optimal combination forecasting models based on grey model and RBF neural network model are constructed. Finally, the grey prediction model is applied. RBF neural network forecasting model and three combined forecasting models are used to forecast the logistics demand of Beijing. The comparison and analysis show that the IOWHA operator combination prediction model based on correlation index can effectively improve the prediction accuracy.
【学位授予单位】：河北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP183;F259.2

【参考文献】