基于需求特性分類(lèi)的電力物資庫(kù)存控制與需求預(yù)測(cè)方法研究
發(fā)布時(shí)間：2018-10-31 14:12
【摘要】：隨著我國(guó)生產(chǎn)水平和信息化水平的逐步提高,電力相關(guān)行業(yè)對(duì)電力物資的需求日益增多,電力行業(yè)面臨著客戶(hù)需求多變、采購(gòu)周期長(zhǎng)、歷史消耗數(shù)據(jù)缺失與失真、需求預(yù)測(cè)與庫(kù)存控制困難等諸多問(wèn)題,這些都給電力物資的需求管理與庫(kù)存管理提出了很大挑戰(zhàn),因此對(duì)電力物資進(jìn)行需求預(yù)測(cè)與庫(kù)存控制越來(lái)越受到企業(yè)的重視。間斷性需求電力物資對(duì)于企業(yè)來(lái)說(shuō)至關(guān)重要,對(duì)其進(jìn)行良好的需求預(yù)測(cè)與庫(kù)存管理不但可以保證企業(yè)生產(chǎn)運(yùn)營(yíng)的正常進(jìn)行,而且可以大大的降低庫(kù)存成本,提高企業(yè)競(jìng)爭(zhēng)力。本文以上海市電力公司某倉(cāng)庫(kù)運(yùn)營(yíng)管理實(shí)際需求為背景,從電力物資的自身特點(diǎn)出發(fā),在對(duì)其需求特性進(jìn)行分析與科學(xué)分類(lèi)的基礎(chǔ)上,針對(duì)庫(kù)存控制問(wèn)題與間斷性需求預(yù)測(cè)問(wèn)題展開(kāi)研究,論文主要包括以下幾個(gè)方面內(nèi)容：(1)以上海市電力公司某倉(cāng)庫(kù)實(shí)際情況為背景,深入企業(yè)進(jìn)行實(shí)地調(diào)研,了解整個(gè)倉(cāng)庫(kù)管理的一般現(xiàn)狀,分析企業(yè)對(duì)電力物資庫(kù)存管理的需求,確定本課題研究?jī)?nèi)容的重要性與可行性。以歷史消耗數(shù)據(jù)為依據(jù),提取并分析電力物資需求特性,并基于此對(duì)其進(jìn)行科學(xué)全面的分類(lèi),設(shè)計(jì)了電力物資總體庫(kù)存控制策略,并針對(duì)不同類(lèi)型的電力物資,構(gòu)建其動(dòng)態(tài)庫(kù)存控制模型,為實(shí)現(xiàn)有效的庫(kù)存管理提供依據(jù)。(2)在深入了解間斷性需求電力物資需求特征的基礎(chǔ)上,將其整個(gè)預(yù)測(cè)研究工作分為兩部分,即需求發(fā)生時(shí)刻預(yù)測(cè)與需求發(fā)生量預(yù)測(cè)。提出使用BP神經(jīng)網(wǎng)絡(luò)模型預(yù)測(cè)間斷性需求電力物資發(fā)生時(shí)刻,并通過(guò)實(shí)驗(yàn)分析其優(yōu)勢(shì)與缺點(diǎn),在此基礎(chǔ)上提出使用遺傳算法優(yōu)化的BP神經(jīng)網(wǎng)絡(luò)模型進(jìn)行預(yù)測(cè)�？紤]間斷性需求電力物資歷史需求消耗數(shù)據(jù)較少的特點(diǎn),提出使用灰色GM(1,1)模型進(jìn)行需求發(fā)生量的預(yù)測(cè),并通過(guò)具體預(yù)測(cè)實(shí)驗(yàn)說(shuō)明模型的有效性及可行性,最后將兩方面結(jié)合起來(lái)描述對(duì)未來(lái)的預(yù)測(cè)情況,降低了預(yù)測(cè)難度,為間斷性需求電力物資的需求預(yù)測(cè)提供一種有效的方法。(3)從企業(yè)需求的角度出發(fā),詳細(xì)設(shè)計(jì)了電力物資庫(kù)存控制與需求預(yù)測(cè)軟件模塊的總體功能及各個(gè)子模塊功能,以及編程實(shí)現(xiàn)了包括基本數(shù)據(jù)管理、需求特性管理、需求預(yù)測(cè)以及動(dòng)態(tài)庫(kù)存控制等功能模塊,達(dá)到了對(duì)電力物資庫(kù)存進(jìn)行有效控制的目的。
[Abstract]:With the gradual improvement of production level and information level of our country, the demand of electric power related industries for electric power materials is increasing day by day. The electric power industry is faced with changeable customer demand, long purchasing period, lack and distortion of historical consumption data. It is difficult to forecast the demand and control the inventory, which brings great challenge to the demand management and inventory management of electric power materials. Therefore, the enterprises pay more and more attention to the demand prediction and inventory control of electric power materials. Intermittent demand for electric power materials is very important for enterprises. Good demand prediction and inventory management can not only ensure the normal operation of enterprises, but also greatly reduce the cost of inventory and improve the competitiveness of enterprises. Based on the actual demand of a warehouse operation and management of Shanghai Electric Power Company, this paper starts from the characteristics of electric power materials, and analyzes the demand characteristics and classifies them scientifically. The paper mainly includes the following aspects: (1) taking the actual situation of a warehouse of Shanghai Electric Power Company as the background, the paper carries on the field investigation and research deeply to the enterprise, which is aimed at the inventory control problem and the intermittent demand forecasting problem. Understand the general situation of the whole warehouse management, analyze the demand of the electric power material inventory management, determine the importance and feasibility of the research content of this subject. Based on the historical consumption data, this paper extracts and analyzes the demand characteristics of electric power materials, and classifies them scientifically and comprehensively, designs the overall inventory control strategy of electric power materials, and aims at different types of power materials. The dynamic inventory control model is constructed to provide the basis for effective inventory management. (2) on the basis of deeply understanding the demand characteristics of intermittent demand, the whole prediction research work is divided into two parts. That is, demand occurrence time forecast and demand occurrence quantity forecast. In this paper, BP neural network model is used to predict the occurrence time of intermittent demand power materials, and its advantages and disadvantages are analyzed through experiments. On the basis of this, the BP neural network model optimized by genetic algorithm is proposed to forecast. Considering the fact that the historical demand data of discontinuous demand for electric power materials are less, a grey GM (1 ~ 1) model is proposed to predict the amount of demand, and the validity and feasibility of the model are illustrated by the concrete prediction experiments. Finally, the two aspects are combined to describe the forecasting situation in the future, which reduces the difficulty of forecasting, and provides an effective method for forecasting the demand of intermittent demand for electric power materials. (3) from the point of view of enterprise demand, The general function and each sub-module function of electric power material inventory control and demand prediction software module are designed in detail, and the function modules including basic data management, demand characteristic management, demand forecasting and dynamic inventory control are realized by programming. To achieve the purpose of effective control of electricity material inventory.
【學(xué)位授予單位】：東北大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2013
【分類(lèi)號(hào)】：TP183;F426.61;F274
，
本文編號(hào)：2302482
[Abstract]:With the gradual improvement of production level and information level of our country, the demand of electric power related industries for electric power materials is increasing day by day. The electric power industry is faced with changeable customer demand, long purchasing period, lack and distortion of historical consumption data. It is difficult to forecast the demand and control the inventory, which brings great challenge to the demand management and inventory management of electric power materials. Therefore, the enterprises pay more and more attention to the demand prediction and inventory control of electric power materials. Intermittent demand for electric power materials is very important for enterprises. Good demand prediction and inventory management can not only ensure the normal operation of enterprises, but also greatly reduce the cost of inventory and improve the competitiveness of enterprises. Based on the actual demand of a warehouse operation and management of Shanghai Electric Power Company, this paper starts from the characteristics of electric power materials, and analyzes the demand characteristics and classifies them scientifically. The paper mainly includes the following aspects: (1) taking the actual situation of a warehouse of Shanghai Electric Power Company as the background, the paper carries on the field investigation and research deeply to the enterprise, which is aimed at the inventory control problem and the intermittent demand forecasting problem. Understand the general situation of the whole warehouse management, analyze the demand of the electric power material inventory management, determine the importance and feasibility of the research content of this subject. Based on the historical consumption data, this paper extracts and analyzes the demand characteristics of electric power materials, and classifies them scientifically and comprehensively, designs the overall inventory control strategy of electric power materials, and aims at different types of power materials. The dynamic inventory control model is constructed to provide the basis for effective inventory management. (2) on the basis of deeply understanding the demand characteristics of intermittent demand, the whole prediction research work is divided into two parts. That is, demand occurrence time forecast and demand occurrence quantity forecast. In this paper, BP neural network model is used to predict the occurrence time of intermittent demand power materials, and its advantages and disadvantages are analyzed through experiments. On the basis of this, the BP neural network model optimized by genetic algorithm is proposed to forecast. Considering the fact that the historical demand data of discontinuous demand for electric power materials are less, a grey GM (1 ~ 1) model is proposed to predict the amount of demand, and the validity and feasibility of the model are illustrated by the concrete prediction experiments. Finally, the two aspects are combined to describe the forecasting situation in the future, which reduces the difficulty of forecasting, and provides an effective method for forecasting the demand of intermittent demand for electric power materials. (3) from the point of view of enterprise demand, The general function and each sub-module function of electric power material inventory control and demand prediction software module are designed in detail, and the function modules including basic data management, demand characteristic management, demand forecasting and dynamic inventory control are realized by programming. To achieve the purpose of effective control of electricity material inventory.
【學(xué)位授予單位】：東北大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2013
【分類(lèi)號(hào)】：TP183;F426.61;F274
，
本文編號(hào)：2302482
本文鏈接：http://sikaile.net/guanlilunwen/shengchanguanlilunwen/2302482.html
最近更新
教材專(zhuān)著
