本文選題：配電網(wǎng)備品備件　切入點：支持向量機(jī)　出處：《華南理工大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：鑒于電網(wǎng)企業(yè)的專業(yè)性與特殊性，配電網(wǎng)備品備件在滿足設(shè)備檢修維護(hù)需要的同時，還需盡量減少庫存積壓和占用庫存資金，以保障電網(wǎng)生產(chǎn)的安全性與經(jīng)濟(jì)性，實現(xiàn)電網(wǎng)企業(yè)的高效運營。因此，合理的庫存需求預(yù)測模型、科學(xué)的庫存管理策略以及庫存定額算法對提高備品備件管理水平是非常必要的。在此背景下，本文以G供電局為研究對象，對配電網(wǎng)備品備件的需求預(yù)測、管理策略及庫存定額展開具體研究。 本文首先通過實踐調(diào)研等方式收集G供電局配電網(wǎng)備品備件庫存管理現(xiàn)狀的相關(guān)信息，梳理資料并進(jìn)行詳細(xì)的現(xiàn)狀問題分析�？偨Y(jié)出G供電局配電網(wǎng)備品備件現(xiàn)行的庫存管理中存在的物資采購計劃具有盲目性、物資分類不合理、物資出入庫記錄存在紕漏、庫存定額不科學(xué)和庫存管理模式過于傳統(tǒng)等問題。 通過研究庫存需求預(yù)測的相關(guān)理論以及配電網(wǎng)備品備件的特征，本文基于支持向量機(jī)回歸預(yù)測法（SVM，，Support Vector Machine）提出適用于配電網(wǎng)備品備件的需求預(yù)測模型。模型有效考慮了物資的歷史需求量、檢修計劃、設(shè)備故障率以及運行環(huán)境這四個影響物資需求的因素，最后以搶修物資柱上斷路器為例驗證需求預(yù)測方法的合理性與可行性。 本文還提出了適用于配電網(wǎng)備品備件的AB分類庫存管理策略以及庫存定額算法。首先將G供電局配電網(wǎng)備品備件劃分成A、B（包括B1和B2）兩大類，研究確定其相應(yīng)的庫存控制管理模型，并以物資需求預(yù)測結(jié)果為基礎(chǔ)，為A、B類物資分別建立各自庫存管理模型下的庫存定額算法，最后以算例補(bǔ)充說明庫存定額管理算法的實際應(yīng)用。此外，本文針對G供電局還引入虛擬庫存管理策略，為解決實物庫存管理存在的缺陷提出了虛擬庫存具體應(yīng)用建議。 算例結(jié)果表明，本文所提的庫存需求預(yù)測模型，配電網(wǎng)備品備件的分類管理方法、庫存管理控制策略以及庫存定額算法能有效改善G供電局現(xiàn)行庫存管理中存在的問題，對配電網(wǎng)備品備件庫存管理具有良好的工程應(yīng)用價值。
[Abstract]:In view of the specialty and particularity of power grid enterprises, the spare parts of distribution network should not only meet the needs of maintenance and repair of equipment, but also reduce the backlog of inventory and occupy stock funds as far as possible, so as to ensure the safety and economy of power grid production. Therefore, reasonable inventory demand forecasting model, scientific inventory management strategy and inventory quota algorithm are very necessary to improve spare parts management level. Taking G power supply bureau as the research object, this paper studies the demand forecast, management strategy and inventory quota of spare parts in distribution network. Firstly, this paper collects relevant information about the status of spare parts inventory management in distribution network of G Power supply Bureau by means of practical investigation and research. Combing the data and analyzing the current situation in detail, summing up the material purchase plan existing in the current inventory management of spare parts in the distribution network of G Power supply Bureau is blind, the material classification is unreasonable, the material entry and storage record is flawed, The stock quota is not scientific and the inventory management mode is too traditional. By studying the theory of inventory demand forecasting and the characteristics of spare parts in distribution network, Based on the support vector machine regression prediction method, this paper presents a demand forecasting model for spare parts in distribution network. The model effectively considers the historical demand of materials and maintenance plan. The failure rate of equipment and the operating environment are the four factors that affect the material demand. Finally, the rationality and feasibility of the demand prediction method are verified by an example of the circuit breaker on the material post for emergency repair. This paper also puts forward the AB classification inventory management strategy and inventory quota algorithm suitable for distribution network spare parts. Firstly, the spare parts of G power supply bureau are divided into two categories: Agna B (including B1 and B2). This paper studies and determines the corresponding inventory control management model, and based on the forecast results of material demand, establishes the inventory quota algorithm for Agna B material under their respective inventory management model. Finally, an example is given to illustrate the practical application of the inventory quota management algorithm. In addition, this paper introduces virtual inventory management strategy for G power supply bureau, and puts forward some specific suggestions for the application of virtual inventory in order to solve the defects of physical inventory management. The results show that the forecasting model of inventory demand, the classified management method of spare parts in distribution network, the control strategy of inventory management and the algorithm of inventory quota can effectively improve the problems existing in the current inventory management of G Power supply Bureau. It has good engineering application value for spare parts inventory management in distribution network.
【學(xué)位授予單位】：華南理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TM73

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 高正平;;電力企業(yè)備品備件儲備管理模式的新思考[J];電力技術(shù)經(jīng)濟(jì);2006年03期

2 周偉;;核電站備品備件的聯(lián)合儲備管理模式分析[J];發(fā)電設(shè)備;2009年05期

3 任博,張恒喜,蘇暢;基于支持向量機(jī)的飛機(jī)備件需求預(yù)測[J];火力與指揮控制;2005年03期

4 馬秀紅,宋建社,董晟飛;基于回歸分析的備件故障率預(yù)測模型[J];計算機(jī)仿真;2003年11期

5 黃遠(yuǎn)兵;蔡啟明;楊瑋龍;黃燕美;;基于支持向量回歸的服務(wù)備件需求量預(yù)測研究[J];物流科技;2006年10期

6 宋效紅;;基于“庫存風(fēng)險-庫存價值”模型的煤炭企業(yè)采購物資分類探討[J];煤炭經(jīng)濟(jì)研究;2009年02期

7 嚴(yán)婷婷,李宏余;基于AHP的庫存產(chǎn)品分類模型研究[J];物流技術(shù);2005年11期

8 丁留明;崔南方;李晉;;考慮設(shè)備關(guān)鍵性的備件庫存ABC分類兩階段改進(jìn)模型研究[J];物流技術(shù);2006年12期

9 洪芳華;高峻峻;陸爽;;基于需求特性分類的電力物資庫存管理策略設(shè)計[J];物流技術(shù);2011年17期

10 史耀媛;史忠科;;基于非單點模糊正則網(wǎng)絡(luò)的時間序列預(yù)測模型[J];西北大學(xué)學(xué)報(自然科學(xué)版);2006年06期

本文編號：1584350