發(fā)布時(shí)間：2018-04-01 04:19

  本文選題：生產(chǎn)/庫(kù)存/分銷(xiāo)系統(tǒng)　切入點(diǎn)：庫(kù)存不精確　出處：《東南大學(xué)》2016年博士論文

【摘要】：完整的生產(chǎn)/庫(kù)存/分銷(xiāo)系統(tǒng)控制通常分為三個(gè)主要環(huán)節(jié):原材料庫(kù)存管理、生產(chǎn)控制及分銷(xiāo)調(diào)度。這三個(gè)環(huán)節(jié)分別包含原材料庫(kù)存、在制品庫(kù)存和成品庫(kù)存。而庫(kù)存不精確(庫(kù)存記錄與真實(shí)庫(kù)存的不符)在上述三個(gè)環(huán)節(jié)均客觀存在。庫(kù)存不精確給物料需求計(jì)劃系統(tǒng)和供應(yīng)鏈管理系統(tǒng)的運(yùn)行帶來(lái)了不利影響,進(jìn)而帶來(lái)了成本的增加和收益的減少。因此,本論文將研究一種魯棒集成補(bǔ)貨、生產(chǎn)控制和分銷(xiāo)策略來(lái)克服庫(kù)存不精確對(duì)生產(chǎn)/庫(kù)存/分銷(xiāo)系統(tǒng)三個(gè)主要環(huán)節(jié)帶來(lái)的負(fù)面影響。本文的主要內(nèi)容概括如下:1.針對(duì)考慮補(bǔ)貨時(shí)延和庫(kù)存不精確的原材料庫(kù)存系統(tǒng),研究了魯棒補(bǔ)貨問(wèn)題。該問(wèn)題以最小化庫(kù)存成本、缺貨成本和物料運(yùn)輸成本為目標(biāo)。首先給出了用于估計(jì)真實(shí)庫(kù)存概率分布的遞歸算法。在庫(kù)存水平概率分布的基礎(chǔ)上,提出了一種魯棒再訂貨時(shí)間/再訂貨量策略(RTQ策略)。文中的理論分析和數(shù)值實(shí)驗(yàn)結(jié)果揭示了如下結(jié)論和管理學(xué)啟示:(1)如果存在不可見(jiàn)的庫(kù)存損失及由此帶來(lái)的不精確的庫(kù)存記錄,在經(jīng)典策略(r,0)策略,(R,S)策略,(s,S)策略)控制下,將會(huì)在有限時(shí)間內(nèi)陷入零服務(wù)狀態(tài)(庫(kù)存水平以概率1小于或等于零);如果庫(kù)存損失的概率分布已知,在所提出的RTQ策略下,即使不進(jìn)行盤(pán)點(diǎn),系統(tǒng)也不會(huì)陷入零服務(wù)狀態(tài)且會(huì)保持較高的服務(wù)水平;(2)如果庫(kù)存記錄是精確的,則RTQ策略是近似優(yōu)的,其性能與最優(yōu)參數(shù)下的(r,Q)策略非常接近。(3)為克服庫(kù)存系統(tǒng)中庫(kù)存不精確帶來(lái)的負(fù)面影響,除了盤(pán)點(diǎn)和部署射頻識(shí)別設(shè)備(RFID),也可采用魯棒補(bǔ)貨策略,這樣不會(huì)增加人力或硬件部署和維護(hù)的成本。2.針對(duì)在制品及成品庫(kù)存不精確的生產(chǎn)系統(tǒng),研究了多機(jī)器多品種制造系統(tǒng)的魯棒生產(chǎn)控制問(wèn)題。該問(wèn)題以最小化在制品和成品的庫(kù)存成本以及成品的缺貨成本為目標(biāo)。該問(wèn)題分以下兩步解決:首先,在庫(kù)存水平精確可知的假設(shè)下,根據(jù)動(dòng)態(tài)規(guī)劃原理推導(dǎo)出了簡(jiǎn)化的最優(yōu)生產(chǎn)控制策略(簡(jiǎn)化策略);基于該控制策略的結(jié)構(gòu)特征和真實(shí)庫(kù)存水平條件概率分布,為克服庫(kù)存不精確對(duì)生產(chǎn)控制的負(fù)面影響,提出了一種基于在制品及成品庫(kù)存水平的條件數(shù)學(xué)期望的魯棒生產(chǎn)控制策略(魯棒策略)。數(shù)值仿真實(shí)驗(yàn)將簡(jiǎn)化策略和所提出的魯棒策略在各自的最優(yōu)參數(shù)下進(jìn)行了對(duì)比。結(jié)果表明,魯棒策略的性能總是優(yōu)于簡(jiǎn)化策略。最后對(duì)兩種策略關(guān)于庫(kù)存誤差分布參數(shù)、機(jī)器故障率等系統(tǒng)參數(shù)進(jìn)行了靈敏度分析。結(jié)果表明,在不同的系統(tǒng)參數(shù)下,所提出的魯棒策略都優(yōu)于簡(jiǎn)化策略。本問(wèn)題的研究結(jié)果表明,在生產(chǎn)控制中,即使在制品種類(lèi)多,且數(shù)量持續(xù)變化難以跟蹤,魯棒控制策略仍能有效減小庫(kù)存不精確帶來(lái)的負(fù)面影響。3.針對(duì)考慮補(bǔ)貨時(shí)延和庫(kù)存不精確的分銷(xiāo)系統(tǒng),研究了魯棒庫(kù)存路徑問(wèn)題(Inventory routing problem,IRP)。與傳統(tǒng)IRP問(wèn)題不同的是,我們考慮了補(bǔ)貨時(shí)延,而各零售商的補(bǔ)貨時(shí)延由分銷(xiāo)路徑?jīng)Q定,因此補(bǔ)貨量決策和路徑?jīng)Q策相互耦合。為解決此問(wèn)題,首先研究了庫(kù)存精確情形下考慮補(bǔ)貨時(shí)延的IRP問(wèn)題。以一個(gè)供應(yīng)商兩個(gè)零售商的簡(jiǎn)單情形為研究對(duì)象,證明了該問(wèn)題的最優(yōu)庫(kù)存路徑策略結(jié)構(gòu)為切換曲線類(lèi)型,狀態(tài)(零售商的庫(kù)存位勢(shì))空間被切換曲線劃分為多個(gè)區(qū)域,每個(gè)區(qū)域?qū)?yīng)各自的最優(yōu)補(bǔ)貨路線及最優(yōu)訂至點(diǎn)(order-up-to-level)。根據(jù)最優(yōu)策略的結(jié)構(gòu),針對(duì)現(xiàn)實(shí)規(guī)模的分銷(xiāo)系統(tǒng),提出了一種啟發(fā)式策略并給出了算例來(lái)說(shuō)明該策略的可行性。其次研究了考慮庫(kù)存不精確和補(bǔ)貨時(shí)延的IRP問(wèn)題。該問(wèn)題的解決分為三層:路徑優(yōu)化層(產(chǎn)生和優(yōu)化待選路徑),補(bǔ)貨量?jī)?yōu)化層(對(duì)每條待選路徑,優(yōu)化補(bǔ)貨量),路徑評(píng)估層(對(duì)每條路徑的性能指標(biāo)進(jìn)行評(píng)估)。提出了一種魯棒再訂貨時(shí)間/階段長(zhǎng)度/再訂貨量算法來(lái)優(yōu)化補(bǔ)貨時(shí)間、補(bǔ)貨量。最后,提出了一種基于遺傳算法的路徑?jīng)Q策方法。4.針對(duì)考慮庫(kù)存不精確的補(bǔ)貨/生產(chǎn)/分銷(xiāo)系統(tǒng),研究了魯棒集成補(bǔ)貨、生產(chǎn)控制及分銷(xiāo)問(wèn)題。為解決此問(wèn)題,首先建立了系統(tǒng)完整的數(shù)學(xué)模型,并在前述工作的基礎(chǔ)上構(gòu)建了集成魯棒補(bǔ)貨/生產(chǎn)/分銷(xiāo)控制策略。通過(guò)遞歸估計(jì)的未來(lái)周期零售終端庫(kù)存水平的概率分布預(yù)測(cè)下次配送的時(shí)間和所需的配送量,以此作為生產(chǎn)環(huán)節(jié)下一階段的生產(chǎn)目標(biāo)。證明了該種情形的最優(yōu)生產(chǎn)控制策略為{0,μ}控制,即若開(kāi)始生產(chǎn)就以最大生產(chǎn)率來(lái)生產(chǎn)。在生產(chǎn)環(huán)節(jié),使用基于仿真的方法優(yōu)化最佳生產(chǎn)開(kāi)始時(shí)間和各周期對(duì)各種原材料需求的概率分布,以此作為原材料庫(kù)存控制環(huán)節(jié)進(jìn)行原材料補(bǔ)貨的依據(jù)。數(shù)值實(shí)驗(yàn)驗(yàn)證了該策略的可行性。通過(guò)對(duì)庫(kù)存不精確程度的靈敏度分析,我們發(fā)現(xiàn)在該策略的控制下,當(dāng)各環(huán)節(jié)庫(kù)存不精確程度的變化時(shí),雖然控制策略的參數(shù)隨之變化,但各環(huán)節(jié)的服務(wù)水平保持不變,體現(xiàn)了策略較好的魯棒性。對(duì)于關(guān)于庫(kù)存不精確問(wèn)題的研究,本文的主要貢獻(xiàn)是:(1)將庫(kù)存不精確情形下的補(bǔ)貨問(wèn)題從現(xiàn)有的報(bào)童模型擴(kuò)展至考慮固定訂貨成本和補(bǔ)貨時(shí)延的一般模型,并提出相應(yīng)的魯棒再訂貨點(diǎn)/再訂貨量策略;(2)研究了在制品及成品庫(kù)存不精確情形下的生產(chǎn)控制問(wèn)題,提出了一種基于在制品及庫(kù)存水平的條件數(shù)學(xué)期望的魯棒生產(chǎn)控制策略。(3)研究了分銷(xiāo)過(guò)程中成品庫(kù)存不精確情形下的庫(kù)存路徑問(wèn)題,考慮了分銷(xiāo)過(guò)程中運(yùn)輸時(shí)延不可忽略的現(xiàn)實(shí)因素,證明了最優(yōu)庫(kù)存路徑策略的結(jié)構(gòu)特征,并提出了魯棒庫(kù)存路徑策略。(4)研究了庫(kù)存不精確情形下的補(bǔ)貨/生產(chǎn)/分銷(xiāo)系統(tǒng)的魯棒集成控制,提出了魯棒集成補(bǔ)貨/生產(chǎn)/分銷(xiāo)控制策略。并通過(guò)算例驗(yàn)證了其可行性和魯棒性。
[Abstract]:Complete production / inventory / distribution control system is usually divided into three major parts: raw materials inventory management, production control and distribution scheduling. These three aspects include the inventory of raw material, WIP and finished goods inventory and inventory inaccuracy (inconsistent inventory records and real stock) in the above three aspects are the objective existence. Adverse effects of inventory inaccuracy to the MRP system and supply chain management system, and reduces the cost and increase the profit. Therefore, this paper will research a robust integrated replenishment, production control and distribution strategies to overcome the inventory inaccuracy of production / inventory / distribution negative impact bring the three main parts of the system. The main contents of this paper are as follows: 1.. Considering the inventory of raw materials inventory replenishment system delay and imprecise, studies the robust problem. The problem with replenishment To minimize the inventory cost, shortage cost and transportation cost as the goal. Firstly, a recursive algorithm is used to estimate the real stock probability distribution. Based on the probability distribution of the inventory level, this paper presents a robust reorder time / quantity strategy (RTQ strategy). Theoretical analysis and numerical experiment results in this paper reveals the following conclusions and management implications: (1) if there is no visible inventory loss and the resulting inaccurate inventory records, in the classic strategy (R, 0), (R, S) strategy, (s, S)) under the control strategy will be in zero in finite time state (the inventory level is less than or equal to zero with probability 1); if the known probability distribution of inventory loss, the proposed RTQ strategy, if not for inventory, the system does not fall into the zero state and will maintain a high level of service; (2) if the inventory records are accurate, then RTQ strategy is a better approximation, the performance and the optimal parameters of the (R, Q) strategy is very close. (3) in order to overcome the negative effects in the inventory system inventory inaccuracy brought about, in addition to checking and deployment of RFID equipment (RFID), also can use robust replenishment strategy, so as not to increase human or hardware deployment and the cost of maintenance for the.2. production system, WIP and finished goods inventory not accurate, robust control problem of multi machine production variety manufacturing system. The problem is to minimize the WIP and finished goods inventory cost and shortage cost of finished products as the goal. The problem is divided into the following two steps: first, to solve the inventory level in the accurate under the assumption, according to the dynamic programming principle is derived for the optimal production control policy (simplified simplified strategy); structural features of the control strategy and the real conditions of the probability distribution based on the inventory level, inventory is not to overcome Precise negative impact on production control, put forward a kind of conditional expectation WIP and FG inventory level based on the robust control strategy of production (robust strategy). Numerical simulation results will simplify the strategy and the proposed robust strategies are compared in their respective optimal parameters. The results show that the performance of robust strategy is always better than the simplified strategy. At the end of two strategies about inventory parameter error distribution, machine failure rate and other system parameters are analysed. The results show that the different parameters of the system, the proposed robust strategies are better than the simplified strategy. This research results show that in production control, even in many types, and the number of continuous it is difficult to change tracking, robust control strategy still can effectively reduce the negative impact of inaccurate inventory.3. considering sales and inventory replenishment system delay is not accurate, Research on robust inventory routing problem (Inventory routing, problem, IRP). Different from the traditional IRP problem, we consider the replenishment time delay, and the delay of each retailer's replenishment is decided by the distribution path, so the replenishment quantity decision and decision path coupling. In order to solve this problem, we study the IRP problem considering time delay inventory replenishment the exact circumstances. In a simple case of one supplier and two retailers as the research object, proved that the problem of optimal inventory strategy for path switching curve type structure, state (retailer potential) switching curve space is divided into multiple regions, each region corresponding to their respective optimal replenishment route and optimal order up to (order-up-to-level). According to the structure of the optimal strategy, the distribution system according to the reality of the scale, this paper proposes a heuristic strategy and an example is given to illustrate the feasibility of the strategy. Secondly considering the IRP problem of inventory inaccuracy and replenishment delay. To solve the problem is divided into three layers: the path optimization layer (generation and Optimization for path), the replenishment quantity optimization layer (for each candidate path, optimal replenishment quantity), path evaluation layer (for each path of evaluation index). A robust reorder time / phase length / ordering algorithm to optimize the replenishment time, the replenishment quantity is put forward. Finally, proposes a path based decision method.4. genetic algorithm considering inaccurate inventory replenishment / production / distribution system, the research of robust integrated replenishment, control and distribution problems the production. In order to solve this problem, the first complete mathematical model of the system is established, and the construction of the integrated robust replenishment / production / distribution control strategy on the basis of previous work. Through the recursive estimation of the future cycles of retail terminal inventory probability Prediction of the distribution of delivery time and delivery time required, as the production of the next stage of production goals. Prove that this kind of situation the optimal production control policy for {0, Mu} control, if the production started with the maximum productivity to production. In the process of production, using the method of Yu Fangzhen optimal production base the probability distribution of demand for raw materials and time of each period, as the raw material replenishment of raw material inventory control link basis. Numerical experiments verify the feasibility of the strategy. Through the sensitivity analysis of inventory inaccuracy degree, we found that in the control strategy, when the change of each link inventory is not accurate when, although parameters control strategies change, but all aspects of the service level remained unchanged, reflecting the good robustness of the strategy. For the research on inventory inaccuracy problem, The main contribution of this paper is: (1) the inventory replenishment problem under imprecise newsboy model from the existing extension to consider the general model of fixed ordering cost and replenishment delay, and put forward the corresponding robust reorder point and order quantity strategy; (2) the problem of production control and product inventory of WIP the exact circumstances, put forward a kind of conditional expectation and WIP inventory levels based on the robust control strategy for production. (3) study on the inventory routing problem during the distribution of finished goods inventory inaccuracy situation, considering the practical factors of transport delay can not be ignored in the distribution process, proved that the structural characteristics of the optimal inventory path the strategy, and put forward a robust inventory routing strategy. (4) studied the robust integrated control of inventory inaccuracy situation replenishment / production / distribution system, this paper presents a robust integrated replenishment / production / distribution control The feasibility and robustness of the strategy are verified by an example.

【學(xué)位授予單位】：東南大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2016
【分類(lèi)號(hào)】：F274;TP13

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 王衛(wèi)民;補(bǔ)貨應(yīng)采用拉動(dòng)而不是推動(dòng)[J];物流技術(shù);2003年08期

2 陳里 ,曲文浩,梁永剛,黃建農(nóng);永遠(yuǎn)的補(bǔ)貨[J];中國(guó)藥店;2005年09期

3 顏文;張仁頤;;論連鎖店的分店補(bǔ)貨模型[J];物流科技;2005年10期

4 劉保政;劉德寶;任蘋(píng);高立群;;易變質(zhì)商品隨機(jī)補(bǔ)貨間隔的補(bǔ)到水平存貯控制策略[J];東北大學(xué)學(xué)報(bào);2006年06期

5 熊燕華;沈厚才;;按單裝配系統(tǒng)中的協(xié)同補(bǔ)貨問(wèn)題研究[J];中國(guó)管理科學(xué);2006年06期

6 余大勇;駱建文;;具有有效期易變質(zhì)品的最優(yōu)補(bǔ)貨策略研究[J];工業(yè)工程與管理;2008年06期

7 ;光泉M化巡補(bǔ)貨系統(tǒng)的應(yīng)用[J];物流技術(shù)與應(yīng)用;2008年08期

8 申風(fēng)平;王燕;;基于時(shí)間的易變質(zhì)品整合補(bǔ)貨發(fā)貨模型[J];物流工程與管理;2011年10期

9 求實(shí);;一九八七年上半年江蘇省針紡織品補(bǔ)貨會(huì)[J];商業(yè)科技;1987年05期

10 王心意;;自動(dòng)訂貨和補(bǔ)貨的現(xiàn)實(shí)應(yīng)用[J];信息與電腦;2002年05期

相關(guān)會(huì)議論文 前4條

1 汪小京;周昌林;;有限庫(kù)存能力條件下的補(bǔ)貨及外包批量研究[A];第十二屆中國(guó)管理科學(xué)學(xué)術(shù)年會(huì)論文集[C];2010年

2 衛(wèi)武斌;孫權(quán);顏兆林;;存在快速補(bǔ)貨的軍事器材庫(kù)存控制模型及其仿真研究[A];中國(guó)運(yùn)籌學(xué)會(huì)第十屆學(xué)術(shù)交流會(huì)論文集[C];2010年

3 高遠(yuǎn)洋;李化星;;帶實(shí)時(shí)補(bǔ)貨混合策略的供應(yīng)鏈契約研究[A];第三屆（2008）中國(guó)管理學(xué)年會(huì)——市場(chǎng)營(yíng)銷(xiāo)分會(huì)場(chǎng)論文集[C];2008年

4 李宇雨;;考慮消費(fèi)者選擇和替代行為的ATO制造商定價(jià)及補(bǔ)貨策略[A];第八屆（2013）中國(guó)管理學(xué)年會(huì)——管理與決策科學(xué)分會(huì)場(chǎng)論文集[C];2013年

相關(guān)重要報(bào)紙文章 前10條

1 童繼銘;服裝行業(yè)ERP零售終端智能補(bǔ)貨 何以難達(dá)成效[N];中國(guó)計(jì)算機(jī)報(bào);2011年

2 ;生鮮訂貨補(bǔ)貨管理及原則[N];中國(guó)商報(bào);2012年

3 王心意;自動(dòng)訂貨和補(bǔ)貨的現(xiàn)場(chǎng)應(yīng)用[N];中國(guó)商報(bào);2002年

4 AMT顧問(wèn) 梁 星;智能補(bǔ)貨 消除庫(kù)存隱患[N];中國(guó)計(jì)算機(jī)報(bào);2008年

5 三材;進(jìn)貨管理的三大流程[N];醫(yī)藥經(jīng)濟(jì)報(bào);2008年

6 ;零售渠道需要“協(xié)同補(bǔ)貨”嗎？[N];電腦商報(bào);2009年

7 記者 丁寧;銀行理財(cái)收益回落 資金回補(bǔ)貨基[N];上海證券報(bào);2013年

8 本刊記者 呂萍;信息技術(shù)拓寬盈利“藍(lán)海”[N];中華合作時(shí)報(bào);2010年

9 李春波;從到貨與庫(kù)存看結(jié)構(gòu)問(wèn)題[N];中華合作時(shí)報(bào);2012年

10 胡文;藥品擺放一點(diǎn)通(上)[N];中國(guó)醫(yī)藥報(bào);2006年

相關(guān)博士學(xué)位論文 前9條

1 黃輝;基于消費(fèi)者異質(zhì)性的ATO系統(tǒng)生產(chǎn)及補(bǔ)貨策略研究[D];重慶大學(xué);2016年

2 徐浩軒;面向網(wǎng)絡(luò)零售商的批量補(bǔ)貨與發(fā)貨模型研究[D];華中科技大學(xué);2016年

3 李明;庫(kù)存不精確情形下的魯棒集成補(bǔ)貨、生產(chǎn)控制及分銷(xiāo)策略[D];東南大學(xué);2016年

4 瞿慧;基于聯(lián)合補(bǔ)貨策略的供應(yīng)鏈縱向集成優(yōu)化模型研究[D];華中科技大學(xué);2015年

5 李宇雨;ATO環(huán)境下制造商的零部件補(bǔ)貨策略研究[D];重慶大學(xué);2010年

6 于建紅;不確定條件下基于Supply-Hub的協(xié)同補(bǔ)貨策略研究[D];華中科技大學(xué);2012年

7 杜少甫;VMI集成環(huán)境下的補(bǔ)貨發(fā)貨策略研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2007年

8 汪小京;VMI與TPL集成環(huán)境下的補(bǔ)貨及發(fā)貨策略研究[D];華中科技大學(xué);2010年

9 陳芝;VMI模式下的庫(kù)存及其補(bǔ)貨模型研究[D];湖南大學(xué);2013年

相關(guān)碩士學(xué)位論文 前10條

1 袁孝濤;汽車(chē)配件多核多級(jí)協(xié)同補(bǔ)貨策略研究及系統(tǒng)實(shí)現(xiàn)[D];西南交通大學(xué);2015年

2 俞濤;A集團(tuán)公司北方配送中心規(guī)劃及流程設(shè)計(jì)[D];復(fù)旦大學(xué);2014年

3 高義;基于拉式補(bǔ)貨的快速消費(fèi)品供應(yīng)鏈分銷(xiāo)系統(tǒng)庫(kù)存控制模式研究[D];東北財(cái)經(jīng)大學(xué);2015年

4 陳卓;煙草行業(yè)工商協(xié)同補(bǔ)貨決策支持系統(tǒng)研究[D];華中科技大學(xué);2014年

5 徐瑾;L公司電腦零部件協(xié)同補(bǔ)貨策略研究[D];北京交通大學(xué);2016年

6 吳素素;某企業(yè)配送中心訂單分批及補(bǔ)貨路徑優(yōu)化[D];湖南大學(xué);2016年

7 張瑾;模塊化分揀下的需求預(yù)測(cè)與補(bǔ)貨問(wèn)題研究[D];華中科技大學(xué);2015年

8 趙陽(yáng);SH公司原材料綜合補(bǔ)貨策略研究[D];南京農(nóng)業(yè)大學(xué);2015年

9 陶順育;面向生產(chǎn)的配送中心補(bǔ)貨系統(tǒng)的研究與開(kāi)發(fā)[D];廣西大學(xué);2016年

10 曹歌;零售業(yè)配送中心的補(bǔ)貨配送優(yōu)化問(wèn)題研究[D];沈陽(yáng)工業(yè)大學(xué);2017年

，

本文編號(hào)：1693902