【摘要】：信息技術(shù)的發(fā)展推動著企業(yè)生產(chǎn)方式的變革。越來越多的中小離散制造企業(yè)希望應(yīng)用信息化手段來管理車間生產(chǎn)現(xiàn)場,實現(xiàn)提高企業(yè)生產(chǎn)效率、降低生產(chǎn)成本和提高產(chǎn)品質(zhì)量的目標(biāo)。然而企業(yè)信息化深入到車間生產(chǎn)管理時,經(jīng)常會遇到很多問題。論文以構(gòu)建車間生產(chǎn)現(xiàn)場駕駛艙控制系統(tǒng)為目標(biāo),使得管理者可以不必親臨生產(chǎn)現(xiàn)場,通過駕駛艙就能獲取車間現(xiàn)場的實時狀態(tài),并可以控制生產(chǎn)過程的運(yùn)行。 構(gòu)建駕駛艙控制系統(tǒng)需要解決三個關(guān)鍵技術(shù)：車間現(xiàn)場生產(chǎn)狀態(tài)的建模、車間生產(chǎn)狀態(tài)的偏差分析以及車間動態(tài)生產(chǎn)調(diào)度。這三個關(guān)鍵技術(shù)也正是車間生產(chǎn)管理信息化所面臨的問題。 車間現(xiàn)場生產(chǎn)狀態(tài)的模型是研究其他關(guān)鍵技術(shù)的基礎(chǔ)。車間生產(chǎn)現(xiàn)場的基本構(gòu)成要素是工件和工作單元,兩者通過工序聯(lián)系在一起。三者的結(jié)合體抽象成信息節(jié)點,分析了信息節(jié)點之間的關(guān)系,并對信息節(jié)點的屬性和行為作了描述。以信息節(jié)點為基本單元,描述了車間現(xiàn)場的生產(chǎn)狀態(tài)。分析了車間現(xiàn)場的生產(chǎn)數(shù)據(jù)與工件狀態(tài)、工序狀態(tài)和工作單元的狀態(tài)的關(guān)系RSD,以及信息節(jié)點狀態(tài)與工件狀態(tài)、工序狀態(tài)和工作單元的狀態(tài)的關(guān)系RSI,建立了車間現(xiàn)場生產(chǎn)數(shù)據(jù)與信息節(jié)點狀態(tài)的對應(yīng)關(guān)系,建立了車間現(xiàn)場生產(chǎn)狀態(tài)的模型。應(yīng)用信息節(jié)點的模型,簡化了車間現(xiàn)場生產(chǎn)狀態(tài)的建模。 車間現(xiàn)場理想的生產(chǎn)狀態(tài)由基準(zhǔn)生產(chǎn)過程確定,基準(zhǔn)生產(chǎn)過程由作業(yè)計劃調(diào)度確定。實際的生產(chǎn)過程按照基準(zhǔn)生產(chǎn)過程確定的節(jié)點激活順序運(yùn)行,并期望實際生產(chǎn)過程與基準(zhǔn)生產(chǎn)過程保持一致。但由于車間生產(chǎn)現(xiàn)場存在大量的隨機(jī)因素,使得實際生產(chǎn)過程經(jīng)常偏離基準(zhǔn)生產(chǎn)過程,形成實際生產(chǎn)狀態(tài)與理想生產(chǎn)狀態(tài)的偏差e(t)。只有同質(zhì)才能比較,因此把偏差e(t)細(xì)化成生產(chǎn)時間偏差和生產(chǎn)成本偏差等,并給出了計算方法。通過引入信息節(jié)點關(guān)聯(lián)模塊的概念,分析了偏差e(t)在車間生產(chǎn)過程中傳播的機(jī)理,這是分析預(yù)估偏差的基礎(chǔ)。 實際的車間生產(chǎn)過程是動態(tài)變化的,因此需要不斷的重調(diào)度才能滿足實際生產(chǎn)的需求。動態(tài)生產(chǎn)調(diào)度以提高企業(yè)生產(chǎn)效率、降低生產(chǎn)成本和提高產(chǎn)品質(zhì)量為目標(biāo),是多目標(biāo)優(yōu)化問題。為了把多目標(biāo)優(yōu)化問題轉(zhuǎn)換成單目標(biāo)優(yōu)化問題,提出了當(dāng)量成本的概念。通過分析時間、成本和質(zhì)量的關(guān)系,把時間、成本和質(zhì)量轉(zhuǎn)換成當(dāng)量成本,并以當(dāng)量成本最小作為調(diào)度問題的目標(biāo)函數(shù),把生產(chǎn)狀態(tài)的偏差不超過設(shè)定的閾值作為約束條件。根據(jù)車間現(xiàn)場的生產(chǎn)狀態(tài),分析了幾種觸發(fā)動態(tài)生產(chǎn)調(diào)度的典型情況,針對這些情況的特點,給出合適的調(diào)度策略和調(diào)度算法,實現(xiàn)當(dāng)量成本最小的調(diào)度目標(biāo)。動態(tài)生產(chǎn)調(diào)度的實時性要求調(diào)度算法必須高效,論文結(jié)合啟發(fā)式規(guī)則和鄰域搜索的高效性,以時間鄰域內(nèi)的最優(yōu)解作為啟發(fā)式搜索方向,提出了信息節(jié)點分段尋優(yōu)算法。 在此研究的基礎(chǔ)上,對駕駛艙控制系統(tǒng)的體系結(jié)構(gòu)進(jìn)行了研究,并結(jié)合實證案例對論文的理論和方法進(jìn)行了闡述。
[Abstract]:More and more small and medium-sized discrete manufacturing enterprises hope to use information technology to manage the workshop production site to achieve the goal of improving production efficiency, reducing production costs and improving product quality. Many problems. The aim of this paper is to construct the cockpit control system of workshop production site, so that managers can obtain the real-time state of workshop site and control the operation of production process without visiting the production site.
The construction of cockpit control system needs to solve three key technologies: the modeling of workshop production status, the deviation analysis of workshop production status and the dynamic production scheduling.
The model of workshop production status is the basis of other key technologies.The basic elements of workshop production site are workpiece and work unit,which are connected by working procedure.The combination of them is abstracted into information nodes,the relationship between information nodes is analyzed,and the attributes and behaviors of information nodes are described. The information node is the basic unit, which describes the production state of the workshop site. The relationship between the production data and the workpiece state, the working procedure state and the working unit state is analyzed, and the relationship between the information node state and the workpiece state, the working procedure state and the working unit state is analyzed. According to the corresponding relation of node state, the model of workshop production state is established. The model of information node is applied to simplify the modeling of workshop production state.
The ideal production state of the workshop site is determined by the benchmark production process and the benchmark production process is determined by the job scheduling. The actual production process runs in the order of node activation determined by the benchmark production process, and the actual production process is expected to be consistent with the benchmark production process. Only homogeneity can be compared, so the deviation e (t) is refined into production time deviation and production cost deviation, and the calculation method is given. By introducing the concept of information node correlation module, the deviation is analyzed. The propagation mechanism of differential e (T) in the workshop production process is the basis for analyzing the prediction bias.
The actual workshop production process is dynamic, so it needs constant rescheduling to meet the actual production needs. Dynamic production scheduling is a multi-objective optimization problem with the objective of improving production efficiency, reducing production costs and improving product quality. By analyzing the relationship among time, cost and quality, the time, cost and quality are converted into equivalent cost, and the minimum equivalent cost is taken as the objective function of scheduling problem, and the deviation of production state is not exceeded the set threshold as the constraint condition. According to the characteristics of these situations, a suitable scheduling strategy and algorithm are proposed to achieve the goal of minimizing the equivalent cost. The real-time performance of dynamic production scheduling requires that the scheduling algorithm must be efficient. A search algorithm for information nodes is proposed.
On the basis of this research, the system structure of the cockpit control system is studied, and the theory and method of the paper are expounded with an empirical case.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2013
【分類號】：TH186

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 林宋;制造系統(tǒng)的Petri網(wǎng)建模[J];北方工業(yè)大學(xué)學(xué)報;2002年03期

2 張倫彥;楊建軍;郭仲福;;基于仿真的單元化制造系統(tǒng)層次建模技術(shù)研究[J];成組技術(shù)與生產(chǎn)現(xiàn)代化;2006年01期

3 田乃碩;多服務(wù)臺休假排隊系統(tǒng)的進(jìn)展[J];燕山大學(xué)學(xué)報;1998年01期

4 胡峰,溫熙森,孫國基;離散事件動態(tài)系統(tǒng)理論及應(yīng)用[J];電腦與信息技術(shù);2001年05期

5 任子武;傘冶;;實數(shù)遺傳算法的改進(jìn)及性能研究[J];電子學(xué)報;2007年02期

6 李歧強(qiáng);具有約束指導(dǎo)的模擬退火算法[J];系統(tǒng)工程;2001年03期

7 段波;趙穩(wěn)莊;仉樹軍;;Petri網(wǎng)在制造系統(tǒng)建模與仿真中的應(yīng)用[J];航空精密制造技術(shù);2008年05期

8 彭昱;陳蔚芳;;在制品管理中條碼技術(shù)的應(yīng)用[J];航空制造技術(shù);2007年02期

9 王丹;張浩;陸劍峰;;基于瓶頸約束的APS算法研究與應(yīng)用[J];機(jī)電一體化;2007年06期

10 楊衛(wèi)波;趙燕偉;;求解TSP問題的改進(jìn)模擬退火算法[J];計算機(jī)工程與應(yīng)用;2010年15期

相關(guān)博士學(xué)位論文 前1條

1 王志國;面向車間管理的離散生產(chǎn)系統(tǒng)精益運(yùn)行駕駛艙研究[D];浙江大學(xué);2009年

，

本文編號：2246641