【摘要】：市場競爭的持續(xù)加劇與環(huán)境污染的日趨惡化,對生產(chǎn)過程的能效水平提出了更高的要求。尤其是多品種小批量的個性化生產(chǎn)模式,進一步提高了能耗控制問題的復(fù)雜性。汽車零部件的涂裝生產(chǎn)過程會伴隨著大量的能量消耗,它已經(jīng)成為制造業(yè)節(jié)能降耗的重要研究方向。因此,本文以汽車涂裝為例,研究大規(guī)模定制生產(chǎn)模式下的涂裝計劃優(yōu)化排程方法,設(shè)計開發(fā)了多品種混合涂裝計劃排程優(yōu)化與能耗在線監(jiān)測系統(tǒng),并通過在線監(jiān)測獲得工藝參數(shù)與能耗數(shù)據(jù),為提高汽車涂裝生產(chǎn)質(zhì)量與能效水平提供了有力保障。本文分析了涂裝生產(chǎn)過程中產(chǎn)生無效能耗的原因,通過優(yōu)化生產(chǎn)訂單的排產(chǎn)序列來降低涂裝過程產(chǎn)生的無效能耗。首先以混合品種產(chǎn)品集的無效能耗最低、涂裝生產(chǎn)時間最短與涂裝成本最小為目標,建立了涂裝線多目標優(yōu)化模型。然后基于多目標非支配快速排序遺傳算法與改進粒子群算法對多目標優(yōu)化模型分別進行了優(yōu)化求解。在傳統(tǒng)遺傳算法的基礎(chǔ)上,通過快速簡易的前向比較操作對染色體種群進行非支配前沿等級的劃分,克服傳統(tǒng)排序方式分層速度過慢的缺點;然后采用小生境技術(shù)中的擁擠距離對同一非支配層的染色體進行排序,保持種群多樣性;最后將求得的Pareto解集通過層次分析法選出最優(yōu)排產(chǎn)序列。同時對比兩種算法的求解結(jié)果與求解效率,最終將改進的遺傳算法應(yīng)用于多品種混合涂裝計劃排程優(yōu)化與能耗在線監(jiān)測系統(tǒng)中的排程優(yōu)化模塊。利用C#開發(fā)語言、MATLAB與SQL Server數(shù)據(jù)庫,開發(fā)實現(xiàn)了多品種混合涂裝計劃排程優(yōu)化與能耗在線監(jiān)測系統(tǒng),主要功能包括系統(tǒng)管理、計劃排程、報表查看、能耗統(tǒng)計、質(zhì)量分析、實時監(jiān)控六大功能模塊。系統(tǒng)管理實現(xiàn)了基礎(chǔ)數(shù)據(jù)的錄入以及人員/用戶/角色權(quán)限的定義與設(shè)置;計劃排程是整個系統(tǒng)的核心,依據(jù)給定的訂單內(nèi)容,通過多目標非支配排序遺傳算法,實現(xiàn)了涂裝生產(chǎn)的排產(chǎn)優(yōu)化;報表查看實現(xiàn)了對涂裝生產(chǎn)線各個工序工藝參數(shù)的存儲、查詢與分析;能耗統(tǒng)計實現(xiàn)了對涂裝車間日生產(chǎn)/各工件生產(chǎn)的能耗統(tǒng)計;質(zhì)量分析對各工件漆膜厚度與漆膜光澤度數(shù)據(jù)進行了平均值與方差的計算與分析;實時監(jiān)控實現(xiàn)了對整條涂裝生產(chǎn)線的動態(tài)實時監(jiān)測。通過在山東某汽車涂裝生產(chǎn)車間的實際應(yīng)用,驗證了該系統(tǒng)的可行性與有效性。
[Abstract]:The continuous intensification of market competition and the deterioration of environmental pollution put forward higher requirements for the energy efficiency level of the production process. Especially, the individualized production mode of multi-variety and small batch further improves the complexity of energy consumption control problem. The painting process of automobile parts will be accompanied by a lot of energy consumption, which has become an important research direction of energy saving and consumption reduction in manufacturing industry. Therefore, taking automobile painting as an example, this paper studies the optimization scheduling method of painting planning under mass customization production mode, and designs and develops an on-line monitoring system for multi-variety mixed coating planning scheduling and energy consumption. The process parameters and energy consumption data are obtained by on-line monitoring, which provides a powerful guarantee for improving the quality and energy efficiency of automobile coating production. In this paper, the causes of invalid energy consumption in coating production process are analyzed, and the invalid energy consumption in painting process is reduced by optimizing the production sequence of production orders. Firstly, aiming at the lowest invalid energy consumption, the shortest coating production time and the minimum coating cost, the multi-objective optimization model of coating line is established. Then the multi-objective optimization model is optimized based on the multi-objective non-dominated fast sorting genetic algorithm and the improved particle swarm optimization algorithm. On the basis of traditional genetic algorithm, the non-dominant frontier grade of chromosome population is divided by fast and simple forward comparison operation, and the disadvantage of slow stratification speed of traditional sorting method is overcome. Then the crowded distance in niche technique is used to sort the chromosomes of the same non-dominant layer to maintain the diversity of the population. Finally, the obtained Pareto solution set is selected by analytic hierarchy process (AHP) to select the optimal scheduling sequence. At the same time, the solution results and efficiency of the two algorithms are compared, and finally, the improved genetic algorithm is applied to the scheduling optimization module of multi-variety mixed coating planning and energy consumption online monitoring system. Using C # development language, MATLAB and SQL Server database, a multi-variety mixed coating planning scheduling optimization and energy consumption online monitoring system is developed and realized. The main functions include system management, planning scheduling, report viewing, energy consumption statistics, quality analysis. Real-time monitoring of six functional modules. The system management realizes the input of basic data and the definition and setting of personnel / user / role permissions. Planning and scheduling is the core of the whole system. According to the given order content, the multi-objective non-dominant sorting genetic algorithm is used to optimize the production scheduling of painting production. The report review realizes the storage, query and analysis of the process parameters of each process of the painting production line, and the energy consumption statistics realizes the energy consumption statistics of the daily production of the painting workshop / the production of each workpiece. The average value and variance of the film thickness and gloss data of each workpiece are calculated and analyzed by quality analysis, and the dynamic real-time monitoring of the whole painting production line is realized by real-time monitoring. The feasibility and effectiveness of the system are verified by the practical application in an automobile painting workshop in Shandong Province.
【學(xué)位授予單位】：山東大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：U466;TP274

【參考文獻】

相關(guān)期刊論文 前10條

1 李玉霞;曹華軍;李洪丞;陶桂寶;;作業(yè)車間碳排放動態(tài)特性及二階優(yōu)化調(diào)度模型[J];計算機集成制造系統(tǒng);2015年10期

2 陳鴻海;蔣增強;左樂;張彥如;;基于近親變異NSGA-Ⅱ算法的多目標柔性作業(yè)車間調(diào)度[J];農(nóng)業(yè)機械學(xué)報;2015年04期

3 蔣增強;左樂;;低碳策略下的多目標柔性作業(yè)車間調(diào)度[J];計算機集成制造系統(tǒng);2015年04期

4 劉祚時;馬力;羅金平;;混合流水車間多目標調(diào)度的小生境粒子群算法[J];機械設(shè)計與制造;2015年04期

5 劉海博;劉檢華;何永熹;郭崇穎;蔣科;;基于小生境粒子群算法的公差多目標優(yōu)化設(shè)計[J];計算機集成制造系統(tǒng);2015年03期

6 趙詩奎;方水良;顧新建;;作業(yè)車間調(diào)度的空閑時間鄰域搜索遺傳算法[J];計算機集成制造系統(tǒng);2014年08期

7 韓煜東;董雙飛;譚柏川;;基于改進遺傳算法的混裝線多目標優(yōu)化[J];計算機集成制造系統(tǒng);2015年06期

8 詹騰;張屹;朱大林;劉錚;鄭小東;;基于多策略差分進化的元胞多目標遺傳算法[J];計算機集成制造系統(tǒng);2014年06期

9 劉曉冰;焦璇;寧濤;梁旭;;基于雙鏈量子遺傳算法的柔性作業(yè)車間調(diào)度[J];計算機集成制造系統(tǒng);2015年02期

10 楊海東;鄭慶仁;劉國勝;郭建華;;基于差分遺傳算法的置換流水車間低碳調(diào)度模型[J];中南大學(xué)學(xué)報(自然科學(xué)版);2013年11期

相關(guān)博士學(xué)位論文 前1條

1 張靜;基于混合離散粒子群算法的柔性作業(yè)車間調(diào)度問題研究[D];浙江工業(yè)大學(xué);2014年

相關(guān)碩士學(xué)位論文 前2條

1 韓會龍;基于遺傳算法的ERP和柔性作業(yè)車間調(diào)度系統(tǒng)的研究與開發(fā)[D];華北電力大學(xué);2014年

2 高一;基于混合遺傳算法的作業(yè)車間多目標優(yōu)化調(diào)度及系統(tǒng)開發(fā)研究[D];東華大學(xué);2014年

，

本文編號：2480873