【摘要】：結(jié)構(gòu)優(yōu)化設(shè)計(jì)已經(jīng)成為工業(yè)界和學(xué)術(shù)界的研究熱點(diǎn)，而研究重點(diǎn)主要在于如何提高設(shè)計(jì)的效率和精度。一種較新的模仿蜜蜂采蜜行為的元啟發(fā)式優(yōu)化算法--蜂群算法，被應(yīng)用于解決機(jī)械結(jié)構(gòu)件的優(yōu)化設(shè)計(jì)問題。先由試驗(yàn)設(shè)計(jì)和序列響應(yīng)面法構(gòu)建目標(biāo)函數(shù)及約束條件的代理模型，再應(yīng)用改進(jìn)蜂群算法求解最優(yōu)設(shè)計(jì)。在優(yōu)化過程中調(diào)用的是滿足精度要求的代理模型，大大減少了有限元計(jì)算次數(shù)，提高了優(yōu)化效率。選取典型實(shí)例對該方法進(jìn)行驗(yàn)證，預(yù)期值與實(shí)際值相比較的結(jié)果表明該方法具備了足夠的求解精度，能夠滿足工程實(shí)際要求。 穩(wěn)健優(yōu)化設(shè)計(jì)可以減小產(chǎn)品的質(zhì)量波動(dòng)，增強(qiáng)設(shè)計(jì)的健壯性。研究了區(qū)間型不確定因素的性質(zhì)，提出了代理模型方法、區(qū)間分析和改進(jìn)蜂群算法相結(jié)合的穩(wěn)健優(yōu)化設(shè)計(jì)方法。選取典型實(shí)例對上述方法進(jìn)行驗(yàn)證，通過將穩(wěn)健優(yōu)化結(jié)果與結(jié)構(gòu)優(yōu)化設(shè)計(jì)結(jié)果相比較可知，雖然犧牲了產(chǎn)品的一部分性能指標(biāo)，但增加了質(zhì)量指標(biāo)的穩(wěn)健性和設(shè)計(jì)變量的可靠度，且該穩(wěn)健優(yōu)化設(shè)計(jì)方法得到的預(yù)期值和實(shí)際值的誤差控制在一定的容差范圍之內(nèi)。 針對結(jié)構(gòu)優(yōu)化設(shè)計(jì)和穩(wěn)健優(yōu)化設(shè)計(jì)的效率問題，，提出了基于有限元分析過程參數(shù)化的優(yōu)化技術(shù)。利用Open CASCADE幾何造型平臺(tái)開發(fā)參數(shù)化有限元軟件平臺(tái)，實(shí)現(xiàn)了幾何模型參數(shù)化、網(wǎng)格模型參數(shù)化和屬性數(shù)據(jù)參數(shù)化等功能，能夠一次建模完成多次有限元計(jì)算，并提取所需的有限元分析結(jié)果作為樣本數(shù)據(jù)點(diǎn)構(gòu)建代理模型，將得到的代理模型通過不同的數(shù)學(xué)處理分別應(yīng)用到結(jié)構(gòu)優(yōu)化設(shè)計(jì)和穩(wěn)健優(yōu)化設(shè)計(jì)中。通過實(shí)例證明此方法具有良好的運(yùn)行效率和運(yùn)行效果，對實(shí)際工程問題的優(yōu)化設(shè)計(jì)有很大的幫助。
[Abstract]:Structural optimization design has become a hot topic in industry and academia, and the research focus is mainly on how to improve the efficiency and precision of design. A new meta-heuristic optimization algorithm, which simulates honeybee honey gathering behavior, is applied to the optimization design of mechanical structures. The agent model of objective function and constraint condition is constructed from experimental design and sequential response surface method, and then the improved bee colony algorithm is applied to solve the optimal design. In the process of optimization, the proxy model that meets the precision requirement is called, which greatly reduces the times of finite element calculation and improves the efficiency of optimization. A typical example is selected to verify the method. The comparison between the expected value and the actual value shows that the method has enough accuracy to meet the practical engineering requirements. Robust optimization design can reduce product quality fluctuation and enhance design robustness. In this paper, the properties of interval type uncertainties are studied, and the agent model method, the robust optimal design method based on interval analysis and improved bee colony algorithm are proposed. A typical example is selected to verify the above method. By comparing the robust optimization result with the structural optimization design result, although some performance indexes of the product are sacrificed, However, the robustness of the quality index and the reliability of the design variables are increased, and the error between the expected value and the actual value obtained by the robust optimization design method is controlled within a certain tolerance range. Aiming at the efficiency of structural optimization design and robust optimization design, a parameterized optimization technique based on finite element analysis (FEM) is proposed. The parameterized finite element software platform is developed by using Open CASCADE geometric modeling platform. The functions of parameterization of geometric model, parameterization of mesh model and parameterization of attribute data are realized. The required finite element analysis results are extracted as sample data points to construct the agent model. The agent model is applied to structural optimization design and robust optimization design by different mathematical processing. Examples show that this method has good running efficiency and effect, and it is very helpful to the optimization design of practical engineering problems.
【學(xué)位授予單位】：南京航空航天大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2012
【分類號】：TH122

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 錢令希,鐘萬勰,隋允康,張近東;多單元、多工況、多約束的結(jié)構(gòu)優(yōu)化設(shè)計(jì)──DDDU程序系統(tǒng)[J];大連工學(xué)院學(xué)報(bào);1980年04期

2 隋允康,鐘萬勰,錢令希;桿-膜-梁組合結(jié)構(gòu)優(yōu)化的DDDU-2程序系統(tǒng)[J];大連工學(xué)院學(xué)報(bào);1983年01期

3 邱志平,顧元憲;不確定凸模型近似算法的一種改進(jìn)[J];大連理工大學(xué)學(xué)報(bào);1997年06期

4 王紅濤;竺曉程;杜朝輝;;基于Kriging代理模型的改進(jìn)EGO算法研究[J];工程設(shè)計(jì)學(xué)報(bào);2009年04期

5 陳立周，翁海珊，任婷婷;基于隨機(jī)模型的產(chǎn)品質(zhì)量設(shè)計(jì)原理與方法[J];工程設(shè)計(jì);1995年04期

6 馬義中,徐濟(jì)超;多指標(biāo)穩(wěn)健設(shè)計(jì)質(zhì)量特性的度量[J];系統(tǒng)工程;1998年06期

7 錢令希,鐘萬勰,程耿東,隋允康;工程結(jié)構(gòu)優(yōu)化的序列二次規(guī)劃[J];固體力學(xué)學(xué)報(bào);1983年04期

8 郭海丁,路志峰;基于BP神經(jīng)網(wǎng)絡(luò)和遺傳算法的結(jié)構(gòu)優(yōu)化設(shè)計(jì)[J];航空動(dòng)力學(xué)報(bào);2003年02期

9 汪樹玉,劉國華,包志仁;結(jié)構(gòu)優(yōu)化設(shè)計(jì)的現(xiàn)狀與進(jìn)展(續(xù))[J];基建優(yōu)化;1999年05期

10 汪樹玉，劉國華，包志仁;結(jié)構(gòu)優(yōu)化設(shè)計(jì)的現(xiàn)狀與進(jìn)展[J];基建優(yōu)化;1999年04期

相關(guān)博士學(xué)位論文 前1條

1 李恩穎;近似模型優(yōu)化體系關(guān)鍵技術(shù)研究及應(yīng)用[D];湖南大學(xué);2009年

相關(guān)碩士學(xué)位論文 前1條

1 隋曉峰;基于幾何造型的參數(shù)化有限元建模和曲面網(wǎng)格生成方法及實(shí)現(xiàn)[D];大連理工大學(xué);2002年

本文編號：2192149