考慮不確定性與多場(chǎng)耦合的結(jié)構(gòu)優(yōu)化

亢戰(zhàn) 白嵩

摘 要：該研究在不確定性結(jié)構(gòu)魯棒性優(yōu)化和壓電智能結(jié)構(gòu)拓?fù)鋬?yōu)化方面的研究結(jié)果。首先基于不確定性的非概率橢球凸模型描述，研究了桁架結(jié)構(gòu)的魯棒性優(yōu)化設(shè)計(jì)問(wèn)題。考慮桁架結(jié)構(gòu)彈性模量的不確定，并用非概率橢球凸模型處理不確定參數(shù)。提出了一種量化的結(jié)構(gòu)魯棒性度量方法。基于該度量模型，提出了結(jié)構(gòu)魯棒性優(yōu)化問(wèn)題的數(shù)學(xué)模型，其目標(biāo)是要達(dá)到在體積約束條件下，選出結(jié)構(gòu)中魯棒性最小的一個(gè)功能函數(shù)，使其魯棒性最大化。數(shù)值算例驗(yàn)證了優(yōu)化模型的正確性和算法的有效性。我們考慮連續(xù)體結(jié)構(gòu)載荷幅度等參數(shù)的有界不確定性，利用非概率橢球凸模型進(jìn)行不確定性參數(shù)的界限描述，研究連續(xù)體結(jié)構(gòu)的魯棒性拓?fù)鋬?yōu)化設(shè)計(jì)的建模與數(shù)值方法。為提高求解效率，利用位移與載荷的線性關(guān)系，提出一種基于解析幾何方法的魯棒性度量方法，從而避免了求解雙層優(yōu)化問(wèn)題。基于該度量方法，優(yōu)化模型的目標(biāo)是在體積分?jǐn)?shù)約束條件下尋求最優(yōu)拓?fù)湫问揭宰畲蠡Y(jié)構(gòu)的位移魯棒性。數(shù)值算例驗(yàn)證了優(yōu)化模型的正確性和算法的有效性。具有狹長(zhǎng)形狀的壓電作動(dòng)器有利于輸出較大的位移，而采用周期拼裝方式實(shí)現(xiàn)這類(lèi)結(jié)構(gòu)則具有制造成本相對(duì)較低的優(yōu)點(diǎn)。我們提出了基于周期拼裝的平面壓電作動(dòng)器結(jié)構(gòu)拓?fù)鋬?yōu)化設(shè)計(jì)的數(shù)學(xué)模型。其中，以位移輸出點(diǎn)作功最大化為設(shè)計(jì)目標(biāo)，考慮了材料體積和控制能耗約束，對(duì)結(jié)構(gòu)基體材料和壓電材料的分布以及控制電壓的分布進(jìn)行優(yōu)化設(shè)計(jì)。該文給出了結(jié)構(gòu)響應(yīng)的設(shè)計(jì)靈敏度分析，并采用基于梯度的數(shù)學(xué)規(guī)劃方法對(duì)優(yōu)化問(wèn)題進(jìn)行求解。數(shù)值算例驗(yàn)證了該文提出的數(shù)學(xué)模型和算法的可用性與有效性。

關(guān)鍵詞：結(jié)構(gòu)優(yōu)化 拓?fù)鋬?yōu)化 不確定性 多場(chǎng)耦合 智能結(jié)構(gòu)

Structural Optimization Considering Uncertainties and Multi-field Coupling

Kang Zhan Bai Song

（Dalian University of Technology）

Abstract：This report presents our recent progress on study of structural robust optimization under uncertainties and topology optimization of piezoelectric smart structures. Based on ellipsoid convex model description of uncertainties， we studied the robust design optimization problem of truss structures. In the study， the uncertainties of material properties of truss structures are considered and modeled by non-probabilistic ellipsoid convex model. A quantified measure of structural robustness was proposed， and based on this measure， the optimization formulation aims at choosing the robustness of the concerned structural behaviors with smallest robustness， and maximizing the chosen robustness under total volume constraint. Numerical example verified the validity of the proposed method. We studied the topology optimization formulation and numerical techniques of continuum structures with bounded loads on the basis of non-probabilistic ellipsoid convex model description of uncertainties. For the purpose of improving the numerical efficiency， by using the linear relationship between the displacements and the loads， a quantified measure of structural robustness is computed using analytical geometric method. Using this measure， the optimization formulation aims at finding the optimal topology layout to maximize the displacement robustness of structure under volume fraction constraint. The validity of the proposed method is verified by a numerical example. Piezoelectric actuators with large aspect ratio are suitable for deliver large displacements. Assembling this type of actuators by means of repetitive components has the advantage of low manufacturing cost. We present a mathematical model for topology optimization of planar piezoelectric actuators with repetitive components. The design objective is to maximum the work exported at the displacement output port. Constraints with regard to the control energy consumption and the material volume are imposed to the optimization problem. The distributions of the actuation voltage as well as the topologies of both host layers and piezoelectric layers are to be optimized. Numerical techniques for sensitivity analysis of structure response are presented and the proposed optimization problem is solved with a gradient-based mathematical programming approach. Illustrative examples are given to demonstrate the validity and applicability of the proposed approach.

Key Words：Structural optimization； Topology optimization； Uncertainty； Multi-field coupling； Smart structure

閱讀全文鏈接（需實(shí)名注冊(cè)）：http：//www.nstrs.cn/xiangxiBG.aspx？id=50248&flag=1