报告题目：Optimization driven designs for engineering applications

报告人：Dr. Dianzi Liu

单位：University of East Anglia (UEA)，UK

时间：2016年7月19日14:00-15:30

地点：综合实验1号楼525

Report Abstract:

Dr. Dianzi Liufirstly will introduce the bi-level optimization strategy for large-scale laminated composite structural designs subject to various manufacturing constraints, especially the blending requirement. At the top level optimization, both the smear technique and lamination parameters approach will be introduced and then applied to seek the optimal solutions in terms of laminate thicknesses and lamination parameters, respectively. At the bottom level optimization, stacking sequence optimization will be performed to match the computed lamination parameters at the bottom level with the ones passed from the top level as well as satisfy the blending requirement and layup rules. Using the bi-level strategy, the optimal stacking sequenceofthe laminated composite structure can be obtained with less computational cost.

The other basic ingredient of the talk will bemetamodel techniquefor engineering design optimization. The mathematical explicit expression for the response of the interest built by Genetic Programming (GP) is the main feature of this technique. Benefiting from theOptimal Latin Hypercube (OLH) sampling approach, the metamodel will be generated usingthe efficient training data and further the optimal solution will be obtained. The potential of metamodel based optimization technique will be demonstrated with a composite fuselage design. At last, the Multipoint Approximation Method (MAM) will be show and demonstrate its capability to efficiently solve themedium-scale, mixed continuous-discrete engineering optimization problems via the benchmark example of the ten-bar truss structure.

Dr. Liu’s Brief Introduction：

Dr. Dianzi Liu is a Lecturer in the Engineering Division at the University of East Anglia (UEA), UKand obtained his PhD in Mechanical Engineering from the University of Leeds in 2010.His PhD research topic was: Bi-level optimization of composite aircraft wing panels subject to manufacturing constraints. He has more than 10 years of experience in structural analysis, composites modelling, numerical computation and design optimization. More recently, his research has been focused on three themes: 1) Developing efficient algorithms for solving mixed continuous-discrete engineering optimization problems; 2) Designing energy harvesting devices for scavenging the bio-kinetic energy; and 3) Improving damage tolerance of laminated composite panels subject to low-speed impact by stochastic optimization. He is the author of more than 20 technical papers since 2006 and is the recipient of the runner-up prize in the ISSMO-Springer Prize competition in 2009.