Efficient Local Level Set Algorithm Combined with Bidirectional Evolutionary Criterion Using Discrete Variables and its Appliance to Topology and Shape Optimization

Abstract

Many level set methods (LSMs) have been used for topology and shape optimization. An efficient topology optimization algorithm is proposed by combining a new bi-directional evolutionary criterion (BEC) with the local LSM (LLSM). The BEC is established using multiple discrete variables (DVs) and introduced into the bi-directional evolutionary structural optimization (BESO) method so that the DV-BESO method obtained can be applied to topology optimization. The LLSM is used instead of the global LSM due to its high efficiency. In the LLSM, the computational efficiency can be further improved by solving a distance regularized equation without the re-initialization procedure. Two benchmarks show the high efficiency of this proposed combined algorithm for topology and shape optimization.