A Co-Evolution Strategy for Optimization of Supporting Structures of Deep Foundation Pits

摘要

A co-evolution algorithm for the topology optimization of trusses was developed by the authors (Ref. [1]). This paper will extend the algorithm to optimization of supporting structures of deep foundation pits, which involves two design spaces, the scheme design space and the detail structure design space. The scheme design is to find optimal supporting types and the structure design is to optimize design parameters of detail structures relevant to specific supporting types, subject to constraints on stresses, deformation, slip -resistance and crash-resistance, etc., so that the cost of supporting structures is minimized. In this paper, two genetic algorithm processes are separately applied to both design spaces. As the two design spaces are not completely independent, a co-evolution strategy is introduced to account for the interactions between them.