Customized Evolutionary Optimization of Compliant Mechanism to Generate Multiple Light Weight Topologies
Abstract
The present work focuses on evolving the multiple light-in-weight topologies of compliant mechanism tracing user defined path. Therefore in this paper, the bi-objective set is formulated first on the optimization frame-work in which the helper objective of maximum diversity is introduced with the primary objective of minimum weight of elastic structures. Thereafter, the evolutionary algorithm (NSGA-II) is customized to efficiently deal with the constraint bi-objective, non-linear and discrete problem of compliant mechanisms. The existing NSGA-II algorithm is modified with various aspects and schemes such as, domain specific population strategy, domain specific crossover, parallel computing, minimumweight local search method etc. The flexibility of identifying the applied and boundary conditions of elastic structures are also coupled with the customized NSGA-II algorithm to promote non-dominated solutions. Two examples of compliant mechanisms tracing (i) curvilinear path and (ii) straight line path are solved and their light-in-weight topologies are presented.
Authors: Deepak Sharma, Kalyanmoy Deb, N. N. Kishore