Abstract
Most optimization problems rely heavily on simulations to evaluate design solutions according to a set of design criteria under stringent constraints. While an inverse problem may also rely on simulation-driven methods, it can sometimes be viewed more properly as a constrained optimization problem because the main aim of inversion is to find the best parameter estimates so as to minimize the differences between predicted results and observations. In this paper, we will take a unified approach to inversion and optimization. We use the latest cuckoo search to solve inverse problems and shape optimization in heat transfer applications. Simulation results show that cuckoo search is also very efficient for inversion.