Optimization of electrochemical machining process parameters using teaching-learning-based algorithm

Abstract

Electrochemical machining (ECM) process has a wide capability to generate complex shapes on different materials which are occasionally difficult to cut. Its ability to machine a variety of materials makes it an extensively accepted non-traditional machining process in modern day manufacturing sector. Thus, selection of the optimal input parameters for an ECM process is crucial for its efficient utilization. In this paper, a comparative analysis is made among four metaheuristics, i.e. firefly algorithm (FA), differential evolution (DE), ant colony optimization (ACO) algorithm and teaching-learning-based optimization (TLBO) algorithm to discover the optimal values of various control parameters for an ECM process. Dimensional inaccuracy, tool life and material removal rate are the three responses considered which are subjected to temperature, choking and passivity constraints. The TLBO algorithm shows the best performance among the others without violating any of the constraints. The paired t-test is also performed to prove the efficacy of TLBO algorithm over the other optimization techniques. The results derived from these algorithms are finally compared with those obtained by the past researchers using other optimization methods for both single and multi-objective optimization problems.