Inverse modelling of electrochemical machining process using a novel combination of soft computing methods

Abstract

Selection of optimal and suitable process parameters is a crucial issue in manufacturing processes especially in electrochemical machining (ECM). Since the utmost target is to find suitable machining parameters for gaining desired machining performances, a new hybrid approach has been applied for inverse modelling of ECM process. Four machining inputs, i.e. voltage, tool feed rate, electrolyte flowrate and concentration; and two machining responses, i.e. surface roughness (Ra) and material removal rate (MRR) are presented as input variables and responses, respectively. In the proposed approach, firstly, comprehensive mathematical equations have been established based on response surface methodology (RSM). The two machining performances are modeled in this step with machining parameters. Then, the differential evolution (DE) algorithm has been used for Pareto-based multi-objective optimization. Finally, group method of data handling (GMDH)-type neural networks is used through the Pareto table for inverse modelling. As a result, four models have been developed for each of the four machining parameters; therefore, each machining parameters is determined according to the machining performance as two new design variables. The results demonstrated that the suggested method is a helpful and promising tool for inverse modelling and determining such important relationships between optimized responses and input variables.