OPTIMIZATION OF ELECTRO-CHEMICAL DISCHARGE MACHINING PROCESS USING GENETIC ALGORITHM

Abstract

Machining process efficiency can be improved by optimizing the control parameters. This requires identifying and determining the value of critical process control parameters that lead to desired responses ensuring a lower cost of manufacturing. Traditional optimization methods are generally slow in convergence and require much computing time. Also, they may risk being trapped at local optima. Compared to these, Genetic Algorithm (GA) is robust, global and can be applied without recourse to domain-specific heuristics. Considering these novel traits of GA, optimization of Electro-Chemical Discharge Machining (ECDM) process has been carried out in this research. Mathematical models using Response Surface Methodology (RSM) is used to correlate the responses and the control parameters. The desired responses are minimum radial overcut (ROC) and minimum heat affected zone (HAZ) while the control parameters are applied voltage, electrolyte concentration and interelectrode gap. The responses are optimized individually as well as a simultaneous multiobjective optimization is solved. A Pareto optimal solution is the output result for multi objective optimization. Here each solution is non dominated among the group of predicted solution points thus allowing flexibility in operating the machine while maintaining the standard quality.