Micro-EDM optimization through particle swarm algorithm and artificial neural network

• Published in: Precision engineering Vol. 73; pp. 63 - 70
• Main Authors: Quarto, Mariangela, D'Urso, Gianluca, Giardini, Claudio
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.01.2022
• Subjects: Artificial neural network; Micro-EDM; Particle swarm optimization algorithm; Micro-EDM; Particle swarm optimization algorithm; Artificial neural network
• ISSN: 0141-6359; 1873-2372
• DOI: 10.1016/j.precisioneng.2021.08.018

In the present study, an artificial neural network (ANN) together with a heuristic algorithm, called particle swarm optimization (PSO), was used to set up a methodology for selecting the optimal process parameters for the μEDM process. The developed methodology is characterized by a double direction functionality responding to different industry needs. Usually, in the industrial scenario, the operators are bound by the project specifications or by the limited availability of time. For this reason, a methodology tested only on a specific workpiece material, that involves limited input parameters or developed for the optimization of a single performance is limiting. The developed 2-steps model leaves operators free to establish which factors to impose for the optimization and allows to define the best solution for the production of a part. The validation of the model shows a good fit between predicted and experimental results. •Development of a bidirectional methodology for optimizing the process parameters as a function of desired results in term of perfomances.•Combination of Artificial Neural Network and Particle Swarm Optimization techniques•Optimization methodology imposing some inputs based on design specification.•Optimization considering external constraints related to process performance.•The developed methodology is based on the construction of Neural Network, its training and validatation and then its implementation through the PSO algorithm for definition of optimal results.