Multi-objective optimization of electrical discharge machining parameters using particle swarm optimization
This manuscript presents an efficient multi-objective optimization method based on using particle swarm optimization together with a desirability function that can be applied where the response variables may have an opposite behavior and where the range of variation of the independent variables as w...
| Autor: | |
|---|---|
| Formato: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2024 |
| País: | España |
| Recursos: | Universidad Pública de Navarra |
| Repositorio: | Academica-e. Repositorio Institucional de la Universidad Pública de Navarra |
| OAI Identifier: | oai:academica-e.unavarra.es:2454/47669 |
| Acesso em linha: | https://hdl.handle.net/2454/47669 |
| Access Level: | acceso abierto |
| Palavra-chave: | Multi-objective optimization Manufacturing Fuzzy modeling PSO ANFIS |
| Resumo: | This manuscript presents an efficient multi-objective optimization method based on using particle swarm optimization together with a desirability function that can be applied where the response variables may have an opposite behavior and where the range of variation of the independent variables as well as those of the responses are subjected to constraints, which has a great deal of industrial interest. For example, maintaining roughness and dimensional tolerances within a tolerance range is determined by the design requirements of the manufactured parts (shape errors, microgeometry errors, etc.) and these requirements must be met in the manufacture of parts. It is demonstrated that it is possible to obtain optimal results in the ranges of variation considered for the independent variables, with regard to those obtained by experimentation. Similarly, models based on Adaptive Network-based Fuzzy Inference Systems are used to solve the problem that may arise from the inadequate fitting of the regression models. Thus, thanks to this present study a fast and efficient method is available for the multiple-optimization of response variables, subject to constraints on both response and independent variables, which are obtained from experiments and modelled by means of soft computing techniques. Furthermore, it is also demonstrated that it is possible to obtain technology tables for various manufacturing processes, which is of great interest from a technological point of view so as to obtain the most suitable processing conditions. |
|---|