Optimization of Incremental Sheet Forming Process Using Artificial Intelligence-Based Techniques

Abstract

Incremental sheet forming (ISF) technology has evolved as an emerging arena of manufacturing during the early 1990s. The development of different variants of this die-less technique of forming has fascinated researchers and scientists to fabricate various realistic accurate objects of various materials. The cultivation of an artificial intelligence (AI) model, capable of investigating the effects of input parameters on the performance of ISF process, is gaining the attention of researchers as an alternative to conventional techniques of predicting and optimizing the response characteristics. The AI model can be built using the input–output data using artificial neural networks (ANN). The ANN model can be applied to optimizing the performance and quality by choosing the appropriate input parameters. In addition, the ANN model can be trained using a data sample having inputs and the corresponding outputs. Therefore, ANN-based techniques need to be executed and explored for this process to enhance the employability and suitability of ISF process to the mainstream of manufacturing industries. This works aims at investigating and predicting the forming force in ISF process using AI techniques. A hybrid learning-based ANN (HLANN) model is proposed to predict the forming force in SPIF. The results of the HLANN model are compared with the experimental results and the results of two state-of-the-art regression models SVM and GPR. The comparison shows that the HLANN model outperforms the regression methods (SVM and GPR). The accuracy of results predicted using HLANN model was found to be 96.90%. The prediction and measurement of forming forces during SPIF process would help in determining the size of forming machinery and additional hardware along with preventing the failures of facilities.