Predict the modelling of electro chemical machining parameters for AA5083/MoS2 composites using Levenberg–Marquardt algorithm

Abstract

ECM is widely regarded as a highly promising and cost-effective manufacturing technique, especially for processing hard-to-machine materials that are challenging to shape using conventional methods. The machining operations were carried out using an ECM machine with a working voltage range of 0.6 to 1.0 V and a feed rate between 15 and 25 mm/min. A copper electrode was employed alongside an NaCl electrolyte solution for calculating material removal rate on AA5083/MoS2 composites. The Highest MRR is observed when voltage 1.0 V, feed rate 25 mm/min and Electrolyte Concentration 400 g/Lit. To improve the accuracy of the predicted output responses, an artificial neural network (ANN) model was designed using the Levenberg-Marquardt algorithm. The structure with a configuration of 3–10–1, confirmed strong regression fit outcomes, The overall correlation coefficients (R) calculated at 0.96348, confirmed a high level of consistency between the experimental data and the predicted value.

Keywords

• ecm process
• material removal rate
• levenberg-marquardt algorithm
• artificial neural network and modelling