Effect of jute fiber length on drilling performance of biocomposites: optimization comparison between RSM, ANN, and genetic algorithm

This current research is a continuation of previous publication by authors on the study of the delamination factor (F-d) of jute/polyester biocomposites based on the response surface methodology (RSM). However, in this study (part II), ANN statistical data analysis on the drilling capacity of a biocomposite according to the drilling parameters mainly by changing the length of the fibers is reported. The interaction of machining factors is studied using 3D surface plots, whereas optimal parameters of the process are predicted with the desirability/RSM and ANN/ genetic algorithm (GA) curve. ANN is used to obtain the best-trained fitness model for (GA) of the biocomposite material reinforced with jute fibers. RSM and ANN models showed good agreement with experimentally obtained results. However, ANN models were found to be more accurate than the RSM due to the higher correlation of ANN coefficients than RSM coefficients during training, validation, and testing.