Modelling and optimization of microhardness of electroless Ni–P–TiO2 composite coating based on machine learning approaches and RSM

Abstract

In this study, experimental investigations on the microhardness of the synthesized electroless Ni–P–TiO2 coated aluminium composite was carried out. The coated samples were characterized by scanning electron microscopy (SEM) for surface morphology and X-ray diffraction (XRD) pattern for phase recognition. The microhardness of the electroless Ni–P–TiO2 coated composite was measured and predicted by various machine learning algorithms. The recorded datasets were used for optimization by Response Surface Methodology (RSM) model whereas, training and testing of the four different Artificial Intelligence (AI) models were executed using machine learning methods. The four AI models applied in this study were Support Vector Machine (SVM), Artificial Neural Network (ANN), Random Forest (RF) and Extra Trees (ET). The objective of this analysis was to quantify the accuracy of microhardness prediction of four types of AI models along with RSM model. The obtained results revealed that the extra trees (ET) model showed outstanding performance amongst the five models for training, testing, and overall datasets with coefficient of correlation (R2), MSE and MAE value of 94.47, 75.38 and 4.67, respectively. This analysis therefore recommends the ET model in the prediction of microhardness of electroless Ni–P–TiO2 composite coating due to its superior and robust performance.

Keywords

• Microhardness
• Electroless Ni–P–TiO2 coating
• Response surface methodology
• Artificial neural network
• Random forest
• Extra trees