Optimization of automated anodizing plant efficiency and process prediction using Random Forest based Levy flight method

Abstract

The anodizing process is mainly carried out by industrialists to protect aluminium components from corrosion and strengthen their physical and chemical behavior. The goal of this study is to minimize the total cycle time of aluminum anodizing by optimizing pre-treatment procedures and forecasting the anodization process time. The experiments were carried out at the GOLDEN ANODIZER premises, a local aluminum anodizing industry in Coimbatore, India. The relationship between various anodizing pre-treatment parameters such as Nitric bath pH, Caustic pH, alkaline pH, Mean time duration was optimized to reduce time without affecting output qualities like Surface finish (µm), Film thickness (µm), and correction (%). Also, the anodizing pre-treatment process parameters were predicted using a hybrid machine learning method called the Random Forest Levy Flight Algorithm (RF-LFA). The proposed RF-LFA shows 2–3 times better prediction performance than existing RF and deep neural networks.