Adaptive Surface Roughness Prediction in Turning Processes with Dynamic Lubrication using Fish Swarm-Intelligent Modified Xgboost Approach

Abstract

Accurately anticipating Surface Roughness (SR) throughout turning operations is a continuous difficulty for the machining industry, particularly under variable lubrication circumstances. Surface finish quality is sometimes subpar because of standard models' low responsiveness to changes in lubrication conditions and machining settings. The present research introduced a novel Artificial Fish Swarm-Intelligent Modified XGBoost (AFSI-MX) methodology to tackle this problem by combining the XGBoost technique's potent prediction powers with the collective thinking of fish swarm behavior. First, the dataset is collected to assess the suggested AFSI-MX technique in relation to SR prediction. This study is carried out using the AFSI-MX approach on the Matlab platform. The suggested AFSI-MX technique is effective in forecasting SR in turning procedures over variable lubrication, as demonstrated by expserimental findings. The suggested AFSI-MX technique outperforms conventional methods in comparison when it comes to managing the complexity and unpredictability present in machining situations