A Hybrid Approach for Predictive Modeling of KPIs in CNC Machining Operations.

In a CNC machining operation, key performance indicators (KPIs) of process, such as machining time, quality, and energy consumption, vary with cutting parameters. This paper explains a methodology for building physics-guided data-driven models for predicting these process KPIs in CNC machining operations from the planning, machining, and quality data. These physics-guided data-driven models are developed by combining data-driven and physics-based models of machining operations. Using hybrid physics-ML method, predictive modelling of energy consumption and surface roughness in CNC milling operation is also explained by conducting experiments. Finally, accuracies obtained by these models are compared with respective physics-based and data-driven models. [ABSTRACT FROM AUTHOR]