Building a Decision Support System for Determining Cutting Parameters in Turning:A Case-Based Reasoning Approach

Determining proper cutting parameters is key to an effective and economical machining operation, especially turning on a lathe. The complex nature of the domain involving multiple physics at play necessitates the use of alternative models using mathematical, statistical, and computational methods. Decision support systems based on experimental data involving these models help to specify the proper levels of cutting parameters in order to achieve desired machinability outcomes. In this work, case-based reasoning approach is adopted, which tries to estimate solution based on past similar but non-identical records stored as cases. The database of cases better known as case base, which is core to this paradigm is created by virtual data generated by statistical regression model, which is initially built by experimental data while turning EN24 steel with uncoated carbide tool insert. The searching of the best similar matching case is done using a well-established k-NN algorithm. The results obtained by this approach are validated by confirmatory runs and benefits of this over counter proposition of using exact search from database are discussed.