RETRACTED ARTICLE: ARX and ARMAX modelling of SBCNC-60 machine for surface roughness and MRR with optimization of system response using PSO

In the process of attaining high-end machines, control of machining systems via optimized machining parameters along with their transient responses is highly essential. By implementing system detection methodologies, a new methodology is proposed to introduce the best mathematical model, which subsumes the best FIT, fewer parameters, minimum MSE, and residuals amongst ARX and ARMAX models for an SBCNC-60 Machine to gratify the controller’s design requirement. From the CNC machine, the real-time measurement data samples are obtained for model detection; then, they are simulated with the aid of MATLAB. Here, for the study of Metal Removal Rate (MRR), the multiple inputs with the single-output system are utilized; similarly, for tuning operation, the Surface Roughness (SR) is measured; subsequently, the MRR is utilized for drilling operation on P8 (H-13, High-Speed-Steels) materials, which were detected by ARX and ARMAX for varying orders. To optimize the output MRR, the best-fit models were selected for control regarding the PID as well as FOPID controller; moreover, in the ‘3’ inputs’ SR, one input differs at a time whilst retaining the other 2 constants at their mid-levels. Better time-domain characteristics were obtained by the PSO tuned FOPID controlled ARX 331 model than the PSO-PID controller for MRR (tr = 6.86 s., Mp = 1.94%, ts = 8.93 s.) and SR (tr = 1.13 s., Mp = 2.47%, ts = 2.68 s.) in case of turning operation, the ARX 311 is the best-suited model for MRR (tr = 0.0818 s., Mp = 1.8%, ts = 2.78 s.) while running for drilling operation. A prominent effect of the varied cutting speed input variable was illustrated by these models; thus, affecting the output performance like MRR and SR for various operations performed during the machining process.