Prediction and Optimization of Surface Roughness in a Turning Process Using the ANFIS-QPSO Method

Area of Science:

• Materials Science and Engineering
• Manufacturing Technology
• Computational Intelligence

Background:

• Surface roughness is a critical parameter in machining, influencing product performance and lifespan.
• Accurate prediction of surface roughness is essential for optimizing machining processes.
• Dry and cryogenic machining offer environmental and performance benefits but require precise control.

Purpose of the Study:

• To develop and validate a novel machine learning approach for predicting surface roughness in AISI 304 stainless steel.
• To evaluate the performance of the proposed method under both dry and cryogenic turning conditions.
• To compare the proposed method against existing techniques for surface roughness prediction.

Main Methods:

• Utilized the Adaptive Neuro-Fuzzy Inference System with Quantum Particle Swarm Optimization (ANFIS-QPSO) machine learning approach.
• Integrated artificial neural networks, fuzzy systems, and evolutionary optimization for enhanced prediction accuracy and robustness.
• Conducted simulations for dry and cryogenic turning processes of AISI 304 stainless steel.

Main Results:

• ANFIS-QPSO achieved high prediction accuracy for dry turning (RMSE = 4.86%, MAPE = 4.95%, R² = 0.984).
• ANFIS-QPSO demonstrated excellent agreement with measured values for cryogenic turning (RMSE = 5.08%, MAPE = 5.15%, R² = 0.988).
• The proposed method outperformed ANFIS, ANFIS-GA, and ANFIS-PSO in prediction accuracy.

Conclusions:

• ANFIS-QPSO is a highly effective method for accurate surface roughness prediction in both dry and cryogenic turning.
• The integration of ANFIS with QPSO offers superior performance compared to other hybrid approaches.
• This method provides a reliable tool for optimizing machining parameters and ensuring surface quality in stainless steel manufacturing.