The role of artificial intelligence in the modeling, analysis and inspection of ultrasonic welding processes — A review

Ultrasonic Welding is a popular welding procedure that uses high-frequency energy to heat joints. It is a complicated process involving a number of variable parameters that can each greatly modify the final weld product. A number of Artificial Intelligence (AI) technologies have thus been employed to regress and classify results such as weld parameters such as failure load, weld quality and joint strength on the basis of different parameters including power output, annealing temperature and vibration amplitude. Artificial neural network models are the most popular and adept at weld modeling on varying materials and composites. This paper reviews and compares the materials, feature extraction techniques and AI architectures and their performances on predicting a host of welding objectives.