چکیده مقاله
Artificial intelligence, particularly deep learning, is revolutionizing welding and nondestructive testing NDT This systematic review analyzes 85 key publications 2018 2024 to explore AI applications in weld optimization and inspection Results demonstrate that convolutional neural networks CNNs achieve over 97% accuracy in detecting defects like cracks and porosity from radiographic images, significantly outperforming manual methods in speed For signal based NDT ultrasonic, eddy current , Long Short Term Memory LSTM networks enhance subsurface flaw detection to 95% accuracy In process optimization, hybrid artificial neural network ANN and reinforcement learning models reduce weld defects by 35% and lower energy consumption by 20% through parameter prediction and real time control The integration of AI enables automated, real time quality monitoring, minimizing human error and enabling predictive maintenance However, major challenges remain, including the need for extensive labeled datasets, the black box nature of complex models affecting interpretability, and the high cost of integrating AI with legacy industrial systems Future research should focus on few shot learning, lightweight models for edge computing, and standardized shared databases Ultimately, merging AI with digital twins and robotics paves the way for fully automated, smart welding systems within the Industry 4 0 framework
کلیدواژهها
نویسندگان
شیوه ارجاع
Azimvand, Ali Asghar,1404,A Review on Prediction and Intelligent Detection of Weld Defects Using Deep Learning: Transition from Traditional Inspection to Data-Driven Automated Systems,8th International Conference on Welding and Non Destructive Testing & 26th National Conference on Welding & Inspection & 15th National Conference on NDT & 4th National Conference on Additive Manufacturing,Tehran
ارائهشده در
مجموعه مقالات هشتمین کنفرانس بین المللی جوشکاری و آزمایش های غیرمخرب، بیست و ششمین کنفرانس ملی جوش و بازرسی، پانزدهمین کنفرانس ملی آزمایش های غیرمخرب و چهارمین کنفرانس ملی ساخت افزایشی20 بهمن 1404 · تهران