Application of artificial neural networks to strip steel surface defect diagnosis

Based on the analysis of strip steel surface quality examination carried at home and abroad, the paper analyzes flaws and corresponding factors beginning with the design of examination system. It studies deeply the related theories and key techniques of strip steel surface quality examination system, applied neural networks for strip steel surface defect recognizing successfully. It is applied successfully to whole flow quality control technique and equipment composite diagnosis system (TQC-DS) in a steel company.