AUTOMATING ONLINE QUALITY CONTROL BY THE USE OF NEW NEURAL NETWORK ALGORITHMS AND NEURO-FUZZY SYSTEMS

摘要

Online quality control based upon empirical process models became a standard quality assurance tool for specific high complex and expensive parts produced in injection molding. The use of this control method is limited to special-educated engineers due to prerequisite knowledge in statistics and data analysis methods. At IKV algorithms based upon artificial intelligence have been developed to solve this problems and provide online quality control for each injection molded part without any specific education of the user. Well-known process knowledge modeled with Neuro-Fuzzy Algorithms reduces the design of experiments to a simple description of the molding. The new IKV Neural Network Algorithm enables a fully automated process modeling.