Research And Implementation Of Defect Correction System Of Plastic Injection Molding

Injection molding is a complicated process and there are many factors which influence the quality of the final products. Therefore, it is difficult to analyze and correct the reasons of defects by theoretical formula. Currently, the defects of product are corrected mainly through"trial and error method"based on the molding personnel's experience. The relationship between the product quality and parameters is nonlinear and time dependent. Considering advantages of artificial intelligence when dealing with the problems, more and more research centers try to apply the artificial intelligence algorithm to the defect correction. Fuzzy reasoning algorithm is one of artificial intelligence algorithms.Based on the molding technologists'thinking during the molding trials, in this paper, a defect correction system is designed and developed. The core of the defect correction system is the fuzzy reasoning model, which is mainly composed of fuzzification, fuzzy reasoning and defuzzification. The system inputs are defect level from the user feedback and real-time machine parameters obtained by the portable real-time data collection system. When completing the conflict resolution, the defect correction system will output the correction values after fuzzy reasoning. Fuzzy reasoning is based on the knowledge and need lots of data in the reasoning process such as membership, rules of inference and so on. This paper describes how to create a knowledge base of the system. In order to make it real-time and automatic when the defect correction system runs, a data collection system is designed and implemented in this paper. This paper describes the data collection system's overall architecture, the design of hardware and software systems. The data collection system can also provide accurate operating information of the injection process for injection molding personnel to improve the quality of injection products. In order to accurately judge the cycle of the data collection curve, this paper designs and implements the curve-fitting and the cycle judgment based on the PSO (Particle Swarm Optimization) algorithms, and the method proves to be effective.The defect correction system needs much data when it runs. Therefore, in order to store, manage, write and read these data with a better method, this paper describes how to define and implement the overall design of the system database, data table and the relationship of different tables. The system uses the MySQL software to establish, implement and manage the database which has nine data tables. Meanwhile, this paper describes the implementation process of the database in Linux operation system and proves the database effective through test, which makes it convenient to transplant the system to the controller of the injection machine in future.