A Software Based On Artificial Intelligence For Analysis Of Injection-molded Faults

For the application field of injection industry, the stability of quality is always the main problem for the user as the process and interaction is complex. The thesis combines with the specialty of injection technology, invents A Technical Research Based on Artificial Intelligence for Analysis of Injection-molded Faults. Through adopting fuzzy logic and neural network, combining with expert theory knowledge and practice experiment, the thesis analyzes the injection technology and the cause of quality defect, discusses and solves the influence of technology from the quality defect entirely.The thesis summarizes the common defect characteristic,reasons and solution methods. It sets up a database based on the knowledge. The system analyzes the process of injection defects, and confirms the suitable analytical program. It will make the system have the function of fast demonstration. The thesis devises advanced expert system to solve the injection defect. Through the theory of expert system, the design can deal with the module—interface,neural network module,analytical machine module,database management module smoothly. Meanwhile, it can perfect the total function for the system.The system adopts Access2003 to realize the function of data base. The user operation interface is executed by the Visual Basic 6.0. The neural network of MATLAB is used to analyze the defect. The program adopts ADO object to visit the data base, builds connection, solving the demand of editing data base softly. Meantime, by using the ActiveX controller, the program realizes the function of MATLAB in the Visual Basic easily.Finally, the thesis includes the process that solving the problem by the system, checks the effectiveness and operation. Through the practice, this technology can help user to deal with the question of quality defect directly, and supply the correct direction to solve it.