類神經模糊在PCB壓合製程最適化之應用

The application of Neuro-Fuzzy on the optimization process of the PCB Laminate

楊義民(Yih-Min Yang)

指導教授 ： 李 錫 捷 博士

元智大學/資訊學院/資訊管理學系/碩士(2002年)

摘要

類神經模糊在PCB壓合製程最適化之應用
學生：楊義民 指導教授：李錫捷 博士
元智大學資訊管理學系碩士班
摘 要
目前電子業在PCB壓合製程上為達到溫度控制的目的，應用最廣泛也最普遍的是使用傳統的PID 控制器，若是要達到恆溫的控制PID控制器是具有其適用性和效益，但是當要達到特定溫度的昇降曲線，使用傳統的PID控制器要達到控制的目的就相當不容易，若還要達到最適化的控制就更加耗時費力。
本研究希望應用類神經網路的學習機能配合模糊控制理論,應用一套PC Base 上的模糊控制器,協助電子業在PCB壓合製程上執行溫度昇降曲線的功能並和工業界普遍使用的可程式控制器(PLC)系統整合以達到最適化的控制。
最後本研究將應用模擬測試的結果證明其可行性，如此將理論和實務密切結合，進而推展至相關應用領域以發揮其最大效益,是本研究主要的方向和目標。

關鍵字

溫昇曲線控制 ； 模糊控制 ； 類神經網路 ； 類神經模糊

並列摘要

The application of Neuro-Fuzzy on the optimization
process of the PCB Laminate
Student：Yih-Min Yang Advisor：Dr. Hsi-Chieh Lee
Department of Information Management
Yuan-Ze University
ABSTRACT
The application of the traditional PID controller is currently the most popular approach for temperature control in the PCB laminate manufacturing process in the electronics industry. Traditional PID controller works quite well when constant temperature control is required. However, it is never a trivial job to apply the traditional PID controller effectively for control specific ascending descending curve. Consequently, it is even more difficult applying PID controller for optimize control.
In this study, neuro-fuzzy approach is utilized in an attempt to optimize the PCB laminate manufacturing process. A PC-based fuzzy controller is designed to realize the function of the temperature ascending/descending curve for the PCB laminate manufacturing process. Neural network is used in the fuzzy controller in order to learn the membership function from the trained data. Meanwhile, this neuro-fuzzy controller can also be integrated with PLC system, the de facto standard that is widely adopted for achieving the optimization control.
Experimental results from simulated PCB laminate process control have shown the feasibility and usefulness of the neuro-fuzzy controller. Due to the nature the process control, it is not too difficult to modify the neuro-fuzzy controller for the optimization of PCB laminate process for other similar process optimization control.

並列關鍵字

Temperature ascending curve control ； Fuzzy control ； Neural Networks ； Neuro fuzzy

參考文獻

[1] K. K. Kumbla, and Mohammad Janshidi, “Real time control of direct drive motor by learning neuro-fuzzy controller”, IEEE International Conference on Computational Cybernetics and Simulation, vo1.2,pp.1674-1679,1997.

[3] P.Z. Grabowski, “Direct torque neuro-fuzzy control of induction motor dive,” proceeding of IEEE 23rd International Conference, vo1.2, pp.557-562,1997.

[5] L-H. Hoang,“An adaptive fuzzy controller for permanent-magnet ac servo drives,” Conference Record of the IEEE on Industry Applications Conference, vo1.1,pp. 104-110,1995.

[7] J. M. Dongmo, F. G. Salam, and G. Erten, “Adaptive fuzzy control of ac motors for electric vehicles and manufacturing systems,”Proceedings of the 1996 IEEE International Symposium on Intelligent Control, pp.408-413,1996.

[8] T.-C. M. , L.-H. Hoang, “Model reference adaptive fuzzy controller and fuzzy estimator for high performance induction motor drives,”g, Conference Record of the 1996 IEEE on Thirty-First IAS Annual Meeting, vo1.1,pp.380-387,1996.