Implementing fuzzy logic controller for guide roll on cold ring rolling

مقدم أطروحة جامعية

al-Khafaji, Muhannad Muhammad Husayn

مشرف أطروحة جامعية

Muhammad, Jammal H.
Wenliang, Chen

الجامعة

الجامعة التكنولوجية

الكلية

-

القسم الأكاديمي

قسم هندسة الإنتاج و المعادن

دولة الجامعة

العراق

الدرجة العلمية

دكتوراه

تاريخ الدرجة العلمية

2011

الملخص الإنجليزي

This work aims to building a combined model based on explicit finite element method and a closed loop feedback controller, which implements fuzzy logic controller, to simulate cold ring rolling process and control the guide roll in real time mode (online control).

The finite element sends feedback signal to the guide roll fuzzy logic controller ; the controller processes the feedback signal data and returns the required control output (guide roll velocity) to the finite element.

The process continues until the target time for finite element simulation is reached.

The developed model is built under ABAQUS / Explicit finite element software environment.

The fuzzy controller was programed using object oriented programing technique by FORTRAN programing language.

The fuzzy controller program is a collection of classes from them the fuzzy controller objects will be generated during the simulation phase.

This program has an ability to read a predesigned fuzzy controller file by MATLAB fuzzy logic toolbox, this characteristic function reduced the time required to design a fuzzy system, as well as, get the full advantages provided by the MATLAB fuzzy logic tool box.

Different types of fuzzy logic controllers (PD-like, PI-like, self-tuning PD-like and self-tuning PI-like) are designed using MATLAB fuzzy logic toolbox and they are implemented to control the guide roll in the combined model.

This entails to investigate the influence of each type on the process stability, the roundness of produced ring, as well as, the performance of these controllers is considered too.

The simulations are executed with two values of feed rate (1MM / sec) and (0.5 MM / esc), in addition, a case study has been applied with the STPDFC and STPIFC to verify the proposed control method.

A proposed method for ring roundness error estimation has been developed to be used as an indication of the ring roundness.

This method is verified by using an analytical method.

The results show that all the fuzzy controllers are good enough to control the guide roll from the metal forming vantage point, i.e.

the produced rings has good roundness for both inner and outer ring circles ; in addition the process stability error indicator was small.

With the fuzzy controllers the stability is increased and all stability results acquired from the proposed method were compared with the results of published researches and they showed good fit.

The best stability result percentage error is obtained from STPIFC.

The roundness is compared with ideal produce ring and the best result is given by STPIFC and the maximum absolute percentage error is (- 1.0397 %).

The fuzzy controllers input / output factors have been manually tuned for the PD-like and PI-like fuzzy controller in order to enhance the controller response performance.

The value of error gain factor is taken to be equal to the set point and the value of the change in error is taken to be.

Whereas the controller output gain factor for the PD-like fuzzy controller is taken and for the PI-like fuzzy controller is G ∆ w = 2.

The input / output gain factors for both the self-tuning PD-like and PI-like fuzzy controllers are taken to be the same as those for both PD and PI fuzzy controller, respectively.

The results showed that the self-tuning PI-like fuzzy controller has best performance from the control engineering points of view.

From the results, a conclusion can be made that the combination of fuzzy logic controller with the finite element and make them work as a one unite to simulate cold ring rolling process served a good performance to examine and addressing the ring rolling process in real time mode.

Also this combination could be applied to other applications in the field of control engineering, as well as, to control metal forming processes.

This is an attempt to simulate the automation of the metal forming processes with the soft computing methods like fuzzy logic and neural networks.

التخصصات الرئيسية

هندسة المواد والمعادن

الموضوعات

• معالجة البيانات
• المنطق الضبابي

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-305296