University of SaskatchewanHARVEST
  • Login
  • Submit Your Research
  • About
    • About HARVEST
    • Guidelines
    • Browse
      • All of HARVEST
      • Communities & Collections
      • By Issue Date
      • Authors
      • Titles
      • Subjects
      • This Collection
      • By Issue Date
      • Authors
      • Titles
      • Subjects
    • My Account
      • Login
      JavaScript is disabled for your browser. Some features of this site may not work without it.
      View Item 
      • HARVEST
      • Electronic Theses and Dissertations
      • Graduate Theses and Dissertations
      • View Item
      • HARVEST
      • Electronic Theses and Dissertations
      • Graduate Theses and Dissertations
      • View Item

      ADAPTIVE TUNING OF A PID CONTROLLER USING NEURAL NETWORKS

      Thumbnail
      View/Open
      El_Baradie_Mostafa_Mohamed_1994_sec.pdf (11.16Mb)
      Date
      1994-07
      Author
      El Baradie, Mostafa Mohamed
      Type
      Thesis
      Degree Level
      Masters
      Metadata
      Show full item record
      Abstract
      System identification and adaptive control have been of interest for several years. The emergence of neural networks has added a new dimension to the paradigm of learning and adaptive control. Neural networks are distinguished by their diversity of applications. Most of the work done in the system identification area and control of chemical processes uses some form of standard back-propagation as a learning algorithm. A different learning scheme called Quick Propagation (QP) has been implemented in this thesis. This algorithm has evolved from the initial work of Scott E. Fahlman [121. The advantage of this algorithm lies in the fact that fewer tuning parameters are used. This makes it easier to implement and more general to accommodate different applications. The simulation studies presented in this work show the effectiveness of this approach in identifying a non-linear chemical process, specifically a nonisothermal continuous stirred tank reactor where an irreversible exothermic reaction is carried out in a perfectly mixed reactor. PID (Proportional, Integral, and Derivative) controllers have been the most popular controllers in the chemical industry. Several studies focused on adaptively tuning the PID gains. In this research work, the error signal was employed as a teaching signal for the Quick Propagation neural network in order to adjust the controller parameters. The results of computer simulation studies are presented, and the performance of the proposed technique is described. From these results, it is observed that the neural network is capable of tuning the PID controller parameters toward a desired target state. Furthermore, the performance of the proposed technique was compared to one of the traditional tuning techniques, the Ziegler and Nichols method. The evaluation was based on subjecting both techniques to different model uncertainties. The simulation studies demonstrated the effectiveness of the proposed approach. In this thesis, an arbitrary non-linear model and a CSTR (continuous stirred tank reactor) model were considered.
      Degree
      Master of Science (M.Sc.)
      Department
      Electrical and Computer Engineering
      Program
      Electrical Engineering
      Copyright Date
      July 1994
      URI
      http://hdl.handle.net/10388/11597
      Collections
      • Graduate Theses and Dissertations
      University of Saskatchewan

      University Library

      © University of Saskatchewan
      Contact Us | Disclaimer | Privacy