MU Digital Repository
Logo

Fuzzy logic supervisory control of discrete eyent System

Venkateswaran, P R (2008) Fuzzy logic supervisory control of discrete eyent System. Phd. Thesis thesis, Manipal Institute of Technology, MIT, Manipal.

[img] PDF
venkata.pdf - Submitted Version
Restricted to Registered users only

Download (77MB) | Request a copy

Abstract

The behaviour mapping of systems has undergone rapid change with advances in technology. The representation has become digital with state description of the behaviour, This manifestation is ~alled Discrete Event Systems (DES). exhibiting properties such as non - determinism, conflict and parallelism. As the complexity increases, they demand better design tools and its implementation as complexity. The solution can be broadly classified into two parts: an effective modeling tool and valid translation for implementation. The solution will in fact be more effective if it can be integrated with the existing approach or modification of existing approaches. Hence, Supervisory Control Theory (SeT) is chosen as a modeling paradigm and Programmable Logic Controllers (PLC) as an implementation tooL In supervisory control, given a model of the system and the desired system behaviour specifications, the objective is to :find a supervisor (controller) such that the controlled behaviour of the system does not contradict the specifications given and does not unnecessarily constrain the behaviour of the system. The control specifications are provided as a set of forbidden conditions that must be avoided in system behaviour. The major problems in implementation of supervisory control brings in are two fold: first, the number of states grows exponentially as the system complexity increases. It also lacks graphical visualization of the system for interpretation. Petri nets are considered to be the alternative approach for analysis, design and implementation of DES. However, the prospect of uncertain and vague information still creates a problem. Hence, to handle the uncertainty, fuzzy logic is integrated into Petri nets to form Fuzzy Automation Petri nets as a modeling tool to handle these uncertainties. The use of Fuzzy Automation Petri nets (FAPN) demands a valid translation of the formalism to Ladder Logic Diagram (LLD) for implementation. This is achieved using Token Passing Methodology. The supervisory controller is thus implemented in PLC. This work investigates the use of Fuzzy Petri nets in supervisory control and suggests a modified and improved version called Fuzzy Automation Petri net (FAPN) as a modeling tool. It presents a systematic approach to the synthesis of Fuzzy Petri net basedsupervisor for the forbidden state problem using supervisory design procedure. The controlled model of the system can be constructed from this FAPN net structure. The implementation is using Flexible Manufacturing System (FMS) as an example of DES. Theresuhs can be interpreted and applied to high level manufacturing systems.. where the role of the supervisor is to coordinate the control of multiple machines or to low level manufacturing systems. where the control function is to switch ON/OFF with respect to the dynamics between different valves. In addition to this design, the conversion formalismfrom FAPN to LLD is also proposed and illustrated.

Item Type: Thesis (Phd. Thesis)
Subjects: Engineering > MIT Manipal > Biotechnology
Depositing User: MIT Library
Date Deposited: 06 Jan 2015 09:19
Last Modified: 06 Jan 2015 09:19
URI: http://eprints.manipal.edu/id/eprint/141414

Actions (login required)

View Item View Item