الفهرس 
ANT COLONY OPTIMIZATIONOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
fuzzy systems have been widely applied to automatic control, pattern recognition and decision analysis. Fuzzy control methodologies have emerged in the recent years especially in automatic control as promising ways to nonlinear control problems. However, a common bottleneck encountered in fuzzy controller design is that derivation of fuzzy rules is often difficult, time consuming, and requires expert knowledge although the human experts find it difficult to examine all the input- output data from a complex system to obtain a set of suitable rules for the fuzzy system. This disadvantage can be solved by automating the design of fuzzy systems. Many metaheuristic learning algorithms have been used to automate fuzzy system design. A new metaheuristic optimization technique, which is the ant colony optimization (ACO) inspired by real ant colony observations has recently been proposed. In the ACO technique, artificial ant colonies cooperate in finding optimum solution. A fuzzy system consists of a set of fuzzy IF-THEN rules that describe the input-output mapping relationship of the system, so the fuzzy controller (FC) design includes designs of antecedent part (IF part) and consequent part (THEN part).The antecedent part can be partitioned in advance without much difficulty. One challenging design task is the determination of the consequent part. The ACO is employed to design the consequent part.
In this thesis, several ACO algorithms have been introduced to design the FC such as the ant system (AS-FC), max-min ant system (MMAS-FC) and ant colony optimization system (ACO-FC). To compare between them they are applied to a nonlinear plant for tracking control. To compare between the FC designed by ACO and classical FC, the two techniques are applied to a DC motor for position control. There are many problems that may be encountered in DC motor such as disturbances, and nonlinearities especially friction, deadzone, saturation. All of these problems are combined to represent control problem requiring an effective and robust controller