Type of Document	Master's Thesis
Author	Fong, Nga Hin Benjamin
URN	etd-12042009-020333
Title	A comparison study of genetic algorithms in feedback controller design
Degree	Master of Science
Department	Mechanical Engineering
Advisory Committee	Advisor Name Title Robertshaw, Harry H. Committee Chair Cudney, Harley H. Committee Member Wicks, Alfred L. Committee Member
Keywords	Genetic algorithms
Date of Defense	1994-12-08
Availability	restricted
Abstract This thesis discusses the use of genetic algorithms as a global search technique to solve three optimization problems: a sixth-order polynomial problem, a single-degree-of-freedom spring-mass-damper (SDOF SMD) system problem, and a loading bridge regulator problem. Genetic algorithms are iterative global search techniques based on the principles of natural selection and population genetics. The theory, design and implementation of the algorithm is discussed in detail. The Simple Genetic Algorithm (SGA) is presented to solve a sixth-order polynomial optimization problem. Results from two traditional numerical techniques will be compared with the SGA results as well as the analytical calculus solution. In addition, the effect of different parametric sizes of the genetic operators are investigated. In the second problem, genetic algorithms are used to design a two-state feedback optimal gain set for a SDOF SMD model with a given initial condition. An improved selection scheme called the stochastic remainder selection without replacement is introduced. An improved GA-based (IGA) feedback controller is designed to control the system. Lastly, a regulator control problem is presented using advanced genetic algorithms (AGA). Two-point crossover and inversion operators are employed. A loading bridge is chosen as the control model. An advanced GA-based full-state feedback controller is designed to control the loading bridge with the given reference input voltage. The conclusions show that SGA is more robust than traditional numerical techniques to solve multi-modal functions. Among the three GA approaches considered, AGA is the most robust one for the design of adaptive feedback controllers.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access  LD5655.V855_1994.F664.pdf 2.78 Mb 00:12:52 00:06:37 00:05:47 00:02:53 00:00:14 next to an author's name indicates that all files or directories associated with their ETD are accessible from the Virginia Tech campus network only.

If you have questions or technical problems, please Contact DLA.