Why Genetic Algorithms Are More Suited to Solve Certain Types of Optimization Problems

Details

Title: Why Genetic Algorithms Are More Suited to Solve Certain Types of Optimization Problems
Creator
Subject
Description: Some technical optimization problems that have been normally solved manually or with other traditional techniques, can be optimized by using different advanced techniques. Techniques have evolved to include artificial intelligence (AI) algorithms to get better results. Genetic algorithms are a great example of one of these optimization techniques that uses AI. While these techniques are designed to solve problems more efficiently than before, there are some problems that are better suited to these algorithms compared to others. By analyzing experimental data, it is possible to discern what types of problems are better solved using genetic algorithms are better at solving. Studies that have isolated genetic algorithmic applications have been chosen to better compare the solutions that were found, straying away from hybridizations of any kind that could influence results and blur the line of responsibility. Five optimization problems were analyzed to find the problems that had the best improved results and the style best suited for genetic algorithms. These optimization problems that were studied include: PID control tuning; traveling salesman shortest path; generation of machine turning code; design of a wind turbine tower; and A-Mazer, a maze-solving algorithm. By breaking down these problems and how genetic algorithms are used to enhance solutions, a comparison can be made. It was found that problems that would have multiple trials within traditional techniques are more suited for genetic algorithms over other optimization problems., Honors Thesis Advisor: Dr. Jennifer Wilburn.
Publisher
Contributor: Peters, Leslie E. (Author), California University of Pennsylvania. Honors Program.
Date: 2020-04-22
Type: Text
Format
Identifier: cali:2225
Source
Language
Relation
Coverage
Rights