Application of Genetic Algorithms In discontinuous curve fitting

• Main Authors: Hung-Jen Chen, 陳宏任
• Other Authors: Nacy Lin
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/43859687229586948170

碩士 === 淡江大學 === 資訊工程學系 === 88 === Discontinuous curve fitting is the process which determining the function of best fit to a set of points. It is useful Representing a set of points as a function when we analysis and apply them. In the area of engineering, discontinuous curve fitting can combine with frequently used analysis and the technique of engineering (e.g., system simulation, image analysis) to solve the problem in industry manufacturing.The purpose of discontinuous curve fitting is selecting the coefficients (or parameters) of a function to minimizing the summation of the error of the considered points.Once the form the function and the expression of the error were determined, discontinuous curve fitting becomes the optimization problem on the set of points. Though least squares method (LSM) was frequently used on curve fitting problems, but the error of the obtained solution by it is not always minimum. This paper attempts to apply genetic algorithms (GAs) on complicated discontinuous curve fitting problems. GAs were introduced by Holland (1960-1970). A similar analogy is applied to evolve solutions to complicated optimization problems. The notable feature of Gas is that it simulates the characteristics of a biological system like self-repair and reproduction.