Publications

Detailed Information

Development of Stability Formulas for Tetrapods and Rock Armors Using Multigene Genetic Programming : 다중유전자 유전프로그래밍을 이용한 테트라포드 및 사석피복재의 안정공식 개발

DC Field Value Language
dc.contributor.advisor서경덕-
dc.contributor.author이재성-
dc.date.accessioned2017-10-31T07:30:14Z-
dc.date.available2017-10-31T07:30:14Z-
dc.date.issued2017-08-
dc.identifier.other000000145515-
dc.identifier.urihttps://hdl.handle.net/10371/137307-
dc.description학위논문 (석사)-- 서울대학교 대학원 공과대학 건설환경공학부, 2017. 8. 서경덕.-
dc.description.abstractTetrapods and rock armors are widely used for armoring rubble mound breakwaters. Calculating the stability number of armor units is a necessary process to determine the optimal weight of armor units. Many stability formulas have been proposed to calculate the stability number of armor units since the Hudson formula in 1950s. Most of them are proposed by performing regression analysis for the parameters of the equation using the data obtained from the hydraulic model tests. In recent years, when there is a large amount of experimental data, machine learning methods have been introduced. For rock armors, an artificial neural network (ANN) model was proposed by Mast et al. (1995) and a combination of the ANN model with other models has also been proposed recently. The ANNs show good results, but it is complicated to calculate the output using the input data, so engineers have difficulty in using it in practice. To solve this problem, this study propose the definite functions to calculate the stability number of Tetrapods and rock armors through symbolic regression using multigene genetic programming (MGGP). This method, also known as multigene symbolic regression, has the advantage of obtaining both the parameters and the structure of the formula without assuming the structure of the formula. The proposed formulas are developed in terms of dimensionless variables and can be applied to both laboratory and field applications, and can be applied to any types of wave breaking. The final formulas are more accurate than the previous stability formulas and are simple to use by engineers.-
dc.description.tableofcontentsCHAPTER 1. INTRODUCTION 1
1.1 Background 1
1.2 Objectives 4
CHAPTER 2. THEORETICAL BACKGROUDS 5
2.1 Stability number of armor units 5
2.2 Parameters that influence stability of armor units 7
2.2.1 Hydraulic parameters 7
2.2.2 Structural parameters 9
2.3 Empirical stability formulas for armor units 12
2.3.1 Rock armors 12
2.3.2 Tetrapods 14
2.4 Multigene genetic programming 17
2.4.1 Genetic programming 17
2.4.2 Multigene symbolic regression 22
CHAPTER 3. MODEL DEVELOPMENT 25
3.1 Experimental data and input data 25
3.1.1 Rock armors 25
3.1.2 Tetrapods 26
3.2 Bootstrapping for training data sampling 28
3.3 Measure of complexity of models 32
3.4 Parameters setting for MGGP model 36
3.4.1 Effect of the maximum tree depth 36
3.4.2 Effect of population size and number of generations 39
3.4.3 Effect of crossover rate and mutation rate 42
3.4.4 Effect of number of multigene 44
3.4.5 Parameter setting for each model 46
CHAPTER 4. RESULTS AND DISCUSSIONS 48
4.1 Bootstrap sampling results for training data 48
4.2 Stability formula for Tetrapods 59
4.3 Stability formula for rock armors 64
CHAPTER 5. CONCLUSIONS 69
REFERENCES 71
국문초록 74
-
dc.formatapplication/pdf-
dc.format.extent9498248 bytes-
dc.format.mediumapplication/pdf-
dc.language.isoen-
dc.publisher서울대학교 대학원-
dc.subjectTetrapod-
dc.subjectrock armor-
dc.subjectstability number-
dc.subjectmachine learning-
dc.subjectgenetic programming-
dc.subjectmultigene genetic programming-
dc.subjectsymbolic regression-
dc.subject.ddc624-
dc.titleDevelopment of Stability Formulas for Tetrapods and Rock Armors Using Multigene Genetic Programming-
dc.title.alternative다중유전자 유전프로그래밍을 이용한 테트라포드 및 사석피복재의 안정공식 개발-
dc.typeThesis-
dc.contributor.AlternativeAuthorJae Sung Lee-
dc.description.degreeMaster-
dc.contributor.affiliation공과대학 건설환경공학부-
dc.date.awarded2017-08-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share