Loading...

Using Machine Learning to Predict the Behavior of Concrete Dams with Data of Monitoring

Momeni Golshad, Mohamad Reza | 2022

67 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56159 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Ghaemian, Mohsen; Amini, Zahra
  7. Abstract:
  8. Karun 4 arch concrete dam with a height of 230 meters is designed and built in one of the most complex natural places. Therefore, monitoring the behavior and evaluating the safety and stability of this dam, both with regard to the height and nature of the dam itself and with regard to its national and international dimensions, is of special importance and sensitivity. This sensitivity is doubled due to some problems and issues encountered in the body of the Karun 4 dam, which of course are considered probable and possible phenomena in arch concrete dams. The Karun 4 Dam construction site is located in the southwest of Iran on the Karun River, immediately upstream of the Karun 3 Dam reservoir and at a distance of 200 meters upstream from the intersection of the Manj and Karun rivers. In this area, thick sediments with a depth of more than 30 meters have formed in the river bed. The dam construction valley has an asymmetrical v shape with a steeper general slope on the right side. Karun 4 Dam with a height of 230 meters is the tallest dam built and in operation in the country, which was built to regulate 7.3 billion cubic meters of water of the Karun River and produce 2107 million kilowatt hours of hydroelectric power per year. The Karun 4 dam's instrumentation system, focusing on measuring the movements of the dam body and foundation rock mass, includes more than 850 different instruments such as pendulums, extensometers, seam gauges, inclinometers, accelerometers, and water, concrete, and air thermometers. To measure the horizontal displacement of the dam, 5 straight pendulums have been placed in the central block of the dam and the left blocks (9 and 17), and the right blocks (10 and 20) [15]. behavioral information of the dam is investigated by reading the precision instruments of the dam. With the recorded data of instruments such as extensometer and pendulum, an attempt has been made to create regression models according to the factors influencing the devices. These factors include reservoir water level, strain, and other time-dependent variables, such as the dam's age. The algorithm used in this research includes linear and non-linear regression in the supervised category. It has been tried that this algorithm has the best fit with the data of the examined devices, including extensometers so that it can predict the future with the lowest error coefficient. It is expected that after the completion of this project, the most important factors of the dam's behavior, including the displacement of the pendulum and extensometers, can be identified and the future can be predicted
  9. Keywords:
  10. Regression Analysis ; Concrete Dam ; Concrete Arc Dam ; Karoon IV Dam ; Machine Learning ; Structural Health Monitoring ; Structural Stability

 Digital Object List

 Bookmark

No TOC