Loading...

Long-Term Water Demand Forecasting for the Tehran City under Uncertainties

Miraki, Ghasem | 2012

715 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 43002 (09)
  4. University: Sharif University of Technology
  5. Department: civil Engineering
  6. Advisor(s): Abrishamchi, Ahmad
  7. Abstract:
  8. Forecasting model of water consumption amounts could be used in order to manage water resources for future condition of city. In this thesis, a model for forecasting water demand for Tehran has been presented by evaluating regression models and intelligent models. In this study, uncertainties which are connected to climate and population changes are taken into account. The considered variables include minimum, maximum and medium temperature, precipitation and solar radiation. Considering objectives of this thesis and various forecasting methods and their advantages and regional conditions of Tehran, in addition to regression analysis, perceptron neural network, probabilistic neural network and adaptive neural-fuzzy networks models are used to forecast water demand of Tehran. In this study, three population scenario-based logistic curve based on effective parameters on population including birth, death and immigration is used to forecast population of Tehran. Lars-wg atmospheric model is used to forecast climate change in three climate scenario. In this thesis with developing, training, testing and calibration of various simulation models, it was determined that adaptive neural-fuzzy networks model is the most appropriate model to forecast water consumption of Tehran and results of forecasting with chosen model in the most of water consumption scenarios indicate an increase in water consumption of future years
  9. Keywords:
  10. Forecasting ; Urban Water Demand ; Tehran City ; Regression Analysis ; Adaptive Neuro-Fuzzy Inference System (ANFIS) ; Perceptron Neural Network

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC