Please use this identifier to cite or link to this item:
http://repository.aaup.edu/jspui/handle/123456789/2740
Title: | Prediction of Water Demand in North of Palestine Using Artificial Neural Networks رسالة ماجستير |
Authors: | Alkelani, Mohammad Tayseer Mahmoud Zaid$AAUP$Palestinian |
Keywords: | networks,algorithm,function neural network |
Issue Date: | 2018 |
Abstract: | Water is very significant to humanity, in all our life fields, such as agriculture, industry, and health. In our daily life, water consumption has been increased in a significant manner. The use of domestic water is considered the most fundamental factor in the chain of water demand and consumption in urban areas. Several models and more computational techniques have been proposed and presented in the last decades to predict the water demand in short or long-term time series, such as linear statistical and mathematical methods for predicting the future quantities of daily, monthly and yearly demand. Other more advanced methods like the regression of time series analysis approaches have been applied. The majority of these methods depends on extrapolating historical trends and linked the demand with socioeconomic variables. It is necessary to evaluate the ability of existing resources to meet future needs and to provide the basis for planning for future systems and improve them to limit the uncertainties in demand predictions. These uncertainties include the population growth, economic change, changes in consumption habits, and climate change especially in our region, hence the prediction of water demand helps water distribution companies and government knowing and to investigate the expected demand and the impacts on the sustainable development planning. According to a set of challenges that face the water authorities, municipalities and the water sector, in general, the most important limitation is lack tools that can predict the water demand in the future. The Jenin city municipality does not have computer applications to assist the future needs estimation and the satisfactory distribution of water. VI In this thesis, we use linear and nonlinear methods to predict water demand in the city of Jenin, for this purpose, we use different types of Artificial Neural Networks (ANNs) with different learning methods to predict the water demand, and compare our results with a known types of statistical methods. In this context a computational technique which depends on artificial neural networks (ANNs) and a hybrid method of NNs with optimization algorithm is used to predict the short-termwater demand in urban regions. The dataset depends on sets of collected data (extracted from Municipalities Databases) during a specific period of time and hence, we propose a nonlinear model for predictingthe monthly and yearly water demand and finally provide the more accurate prediction model compared with other linear and nonlinear methods. We aim to create a model capable of making an accurate prediction for water demand in the future for the Jenin city. This prediction is made with a time horizon (months or years) depending on the extracted data. This data will be used to feed the neural network model to implement mechanisms and a system that can be employed to predict a short-term for water demand, unlike the current situation where the water authority in Jenin organize the water where supply one day on a week for each region in the city, hence our idea is to organize the water distribution for each region in Jenin city depending on the past consumption. This method is based on the use of K-means clustering algorithm to classify the regions, so the authorities can supply a water for each region with a number of days to ensure a fair distribution.Two applied models of artificial neural networks are used; Multilayer Perceptron (MLP) and Radial Basis Function Neural Networks (RBFNNs) with different learning and optimization algorithms, and one type of linear statistical method called Autoregressive Integrated Moving Average (ARIMA) are applied on the water demand data collected from Jenin city to predict the water demand in VII the future. The obtained results demonstrate that the MLPNN type is surpassed the RBFNN and ARIMA models in the prediction the water demand values according to lowest error and goodness fit. Furthermore, the experimental result of clustering the regions in the city provides an efficient approach to supply water fairly. |
Description: | Master`s degree in Computer Science |
URI: | http://repository.aaup.edu/jspui/handle/123456789/2740 |
Appears in Collections: | Master Theses and Ph.D. Dissertations |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
محمد الكيلاني.pdf | 10.72 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
Admin Tools