Prediction model based on time series for chloride and nitrate concentration in municipal groundwater wells in Gaza Strip

العناوين الأخرى

نموذج توقعي معتمد على السلاسل الزمنية لتركيزات الكلورايد و النترات في الآبار الجوفية البلدية في قطاع غزة

مقدم أطروحة جامعية

al-Astal, Shukri Muhammad Abd al-Salam

مشرف أطروحة جامعية

Barhum, Tawfiq Sulayman

الجامعة

الجامعة الإسلامية

الكلية

كلية تكنولوجيا المعلومات

القسم الأكاديمي

تكنولوجيا المعلومات

دولة الجامعة

فلسطين (قطاع غزة)

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2017

الملخص الإنجليزي

The groundwater in Gaza strip has a high Total Dissolved Solid (TDS), chloride concentration and nitrate concentration.

Besides that, the high cost and poor data in Gaza strip are considered as problem for water management.

In the part of the target area, there are many researches for using AI techniques to predict chloride and nitrate in groundwater wells.

However, it is not depending on time series and need more data for these wells which not available.

An artificial prediction model for nitrate and chloride concentration in groundwater wells using Data mining techniques in time series is built to improve management of groundwater wells in Gaza strip despite of limitation of data.

In this research, many intelligent techniques as Artificial Neural Network, Support Vector Machine, K-Nearest Neighbours and Linear Regression are applied then determine the optimum one in this case by using Statistics Coefficient of Determination (R2 ) and Root Mean Square Error (RMSE).

Data collected from Palestinian Water Authority that contains historical reads for groundwater wells in 1974 to 2013 period, then data preprocessed with domain expert contribution.

The optimum model has RMSE= 64.68 and R2 = 0.954 for chloride and 33.33 and 0.828 for nitrate respectively by LR .

Finally, final model evaluation took place using 2014 and 2015 PWA data for the targeted field.

The result was RMSE less than 84 and R2 more than 0.9 for chloride.

In additional, RMSE less than 36 and R2 more than 0.86 for nitrate.

عدد الصفحات

51

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-905914