مقارنة بين طرائق تحليل وتنبؤ السلاسل الزمنية وتطبيقها على مبيعات الشركة العامة لتوزيع كهرباء بغداد == Comparative Studies Among The Methods of Analysis And Forecasting Time Series With Application On Electricity Sales In Baghdad City
Author name:
اسيل سمير محمد محمود
Supervisor name:
عبد المجيد حمزة الناصر
General topic:
Administration and Economics
Specific topic:
Statistics
Degree:
Doctorate
University:
University of Baghdad - Faculty Of Administration And Economics - Department Of Statistics
Language:
Arabic
University location:
Baghdad
First pages:
07T3412 - p.pdf
Abstract:
يعد التنبؤ بالسلوك المستقبلي للسلاسل الزمنية من الموضوعات الهامة في العلوم الاحصائية، وذلك للحاجة اليه في مجالات الحياة جميعا، مثل التنبؤ بالحالة الجوية ودرجات الحرارة، حالة السوق والاسعار، تدفق المياه، واستهلاك الطاقة الكهربائية. وقد تزايد الاهتم | Forecasting of future behavior of time series is one of the important subjects in statistical science, because of its need in the different fields of life, like forecasting of weather state and air temperature market state and price water flow consumption electrical power, in the recent years there is an increased interesting in forecasting, and some new techniques like Artificial Neural Networks and Fuzzy Trend Sets. These techniques are able for learning and self - adaptation with any model, and don’t need assumption on the natural of time series. On the other side, the classical forecasting methods like Box - Jenkins method, exponential smoothing and adaptive filtering need certain conditions. A condensed study was done to compare between ordinary methods, namely; Box & Jenkins, ES, and Adaptive Filtering with modern methods namely; ANN, and Fuzzy Trend Set, where some new results are obtained and new method was proposed. A raw data electrical power in Baghdad city is used to perform this comparison through the application of the two programs Statistica and Matlab. From the practical application it found that proposed method gives better and more efficient results than others.
