09 فروردین 1403
پرويز حاجياني

پرویز حاجیانی

مرتبه علمی: استادیار
نشانی: دانشکده کسب و کار و اقتصاد - گروه علوم اقتصادی
تحصیلات: دکترای تخصصی / اقتصاد کشاورزی
تلفن: 09177743049
دانشکده: دانشکده کسب و کار و اقتصاد

مشخصات پژوهش

عنوان
مدلسازی و پیش بینی تقاضای انرژی در ایران: مقایسه روش شبکه عصبی مصنوعی و الگوهای سری زمانی
نوع پژوهش پارسا
کلیدواژه‌ها
Key words:Forcast, Energy Demand, Artificial Neural Network, Time Series Models.
پژوهشگران صیادی برازجانی فرزانه (دانشجو) ، پرویز حاجیانی (استاد راهنما) ، حجت پارسا (استاد مشاور) ، احمد کشاورز (استاد مشاور)

چکیده

Energy is considered as one of the most critical and effective factors in today's modern life of human beings, which is also influential in the level of welfare as well as quality, efficiency, and production. Regarding its importance, government and energy-based organizations in developed and developing countries concentrate mostly on the prediction of energy consumption. False prediction of energy consumption could lead to extra capacity or failure in energy provision as well as high prices imposed on the society; therefore designing appropriate energy consumption patterns accurately is a cardinal and essential pattern. In the present study, Time Series patterns and Artificial Neural Network Model are used to predict energy demand. The data used in this study constitute energy demand variable, gross domestic product, urban population, energy carriers' prices, imports and exports of goods and services in the years 1357-1393. According to Autoregressive Distributed Lag Model, the results demonstrate that urban population and gross domestic product increase energy consumption and imports of goods and services lead to the decrease of energy consumption. In this manner, the results of long-term relationships of this model express that urban population is the most effective variable in energy consumption and ARDL Model and Artificial Neural Network Model are predicted in the next level; the results of which demonstrate that Artificial Neural Network Model is more proficient in comparison with ARDL Model..