List of Articles Hamid Jabari

• Open Access Article
  • Abstract Page
  • Full-Text
  
  1 - Short-Term Forecasting of Wind Farm Power Production Using a Modified Artificial Neural Networks Based Algorithm in Python: A Case Study in Manjil
  Hamid Jabari Ardalan Shafiei-Ghazani Farkhondeh Jabari
  This paper presents a new approach for short-term forecasting of wind farm power generation using artificial neural networks under Python programming language. In this method, weather conditions such as wind speed, wind direction, temperature and air pressure are select More

  This paper presents a new approach for short-term forecasting of wind farm power generation using artificial neural networks under Python programming language. In this method, weather conditions such as wind speed, wind direction, temperature and air pressure are selected as key features affecting the power production of the wind farm. To achieve a relatively accurate estimate, the root mean squared error of the predicted values is calculated and minimized as the objective function. The speed and accuracy of the proposed algorithm have been evaluated by conducting a case study on a wind farm located in Manjil, Iran. The power production of the wind power plant is predicted for a time horizon of one week and hour by hour using the wind speed, wind direction, temperature and air pressure during 8592 hours (total hours of a year minus hours of a week). The root mean squared error, the highest relative error percentage, the time resolution of the forecasts and the calculation time of the proposed algorithm are compared with other algorithms published in recent years, which shows the effectiveness and high accuracy of the results in a short calculation time. The power production of the wind farm was predicted hour by hour during a week and 168 data points were obtained, the root mean squared error in the optimal scenario is equal to 0.010817. The calculation time of the forecasting algorithm is less than 1 minute, and the maximum relative error in the proposed method is 2.3%, which demonstrates that the uncertainties associated with the power production of the wind farm can be reduced by using this short-term forecasting approach. Manuscript profile