Solar Forecasting for Power System Operator
- Resource Type
- Conference
- Authors
- Wanady, Irene; Viswanath, Aparna; Mahata, Kaushik
- Source
- 2018 IEEE Electrical Power and Energy Conference (EPEC) Electrical Power and Energy Conference (EPEC), 2018 IEEE. :1-7 Oct, 2018
- Subject
- Components, Circuits, Devices and Systems
Engineering Profession
Power, Energy and Industry Applications
Transportation
Humidity
Predictive models
Forecasting
Time series analysis
Atmospheric modeling
Mathematical model
Solar Forecasting
Singapore
ARMA
maximum likelihood
instrumental variable
- Language
This paper aims to build a solar forecasting model for the power system operator to allow them to make informed decisions on the electricity market dispatch. Detailed literature review on meteorological and atmospheric sciences is performed to understand the various factors which affect the solar irradiance level. These parameters are classified into four types. The first type is the meteorological parameters which vary with the date and time and the second type is the parameters which depend on the location. Using known equations and existing empirical models, the parameters classified in these two types is determined. The third classification is the parameters which are affected by the weather and this includes temperature and relative humidity. In this paper, statistical prediction method will be used to forecast these two parameters. Temperature and humidity are related to each other and therefore, vector time series is used in the prediction method. Stationary time series data will be used in the ARMA model fitting. The innovation series was found before maximum likelihood and instrumental variable method are used to determine the suitable parameter for the ARMA model. The last classification for this paper is the parameter for the cloud cover. Image processing of satellite images will be used to determine this cloudiness parameter. Solar irradiance is then calculated using the combination of all these parameters. This method is illustrated by using Singapore weather data.