Abstract
In this paper, super-short-term prediction of solar power generation for applications in dynamic control of energy system has been investigated. In order to follow and satisfy the dynamics of the controller, the deployed prediction method should have a fast response time. To this end, this paper proposes fast prediction methods to provide the control system with one step ahead of solar power generation. The proposed methods are based on univariate time series prediction. That is, instead of using external data such as the weather forecast as the input of prediction algorithms, they solely rely on past values of solar power data, hence lowering the volume and acquisition time of input data. In addition, the selected algorithms are able to generate the forecast output in less than a second. The proposed methods in this paper are grounded on four well-known prediction algorithms including Autoregressive Integrated Moving Average (ARIMA), K-Nearest Neighbors (kNN), Support Vector Regression (SVR), and Random Forest (RF). The speed and accuracy of the proposed algorithms have been compared based on two different error measures, Mean Absolute Error (MAE) and Symmetric Mean Absolute Percentage Error (SMAPE). Real world data collected from the PV installation at the University of California, Riverside (UCR) are used for prediction purposes. The results show that kNN and RF have better predicting performance with respect to SMAPE and MAE criteria.
Highlights
The goal is to predict the solar-generated power one-step ahead to be used in dynamic control of solar + energy storage systems for solar intermittency compensation
Due to fast response criterion of prediction, the prediction algorithms only rely on historical values of the time series and compare local temporal patterns to make the prediction
A modified version of the blocked cross-validation, which maximizes the use of available data while preserving the temporal order in time series data, is proposed to design the prediction parameters of each algorithm
Summary
Renewable energy resources have been identified as essential resources to meet our energy needs; its capacity to replace fossil-fuel-based power generation has been hampered by its intermittency and the difficulty of predicting its availability [1]. Including renewable energy as part of our energy supply requires reliable prediction of its availability for power generation. Employing prediction techniques would yield higher performance of the real time control of renewable generating plants as well as compensating devices. The multivariate methods usually estimate the solar power based on multi-input parameters, which influence solar power generation such as solar irradiance, cloudiness and clearness indices, temperature, wind speed, relative humidity, etc. For high-speed dynamic control, which requires short-term solar power prediction, univariate methods are more effective as they do not rely on a prolonged data acquisition process. Univariate methods only look at previous recorded data for predictions, there is usually a tradeoff between accuracy, cost and speed of the prediction methods
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
