PV power forecast using a nonparametric PV model

Abstract

Forecasting the AC power output of a PV plant accurately is important both for plant owners and electric system operators. Two main categories of PV modeling are available: the parametric and the nonparametric. In this paper, a methodology using a nonparametric PV model is proposed, using as inputs several forecasts of meteorological variables from a Numerical Weather Forecast model, and actual AC power measurements of PV plants. The methodology was built upon the R environment and uses Quantile Regression Forests as machine learning tool to forecast AC power with a confidence interval. Real data from five PV plants was used to validate the methodology, and results show that daily production is predicted with an absolute cvMBE lower than 1.3%.

The Photovoltaic Installation Application in the Public Utility Building

“…Angle of 15°, which was adopted for public building allows for the relative annual energy yield of approximately 818 kWh/kWp. For comparison, the annual energy yield of installed PV system for southern Poland is not less than 942 kWh/kWp [18][19][20].…”

Quantile Regression Post-Processing of Weather Forecast for Short-Term Solar Power Probabilistic Forecasting

“…Unfortunately, their discussion of probability forecasting results does not use any standard metrics and is therefore not directly comparable to our results. In [12], the proposed methodology had a skill score between 33% and 36% depending on the plant examined, for five plants located in northern Spain. The hourly forecast for a 662 kW plant in Northern Italy was evaluated at [15] using several forecasting techniques and an ensemble of the forecasts obtained from these models.…”

“…Bracale et al in [11] obtained a probability distribution for predicting the power generated from a Bayesian autoregressive model that takes into account the dependence on solar radiation and some weather variables, such as cloud cover and humidity. Most recently, Almeida et al [12] studied the accuracy of the non-parametric probability forecasts using the quantile regression forest methodology for five PV plants in northern Spain. The authors used different variables of the meteorological model and showed the importance of the radiation data and the scarce utility of the increase in the number of variables used in the modeling.…”

Solar irradiation nowcasting by stochastic persistence: A new parsimonious, simple and efficient forecasting tool

“…One way to solve or to reduce this problem is to 41 forecast this PV output power [9]. A good forecast helps the grid manager to plan the other energy 42 capabilities to compensate for the PV plants power variations [10]. The forecasting quality of the 43 ouput PV plant is strongly linked to the global horizontal irradiation (GHI) forecasting accuracy [11].…”