Researchers in Spain have developed a new PV prediction method that uses only direct radiation as a parameter. They found it “comparable, if not superior” to four established forecasting techniques. The method can help homeowners with PV systems decide when to use electricity-intensive appliances and cleaning systems.
A research group led by Spain’s Polytechnic University of Valencia has developed a new single-parameter power prediction method for residential PV installations.
The proposed approach defines interval forecast data rather than absolute numbers, the scientists said, noting that it recognizes and transparently communicates the natural variability in solar PV energy generation.
“The choice of a single-parameter model was a strategic decision aimed at simplifying the prediction process,” the research group emphasizes. “While multi-parameter models may provide more nuanced insights, they often introduce greater computational complexity and resource demands. Our streamlined model promises ease of integration and user-friendliness, crucial for residential users and small-scale PV installations.”
The core aspect of the new method is the selection of comparable days in the past with respect to direct radiation to predict the energy generation of a given day. For each forecast, a confidence level of 80% and a total of 10 comparable days are selected. After identifying similar days, the method uses a quantile-based approach to determine the prediction intervals, setting an upper and lower bound. In statistics, quantiles are used to divide the range of a probability distribution into continuous intervals with equal probabilities.
The system was trained and tested using a case study of a residential installation in Spain, consisting of 12,450 W panels and a 5 kW inverter for self-consumption, all installed in 2018. The hourly PV generation was recorded in the years 2019. 2020, 2021 and 2022. Hourly meteorological data for the area was obtained from the Open Meteo database.
The forecasting technique was used to predict PV energy generation in 2020, based on the algorithm to always search for similar days within a range of two years before the target day. During the same period, it was compared with four classical forecasting methods: linear regression model (Alt1); gradient boosting regressor (Alt2); gradient boost with delays (Alt3); and long short-term memory (LSTM) network (Alt4).
“The performance of the models was evaluated using Key Performance Indicators (KPIs), such as forecast accuracy, forecast interval width, actual confidence level and mean error. This thorough approach ensured a balanced assessment, highlighting the strengths and weaknesses of each method,” the researchers said.
The proposed method achieved a mean absolute error (MAE) of 0.1490 kW, a mean square error (MSE) of 0.0917 kW2, a root mean square error (RMSE) of 0.3029 kW, a mean width of intervals (AWI ) of 0.3365 kW, a coverage probability (CP) of 91.55% and an overall interval error (OIE) of 0.3789 kW. Alt1 showed a MAE of 0.3374 kW, an MSE of 0.2428 kW2, an RMSE of 0.4928 kW, an AWI of 0.9312 kW, a CP of 78.69% and an OIE of 0.4117 kW.
Alt2 had a MAE of 0.2558 kW, an MSE of 0.2044 kW2, an RMSE of 0.4521 kW, an AWI of 0.7464 kW, a CP of 80.12% and an OIE of 0.4031 kW. Alt3 recorded a MAE of 0.1379 kW, an MSE of 0.0768 kW2, an RMSE of 0.2771 kW, an AWI of 0.4890 kW, a CP of 91.72% and an OIE of 0.2355 kW. Alt4 showed an MAE of 0.1282 kW, an MSE of 0.0684 kW2, an RMSE of 0.2616 kW, an AWI of 0.3522 kW, a CP of 80.72% and an OIE of 0.2642 kW.
After analyzing the numerical results, the researchers examined how the proposed approach could help PV system owners achieve energy savings. According to their results, the average monthly energy bill fell from €44.3 ($47.96) to €37.48, while energy imported from the grid fell by 45.79 kWh, from 278 kWh to 232.21 kWh.
“By simply adjusting the operating schedules of the pool filtration system, washing machine and dishwasher to match peak solar energy production times, homeowners have been able to harness more solar energy, reducing dependence on the grid and reducing overall energy costs reduced. they concluded. “With advances in home automation technology, even greater results can be achieved.”
Their findings were presented in “Interval-based solar photovoltaic predictions: a single-parameter approach with direct ray focus,” published on Renewable energy. The group included scientists from the Spanish Polytechnic University of Valencia, the University of Valencia and that of Ecuador Politecnica Salesiana University.
This content is copyrighted and may not be reused. If you would like to collaborate with us and reuse some of our content, please contact: editors@pv-magazine.com.
