Two-Stage Model-Based Predicting PV Generation with the Conjugation of IoT Sensor Data

Area of Science:

• Renewable Energy Systems
• Artificial Intelligence in Energy
• Sensor Networks and IoT

Background:

• Meteorological data offers broad environmental insights for photovoltaic (PV) power prediction but lacks localized accuracy.
• Existing research often overlooks the critical timing of Internet of Things (IoT) sensor data acquisition for PV forecasting.
• Real-world IoT data availability can be inconsistent, necessitating prior prediction for accurate PV generation forecasts.

Purpose of the Study:

• To develop a novel short-term photovoltaic (PV) prediction scheme leveraging Internet of Things (IoT) sensor data.
• To address the limitations of meteorological data by incorporating localized environmental changes detected by IoT sensors.
• To enhance PV prediction accuracy by proposing a two-stage neural network model that accounts for the temporal aspects of IoT data.

Main Methods:

• A two-stage neural network model is proposed for PV prediction.
• The first stage predicts future IoT sensor data using available environmental data.
• The second stage utilizes both predicted IoT data and environmental data for PV generation forecasting, optimizing prediction schemes at each stage.

Main Results:

• The proposed two-stage model significantly enhances the accuracy of short-term PV power prediction.
• The scheme demonstrates an improvement of over 12% in prediction accuracy compared to baseline methods relying solely on meteorological data.
• The model effectively integrates predicted IoT sensor data for more precise PV generation forecasts.

Conclusions:

• The novel two-stage prediction scheme offers a robust solution for accurate short-term PV power forecasting.
• Incorporating predicted IoT sensor data alongside environmental data is crucial for overcoming the limitations of traditional forecasting methods.
• This approach provides a more reliable method for predicting PV generation, especially in areas with dynamic environmental conditions.