Intelligent forecasting system for rooftop solar plants
The dependency of the off-grid customer on a non-conventional energy sources like solar energy is always not suggested due to the intermittent nature of the power generated. Forecasting the output power helps even the amateur solar roof top consumers to make plan of their self-consumption strategies...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIP Publishing LLC
2023-10-01
|
| Series: | AIP Advances |
| Online Access: | http://dx.doi.org/10.1063/5.0165153 |
| _version_ | 1797634182463094784 |
|---|---|
| author | Vasudha Hegde Prajwal S |
| author_facet | Vasudha Hegde Prajwal S |
| author_sort | Vasudha Hegde |
| collection | DOAJ |
| description | The dependency of the off-grid customer on a non-conventional energy sources like solar energy is always not suggested due to the intermittent nature of the power generated. Forecasting the output power helps even the amateur solar roof top consumers to make plan of their self-consumption strategies and also make necessary arrangements for alternate source of energy while minimizing their dependencies on energy storage devices to a great extent. One of the major difficulties faced during the forecasting is procurement of the authentic and accurate digital data since the accuracy of the implemented algorithm depends on many parameters. This work presents a low cost solution for roof top solar plants using intelligent technique. This implemented system has two parts, first one for acquisition and digitization of data with Internet of Things (IoT) and Raspberry Pi and the other part is building a day ahead forecasting model using the deep learning 2D convolution approach with the temporal resolution of 1 h. The data acquisition system collects the power related parameters, such as voltage, current, frequency, active power, and power factor and the related weather parameters, such as temperature, light intensity, and humidity and then appends into Google Sheets using IoT. The error in forecasting the solar power output using these data for 12 months is found to be 11.55%. |
| first_indexed | 2024-03-11T12:03:56Z |
| format | Article |
| id | doaj.art-ae1573b4034b410ab80b4a7f4cda2d13 |
| institution | Directory Open Access Journal |
| issn | 2158-3226 |
| language | English |
| last_indexed | 2024-03-11T12:03:56Z |
| publishDate | 2023-10-01 |
| publisher | AIP Publishing LLC |
| record_format | Article |
| series | AIP Advances |
| spelling | doaj.art-ae1573b4034b410ab80b4a7f4cda2d132023-11-07T17:47:01ZengAIP Publishing LLCAIP Advances2158-32262023-10-011310105102105102-810.1063/5.0165153Intelligent forecasting system for rooftop solar plantsVasudha Hegde0Prajwal S1Department of E&EE, Nitte Meenakshi Institute of Technology, Bengaluru, Karnataka, IndiaREConnect Energy Solutions Pvt. Ltd, Bengaluru, Karnataka, IndiaThe dependency of the off-grid customer on a non-conventional energy sources like solar energy is always not suggested due to the intermittent nature of the power generated. Forecasting the output power helps even the amateur solar roof top consumers to make plan of their self-consumption strategies and also make necessary arrangements for alternate source of energy while minimizing their dependencies on energy storage devices to a great extent. One of the major difficulties faced during the forecasting is procurement of the authentic and accurate digital data since the accuracy of the implemented algorithm depends on many parameters. This work presents a low cost solution for roof top solar plants using intelligent technique. This implemented system has two parts, first one for acquisition and digitization of data with Internet of Things (IoT) and Raspberry Pi and the other part is building a day ahead forecasting model using the deep learning 2D convolution approach with the temporal resolution of 1 h. The data acquisition system collects the power related parameters, such as voltage, current, frequency, active power, and power factor and the related weather parameters, such as temperature, light intensity, and humidity and then appends into Google Sheets using IoT. The error in forecasting the solar power output using these data for 12 months is found to be 11.55%.http://dx.doi.org/10.1063/5.0165153 |
| spellingShingle | Vasudha Hegde Prajwal S Intelligent forecasting system for rooftop solar plants AIP Advances |
| title | Intelligent forecasting system for rooftop solar plants |
| title_full | Intelligent forecasting system for rooftop solar plants |
| title_fullStr | Intelligent forecasting system for rooftop solar plants |
| title_full_unstemmed | Intelligent forecasting system for rooftop solar plants |
| title_short | Intelligent forecasting system for rooftop solar plants |
| title_sort | intelligent forecasting system for rooftop solar plants |
| url | http://dx.doi.org/10.1063/5.0165153 |
| work_keys_str_mv | AT vasudhahegde intelligentforecastingsystemforrooftopsolarplants AT prajwals intelligentforecastingsystemforrooftopsolarplants |