The existing approaches for analyzing monitoring data are based on ratios between the power of a PV system and the solar radiation that it receives at a given moment. Many of the possible causes of failures or performance problems that can provoke energy losses are not possible to detect from this kind of approaches.
Our toolbox makes extensive use of cross-relations between spatial and historical comparisons of PV production data. It extracts information from the evolution of the performance of the PV systems over time, and use cross-relations between its components, such as inverter and PV modules.
WebPV builds an Artificial Neural Network (ANN) between all the PV systems present in the database so that any information obtained at an installation can be used to complete the information gathered at another installation constituting the neural network. The ANN allows crossing multiple correlations between its thousands of PV systems, and point out much more accurately the causes of the performance failures that would be otherwise impossible to discover.
In other words, existing approaches are providing a series of bloody pictures to their client, while we provide the whole movie and we identify the crime scene, the victims and the killer. So to speak, we practice high-level criminology.
WebPV provides a toolbox (software) that analyzes the data coming from the already existing data acquisition systems (hardware). It returns a complete performance analysis of the PV systems, including the detection of hidden problems affecting energy performance and PV system lifetime. The toolbox provides these analyses as a service for monitoring companies, PV systems installers or PV systems owners who already possess the monitoring hardware.