Photovoltaic energy production estimation

As part of the design, development, and exploitation of photovoltaic farms around the world, our client collects, aggregates and manually processes a large amount of meteorological, environmental and electrical data. In order to improve data quality and to accelerate its massive processing, MP DATA was asked to design and implement a tailor-made and optimized tool for its solar engineering experts.

Following a work of qualification and framing of needs, we have implemented a process of automation and optimization of all the data processing stages in order to establish a complete workflow.

Therefore, this workflow was implemented through a turnkey application. The developed solution was integrated into the client's IS and was connected to the various data channels (file systems, client databases, webservices and external APIs).

Before, a handful processes were performed manually (detection and correction of outliers, imputation of missing data, sampling, predictions). Using our knowledge in the Data Science and Time Series fields and our data processing tools, a large number of operations and calculations have been automated: outliers detection, data drift correction, missing data imputation, data sampling and forecast.