A Use of Artificial Intelligence for Improving PV Array Performance (Empirical Approach)

Abstract

This paper reflects research to design an Artificial Intelligence (AI) platform for controlling the cooling of Si-based PV modules in commercial solar arrays. The AI technology used for this purpose is based on Adaptive Multi-Agent Systems (AMAS). The information fed to the control platform combines locally-collected and exogenous sources, such as grid data and weather forecasts. Elaborating on an existing PV modules cooling equipment driven by traditional industrial algorithms, the AI control platform substantially improves the use of information to make or recommend operational decisions. Thanks to AMAS techniques, the platform can anticipate events by analysing a broader range of information, thereby improving the economics of PV modules cooling. Introducing Artificial Intelligence in this and related areas of solar plant operations has the potential of further improving resource utilization and maintenance efficiency.