Data-Driven Approach for Condition Monitoring and Improving Power Output of Photovoltaic Systems

Increasing renewable energy targets globally has raised the requirement for efficient and profitable operation of solar photovoltaic (PV) systems. In light this requirement, paper provides a path evaluating operating condition improving power output PV system in grid integrated environment. To achieve this, different types faults grid-connected systems (GCPVs) their impact on loss associated with electrical network are analyzed. A data-driven approach using neural networks (NNs) is proposed to root cause analysis localize fault component level system. The localized combined parallel adaptive neurofuzzy inference units (ANFIUs) develop mismatch-based control unit (PMCU) GCPV. framework, 10-kW single-phase GCPV simulated training NN-based anomaly detection 14 deviation signals. Further, developed algorithm PMCU implemented experimental setup results identified 98.2% accuracy 43000 observations/sec prediction speed trained classifier, improved reduced voltage current harmonics operation.