Time-varying Multivariate Statistical Process Control for Solar Photovoltaic Monitoring and Fault Detecting & Diagnosing Systems

TL;DR

This paper introduces a time-varying multivariate statistical process control method tailored for monitoring and diagnosing faults in solar photovoltaic systems, addressing the challenge of time-dependent variable behavior.

Contribution

It proposes a novel TMSPC approach specifically designed for time-varying processes like solar PV systems, extending traditional MSPC methods.

Findings

Effective fault detection in solar PV systems

Demonstrated with real-world data from Nakhon Phanom University

Improves process monitoring accuracy

Abstract

Time-varying multivariate statistical process control (TMSPC) has been proposed as a tool for process monitoring, fault detecting & diagnosing of time-varying system. It is a modification of multivariate statistical process control (MSPC) of which is designed for the process that variables operated on normal operational condition which independence of time. However, in some processes, such as solar photovoltaic system, the process variables, i.e., temperature, voltage and current, are time-varying. Therefore, TMSPC has been proposed as monitoring and diagnosing tool for time-varying process. The proposed technique has been demonstrated with solar photovoltaic system located at Research & Development Institute, Nakhon Phanom University, Thailand (RDI-NPU).