Detection of Abnormal Status of PV Modules at PV Stations with Complex Installation Conditions

摘要

Fault diagnosis of PV arrays is important to improve reliability, efficiency, and safety of PV stations. Instead of conventional thresholding methods and artificial intelligent (AI) machine learning approaches, an innovative Gaussian Mixture Model (GMM) based fault detection approach is proposed in this paper. GMM is applied to represent the probabilistic distribution functions (PDF) of different PV module output, and based on Sandia PV Array Performance Model (SPAM), an orientation independent vector C is then developed to eliminate the probability distribution differences of power outputs caused by varying azimuth angles and tilt angles. Three methods (a pseudo method, a method of fitting and a method of group testing) are proposed to obtain PDF of the orientation independent variable. Jensen-Shannon (JS) divergence, which captures the differences between probability density of C of each PV module, are generated and used as a fault indicator. Simulation data acquired from SPAM are used to assess the performance of the proposed approaches, which are later compared in terms of ability to detect, the response time and the generalization capability. Results show that the proposed approaches can successfully detects faults in PV systems, but the method of fitting and method of group testing can detect faults more accurately. This work is especially suitable for the PV modules that have different installation parameters such as azimuth angles and tilt angles, and it does not require installation of irradiance or temperature sensors.