Detection and Localization the Open and Short Circuit Faults in PV System: A MILP Approach

Abstract

With increasing the utilization of Photovoltaic (PV) modules, an essential need for fast response and efficient PV fault diagnosis approaches has been growing not only to guarantee system reliability but also to reduce maintenance costs, computational burden, and human effort. For this purpose, in this paper, a mixed-integer linear programming technique is proposed as a new approach for fault detection and diagnosis in the PV array. This approach is capable of identifying, discriminating, and localizing the open circuit and short circuit faults. For this, three non-uniform environmental conditions considering low, medium, and high solar irradiance and module temperatures levels are experienced. Moreover, the PV mathematical model is conducted to calculate the PV array power, which is compared with the measured one to find exactly the faulty module location. The proposed model is validated experimentally using a moderate PV string to identify the possible fault patterns. Additionally, the proposed approach is compared with another two traditional approaches to check the optimization accuracy and computational time during the physical tests.