Sample entropy-based fault detection for photovoltaic arrays

作     者：Khoshnami, Aria Sadeghkhani, Iman 

作者机构：Islamic Azad Univ Smart Microgrid Res Ctr Najafabad Branch Najafabad *** Iran Islamic Azad Univ Dept Elect Engn Najafabad Branch Najafabad *** Iran 

出 版 物：《IET RENEWABLE POWER GENERATION》 (IET. Renew. Power Gener.)

年 卷 期：2018年第12卷第16期

页      面：1966-1976页

核心收录：

学科分类：0820[工学-石油与天然气工程] 0808[工学-电气工程] 08[工学] 

主　　题：power generation control power generation faults power generation protection photovoltaic power systems entropy fault diagnosis fault currents power grids maximum power point trackers sample entropy-based fault detection photovoltaic arrays photovoltaic systems islanded PV system control scheme protection scheme PV arrays fault currents low-mismatch fault high-impedance fault low-irradiance fault maximum power point tracking algorithm potential fire hazards power loss fault detection scheme sample entropy-based complexity time series normalised fault-imposed component PV power line-to-line fault line-to-ground fault open-circuit fault weather disturbances partial shadings training dataset grid-connected PV system 

摘      要：Despite the growing deployment of photovoltaic (PV) systems, they are still facing challenges in developing the proper control and protection schemes. One of the main protection challenges of PV arrays is their low fault currents under low-irradiance, low-mismatch, and high-impedance faults. In addition to these conditions, the operation of the maximum power point tracking algorithm may lead to the faults within the PV array remain undetected, resulting in potential fire hazards and power loss. This study presents a fault detection scheme based on monitoring the output power of the PV array. Using the sample entropy-based complexity, the irregularity of the time series of the normalised fault-imposed component of PV power is quantified as the fault detection criterion. The proposed protection scheme is capable of distinguishing the line-to-line, line-to-ground, and open-circuit faults from the weather disturbances and partial shadings. Also, it does not require the training dataset and the prior information about the PV array and is effective for both grid-connected and islanded PV systems. Extensive time-domain simulation results demonstrate high accuracy of the proposed fault detection scheme.