Photovoltaic fault detection and diagnosis based on adaptive-sampling rate and operating voltage approaches / Ahmad Rivai
Over the last few decades, the photovoltaic power generation system (PV system) has gotten a lot of attention since sunlight as its source is clean, sustainable, and abundant. PV system also has lower maintenance and operational costs than other types of power generation systems. On the other hand,...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Published: |
2022
|
Subjects: | |
Online Access: | http://studentsrepo.um.edu.my/14510/2/Ahmad_Rivai.pdf http://studentsrepo.um.edu.my/14510/1/Ahmad_Rivai.pdf http://studentsrepo.um.edu.my/14510/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Over the last few decades, the photovoltaic power generation system (PV system) has gotten a lot of attention since sunlight as its source is clean, sustainable, and abundant. PV system also has lower maintenance and operational costs than other types of power generation systems. On the other hand, it suffers from high degradation rates because of the long-term operation, which is around 0.5 percent per year. Moreover, the PV system is commonly used in outdoor environments which makes it exposed to all kinds of weather, it is susceptible to a variety of faults. The PV system’s faulty drastically reduces power generation, accelerates system aging, and potentially ruins the entire system’s availability. As a result, PV systems fault detection and diagnosis (FDD) are essential to improve safety, reliability, and efficiency.
This thesis presents photovoltaic (PV) FDD based on the adaptive sampling rate and operating voltage approaches. They are offline and online PV FDD methods, respectively. The adaptive-sampling rate method is the offline approach employed for a multi-channel PV I-V curve tracer. The real-time operating voltage method is the online approach for PV string failure analysis. The I-V curve tracer’s performance has been improved using the adaptive-sampling-rate method, especially on data resolution and acquisition speed. Furthermore, more measurement points and a larger range of voltage and current can be measured using the proposed offline method. Quickly comparing and identifying faults in one or more PV modules is critical. A multi-channel PV I-V curve tracer prototype has been developed and used with an Arduino microcontroller and 30 PV modules for conducting the experimental tests. The I-V curve tracer’s performance and functionality have been confirmed by simulation and experiment. The I-V curve tracer prototype can quickly produce a smooth I-V curve within 57 ms with up to 256 measurements.
The PV module FDD which is based on the operating voltage approach is a simple and effective method that is realized by monitoring the real-time operating voltage of each PV module. All PV modules’ operating voltages are monitored using a self-powered wireless sensor network (WSN). A grid-connected PV system is used to test the proposed method. Each PV module is monitored for detecting faulty modules. The impact of various electrical fault scenarios on PV string characteristics is presented in the PV string failure analysis. As compared to the standard module, the results show that the faulty or degraded module has a lower operating voltage. A graphical user interface (GUI) application software that manipulates colours to graphically displays each module's real-time operating voltage has been designed to allow users to identify faulty modules quickly. The PV module electroluminescence (EL) imaging system further validates the faulty module identification method by clearly illustrating individual cell conditions.
Both offline and online methods have successfully eliminated the fault location investigation by the operators and shortened the fault problem-solving time. The offline and online methods reduced one-quarter and two-fifths of steps, respectively, from the conventional failure-solving process.
|
---|