Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.745174
Title: Fault detection and performance analysis of photovoltaic installations
Author: Dhimish, Mahmoud
ISNI:       0000 0004 7232 5269
Awarding Body: University of Huddersfield
Current Institution: University of Huddersfield
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
The cumulative global photovoltaic (PV) capacity has been growing exponentially around the world, especially due to the installation of grid connected photovoltaic (GCPV) plants. Fault detection and analysis are important for the efficiency, reliability and safety of solar photovoltaic (PV) systems. Even This thesis reports the results of the research work conducted to invent novel fault detection algorithms and evaluate their deployment in multiple existing PV installation, and empirically validate their performance. A major contribution of this thesis is the development of PV fault detection algorithms based on two indicators named power ratio (PR) and voltage ratio (VR). Both ratios are used to identify the type of the fault that occurs in the PV modules, in PV string, and/or in maximum power point tracking (MPPT) unit. Three AI based algorithms were also used to detect faults in PV modules. The first algorithm uses six regions of the power and voltage ratio in order to detect faults in PV systems. The average detection accuracy for the algorithm is equal to 94.74%. However, Mamdani Fuzzy Logic system has been used to enhance the occurrence of fault detection in the PV installations which resulted in an increase to 99.12%. The second proposed PV fault detection algorithm detects defective bypass diodes in PV modules using Mamdani Fuzzy Logic. Whereas, a third PV detection algorithm is based on artificial neural networks (ANN) networks. Four different ANN models have been modelled, which can be classified as follows: - 2 inputs, 5 outputs using 1 hidden layer - 2 inputs, 5 outputs using 2 hidden layers - 2 inputs, 9 outputs using 1 hidden layer - 2 inputs, 9 outputs using 2 hidden layers The output results for the last ANN network had the highest overall fault detection accuracy of 92.1%. In this thesis, the development of two hot spot mitigation techniques used in PV modules will be discussed. These techniques are capable of enhancing the output power of PV modules which are affected by hot spots and partial shading conditions. The detection of hot spots was captured using i5 FLIR thermal imaging camera. Finally the thesis describes the impact of PV micro cracks on the output power of PV modules. A new statistical analysis approach using T-test and F-test was used to identify the significance impact of the cracks on the output power performance of the PV modules. This is developed using LabVIEW software.
Supervisor: Holmes, Violeta ; Mehrdadi, Bruce ; Dales, Mark Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.745174  DOI: Not available
Keywords: T Technology (General)
Share: