Electromechanical impedance based strategy for health monitoring of solar cells [EMBARGOED]

Abstract

Structural Health Monitoring (SHM) based on electromechanical impedance (EMI) has been widely used in different engineering domains for detecting faults, especially for structural faults that happen due to many reasons such as various natural conditions and operating cycles. However, the SHM based on EMI technique studies on the structural faults in photovoltaic (PV) solar cells is very limited if not almost non-existent. Insightfully, the PV systems have several sensitive parts that are necessary to build a complete source of energy, for example, solar cells, supports, frames, layers, and others. In fact, any fault in such parts of the PV system negatively affects the whole system which leads to serious disasters. Along with the importance of providing modern and uncostly integrated techniques for detecting unforeseen faults in PV solar cells and maintaining public safety. This study gives a new monitoring system and detecting structural faults in PV solar cells based on EMI technique. The principle of this technique is based on Piezoelectric (PZT) smart materials that allow to sense the structure by sending a high-frequency and a lower voltage to monitor the changes of EMI signatures which eventually reflects changes occurring in the structure. Moreover, one of the main purposes of this study is to explore the effectiveness and significance of the integration EMI technique as a permanent monitoring system in PV solar cells from both simulation and experiment standpoints. The numerical simulation is to test the capabilities of the Finite Element Analysis (FEA) in simulating different cases of EMI application in a healthy and unhealthy state of the solar cells. The unhealthy state of the solar cell is expressed by the presence of structural damages such as cracks, corrosion, and delamination in a solar cell as well as monitoring defects of supports and frames as they are sensitive structures. Furthermore, the work was delved into depth by analysing the effect of different variables such as forms and sizes of cells, positions of PZT as well as the depth and location of the cracks. The Validation of results has been done either by employing published results such as cases of PZT-free patches or experimentally using the new generation of portable touch screen PV520A impedance analyzer. This work was finalized by identifying the location of damage based on machine learning (ML), which was developed as a supervised algorithm that allows to convert the results of EMI technique into a database. The confusion matrix was used to measure the performance of the ML algorithm, where the result was very good with an accuracy of 96.73%. Finally, the dissertation ends with suggestions for future work, where the high cost of providing monitoring technology in different domains that operate based on solar energy, makes EMI techniques one of the main alternatives which can be relied upon in the structures of these domains. Indeed, EMI is a technique that has the capability to be an integrated monitoring part at the lowest cost in PV solar cells.