Emerging Trends in Optimisation for Reliability and Maintenance of Photovoltaic Systems

The development of renewable energy especially solar and wind energy over the recent years has gained global attention as an alternative method of generating energy experiencing exceptional growth in its production. Like any standard production system, solar energy generation components are subject to random failure resulting to production and supply of demand interruptions. The absence of a suitable maintenance plan is identified as one of the major causes of production failures. The aim of this study is to define a joint production and maintenance strategy. The strategy adopted consists of selecting components on which maintenance should be performed on during preventive maintenance time. While the criterion for the selection is based on reliability of the components by imposing reliability threshold. A perfect maintenance strategy is applied on the selected components to minimize the maintenance downtime and maximize reliability of the system. The methodology we developed entails solving the problem of energy production and maintenance optimization simultaneously optimization by using machine learning method: artificial neural network (ANN) in order to satisfy a fluctuating energy demand over a finite horizon. A one-year reliability assessment was performed to determine the system reliability. It was observed that the system reliability decreases with increase in the system capacity where it is 79.7% for 100 kW plant and 63.7% for 200 kW plant. Consequently, the optimal maintenance plan (N*) is found to be increasing with increase in the system capacity: N* = 2, 3 and 6 for100 kW, 1000 kW and 2500 kW plants respectively. This also implies that the failure rate of the system will be reduced.