An integrated maintenance and power generation forecast by ANN approach based on availability maximization of a wind farm

Abstract

Climate change, global warming and high costs of fossil fuels caused an urgent need to shift from traditional sources of energy generation to renewable ones. Wind energy has been identified as one of the most attractive renewable sources that supply affordable, inexhaustible, and clean energy to the economy. To guarantee wind plant availability, the maintenance downtime has to be minimized. In doing so, a systematic preventive maintenance strategy integrated with wind power generation forecasted by artificial neural network (ANN) technique was developed by selecting the components which maintenance action will be performed upon based on a set minimum reliability margin. In this work, four components of the turbine (bearing, main shaft, gearbox and generator) were selected to represent the reliability of the turbine. An algorithm was developed to automatically select the component whose reliability falls below the minimum selected margin and a perfect maintenance is carried out by replacing the component. We try to develop the model that yields the optimal number of preventive maintenance (PM) actions that minimizes maintenance downtime which yields maximum availability The uniqueness of the work lies with the consideration and integration of the influence of turbine power generation rate into the PM planning strategy. Finally, numerical example was presented and sensitivity analysis was performed by varying some parameters to validate the algorithm.

Keywords

• Reliability
• Availability
• Imperfect maintenance
• Priority selection
• Artificial neural network (ANN)