Optimal maintenance policy for multi-level wind farm systems based on distributed model predictive control

This study introduces a distributed optimization framework for maintenance strategy design in multi-level wind farm systems, employing a model predictive control (MPC) approach. The considered system comprises multiple geographically dispersed wind farms, wherein each wind turbine experiences degradation modeled as an independent and identically distributed stochastic process. Preventive maintenance is executed when a turbine's degradation state reaches a predefined threshold, thereby restoring its operational condition. The optimization framework incorporates constraints related to the capacity of available maintenance resources and compliance with regional electricity protocols. By formulating the objective of maximizing production rewards, the optimal operation and maintenance policy is derived. Simulation-based evaluations demonstrate the effectiveness of the proposed methodology in significantly lowering overall maintenance costs while enhancing system reliability.