Optimal management of reactive power sources in far-offshore wind power plants

This paper introduces a new approach for the optimal management of reactive power sources, which follows a predictive optimization scheme (i.e. day-ahead, intraday application). Predictive optimization is based to the principle of minimizing the real power losses, as well the number of On-load Tap Changer (OLTC) operations for 24 time steps ahead. The mixed-integer nature of the problem and the restricted computing budget is tackled by using an emerging metaheuristic algorithm called Mean-Variance Mapping Optimization (MVMO). The evolutionary mechanism of MVMO is enhanced by introducing a new mapping function, which improves its global search capability. The effectiveness of MVMO (i.e. fast convergence and robustness against randomness in initialization and factors used in evolutionary operations) and the achievement of optimal grid code compliance are demonstrated by investigating the case of a far-offshore wind power plant, interconnected with HVDC link.