Prediction-Corrected Multi-Objective Dispatch for Offshore Wind Farm Clusters under Abnormal States

Le résumé fourni par la source

Offshore wind farm clusters operate under distorted available-power boundaries, large output fluctuations, and competing dispatch objectives during typhoons, high-volatility periods, and transmission-channel constraints. This paper proposes a prediction-corrected, multi-objective coordinated dispatch strategy for these abnormal states. The method first uses a hybrid physics-based and data-driven model to correct station-level available-power boundaries and thereby define a more credible dispatch feasible domain. It then coordinates operational safety, generation fairness, transmission-channel utilization, and turbine lifetime protection. A threshold-based framework identifies normal, high-volatility, typhoon-risk, and transmission-channel fault scenarios. For each scenario, the framework assigns an objective-weight vector, and an improved non-dominated sorting genetic algorithm II (NSGA-II) solves the resulting dispatch problem. In the case study, prediction correction achieves a capacity-referenced accuracy (ACC) of 89.748%. In the normal scenario, the method reduces average total ramping by 36.96%. In the high-volatility and typhoon-risk scenarios, it reduces the maximum load-rate range by 34.75% and 44.31%, respectively, and the average load-rate standard deviation by 46.70% and 53.44%, respectively. The results show that the method adjusts dispatch priorities to the dominant constraint in each abnormal state and improves the operational safety of offshore wind farm clusters under complex meteorological and operating conditions.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.