【摘要】 针对海上风电场双馈发电机的功率调节方式和可调范围,以风电场线路损耗最小为优化目标,通过改进粒子群算法获得功率分配优化数据,并以此优化数据作为人工神经网络聚合模型的学习数据。以风电场出口电网侧有功、无功需求和各风机的风速作为人工神经网络的输入,将各双馈发电机直接功率控制的有功和无功参考值作为输出,搭建了某海上风电场34台3 MW双馈发电机构成的四层36-102-102-68节点人工神经网络聚合模型。仿真结果表明:所提出的人工神经网络聚合模型可用于模拟风电场双馈发电机最优功率分配,具有较高的拟合精度;同时通过比较不同算法,得出改进粒子群算法对减小线路损耗具有显著效果。更多还原

【Abstract】 Aiming at maximizing line loss of the offshore wind power plant, power distribution optimization data were obtained through the improved particle swarm algorithm under consideration of the power regulation method and adjustable range of DFIG, and the optimized data were used as the learning data of the aggregation model of the artificial neural network. A four-layer 36-102-102-68 node aggregation model of the artificial neural network composed of 34 sets of 3 MW DFIG was established at an offshore wind power plant, taking the active and reactive power demand of the plant exit grid side as well as the wind speeds of the wind power generators as the input of the artificial neural network and using active and reactive reference values of the direct power control(DPC) of DFIG as the output. Simulation results showed that the proposed aggregation model for the artificial neural network could be applied to simulate optimal power distribution of DFIG in offshore wind plants with high fitting accuracy. Furthermore, comparison of different algorithms indicated that the improved particle swarm algorithm could produce significant effects on the reduction of line loss.更多还原