摘要
为了使交流励磁变速恒频风力发电系统在无需设置风速计的情况下,双馈电机实现最大风能跟踪的目的,提出了一种新的基于神经网络的无风速检测方案。通过利用高斯径向基神经网络预测为最大风能追踪提供了一个准确的风速数据,从而准确获取最大功率点下对应的电机转子转速,以减少系统的损耗和成本,提高系统的可靠性。本文针对3.6MW风力发电系统的数学模型,建立了基于定子磁链定向矢量控制的系统仿真模型。结合风速预估和有功功率、无功功率的解耦控制实现了最大风能跟踪。
To provide adequate wind speed information in variable speed constant frequency （VSCF） wind power generation without anemometers, a novel sensorless wind speed estimation through neural network is presented to track maximum wind energy or maximum output power point （MPPT） with doubly fed induction generator （DFIG）. The wind speed is estimated by Gaussian radial basis functions （RBF） Neural Network and the rotor speed is decided accordingly to complete the MPPT. This kind of control systems is less costly and more reliable. The mathematical model of 3.6 MW wind power system is analyzed and the simulation model is set up by stator flux orientation vector control technique. Maximum wind energy is tracked in this sensorless system by decoupling control of active power and reactive power.
出处
《电机与控制学报》
EI
CSCD
北大核心
2010年第2期20-24,共5页
Electric Machines and Control
基金
山东省科技攻关项目(2007GG10007006)
关键词
风力发电
风速预测
最大风能跟踪
径向基神经网络
wind power generation
wind speed estimation
maximum output power point
RBF neural network
