基于RBF神经网络的桨距角控制策略  

PITCH ANGLE CONTROL STRATEGY BASED ON RBF NEURAL NETWORK

作　　者：王哲[1] 田艳丰[2] 陈雷[3] 郭庆鼎[1] 

机构地区：[1]沈阳工业大学电气工程学院,沈阳110178 [2]沈阳工业大学信息科学与工程学院,沈阳110178 [3]沈阳工业大学风能技术研究所,沈阳110023

出　　处：《太阳能学报》2011年第5期623-626,共4页Acta Energiae Solaris Sinica

基　　金：辽宁省科技厅基金(2008S169)

摘　　要：针对风力发电系统数学模型复杂,受参数变化和外部干扰严重,具有非线性、时变、强耦合的特点,将神经网络控制引入到风力发电控制系统中,将神经网络控制器作为变桨距系统的控制器,在风速高于额定风速的情况下,根据风速的变化调整桨叶桨距角,从而调节发电机的输出功率,使风力发电机组的输出功率保持稳定。最后,利用Simulink搭建整个控制系统,对系统进行仿真。结果表明,使用RBF神经网络变桨距控制器的风力发电系统确实具有更好的动态性能、收敛速度和静态误差,为今后进一步研究奠定了一定基础。The mathematical model of the wind power generation system is complex, strongly affected by the varying pa- rameter and exterior disturbing, and has the features of nonlinear, time changing and coupling. According to the above characteristics, the Neural Network control was introduced into the wind power system. By using the RBF Neural Network controller as the controller of the variable-pitch system, when the wind speed is higher than the rated speed, the angle and the power output are adjusted according to the change of the wind speed, which makes the output of the wind power unit remain stable. Simulink was used to build the whole control system. The simulation result showed that the new wind power system has the better dynamic function, converge speed and less error.

关 键 词：PID 桨距角 径向基函数 SIMULINK 风力机 

分 类 号：TM315[电气工程—电机]