摘要
使用基于现场数据的神经网络建模方法建立变桨距风电机组动态模型。鉴于风速高于额定风速时,传统PI控制器由于其固定的控制参数无法对时变性和非线性严重的风电系统进行精确恒功率控制,提出一种基于小世界优化算法的自适应在线整定PI控制策略。该策略利用BP神经网络对被控对象进行在线辨识,为确定控制参数提供精确的实时变化信息,同时引入小世界优化算法实现对PI参数快速、准确在线整定,以达到最优控制效果。仿真结果表明该控制策略可有效控制风电机组在高风速下的输出功率,效果优于单纯的PI控制。
A neural network model of variable-pitch wind turbine was built based on field data. In view of the shortcomings that the traditional PI controller due to its fixed control parameters can not achieve accurate constant power control for severe time-varying and nonlinear system when wind speed is higher than the rated speed, an adaptive PI on-line tuning control strategy based on the small-world optimization algorithm was proposed. The strategy uses BP neural network for the on-line identification of controlled object in order to provide accurate real- time information for the parameter tuning. Meanwhile, the small-world optimization algorithm is used to optimize the PI parameters dynamically so as to seek for optimal control target. Simulation results showed that the pitch control strategy proposed in the paper can effectively stabilize the output power and the control effects of the modified PI control are better than simple PI controller.
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2013年第9期1579-1586,共8页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(50776005)
关键词
变桨距风电机组
神经网络辨识
自适应PI控制
小世界优化算法
在线辨识
variable-pitch wind turbines
neural network identification
adaptive PI control
small-worldoptimization algorithm
on-line identification
