变速风力发电机组的模糊-单神经元PID控制

Fuzzy Single-neuron PID Control of Variable-speed Wind Turbines

对风力发电机组的数学模型及常规PID控制器进行了分析设计,由于变速风力发电机组的转速变化范围很宽,表现出高度的非线性性和时变性,常规的PID控制器难以在全范围内得到理想的控制性能。为此,提出采用单神经元智能控制器来替代常规的PID控制器,以改善机组控制性能。在分析单神经元控制器的结构和控制原理基础上,为了进一步提高单神经元控制器的动静态性能,引入了模糊控技术,实现了单神经元控制器输出增益的参数自整定。在仿真基础上,建立了一套完整的实验系统,对几种控制方法进行了实验研究。仿真和实验结果表明,基于模糊自整定的单神经元控制器可有效地改善风力发电机组的控制性能,具有较强的自适应能力和鲁棒性。

The mathematical model of wind energy conversion system （WECS） was firstly analyzed and the conventional PID controller was designed. However, due to its widely operational range of rotor speed, the system exhibits high nonlinearity and time variant characteristics, which deteriorates the performance of wind turbines with traditional PID controller. Therefore, a single-neuron intelligent controller was adopted to improve the system performance. By analyzing the structure and operational principle of the single-neuron controller （SNC）, we noted that the output gain of SNC had a great effect upon the system dynamic and static performance. Consequently, a fuzzy controller was introduced to regulate the output gain of SNC online, which could further optimize the WECS. Finally, an experimental platform was established after computer simulation, both simulation and experimental results verify that the proposed fuzzy single-neuron controller （FSNC） has excellent adaptability and robustness and it can greatly improve the performance of WECS.