风力机系统的神经网络模型辨识

NEURAL NETWORKS MODEL IDENTIFICATION OF THE WIND TURBINE SYSTEM

应用人工神经网络的建模方法，采用多层感知器的模型结构，利用自适应学习速率的ＢＰ学习算法，辨识出风力机系统的功能模型，并把辨识模型的仿真结果与系统实验测量数据相对比，开展了与经典系统辨识方法的比较研究，以检验神经网络模型的可靠性。实验结果表明，这种新的风力机系统建模方法具有很高的精度。

In this paper,the wind turbine system experiment model is presented.The model is made up with a multilayer perception which is trained by adaptive learning rate BP algorithm.The model prediction result is compared to experiment data and also compared to the prediction result which is from linear experiment model,based on modern identification theory.