摘要
风电机组的发电性能会随着风电场运行时间的增加而衰退,但这种衰退现象很难被察觉。文章利用自适应神经模糊推理系统建立风电机组的功率曲线预测模型,并计算预测模型和实测结果的残差模型,然后利用Z test进行残差分析并将其p值作为发电性能的指标,进行发电性能诊断。
The longer the operating hours in the wind farm were, the worse the power performance of the wind turbine would be. But this recession phenomenon was hard to be found. The power curve predication model of the wind turbine was established by using the adaptive neural fuzzy inference system（shortened as AN FIS）, the residual model between the computational predication model and actual measurement result was calculated, and then, the Z test was conducted and used to analyze the residual model while its p value would be used as the evaluation index to diagnose the power generating performance.
出处
《电机技术》
2017年第4期18-22,31,共6页
Electrical Machinery Technology
