论文部分内容阅读
常规的希尔伯特–黄变换(Hilbert-Huang transform,HHT)方法能较好地分析水轮机轴系信号,但经经验模态(empirical mode decomposition,EMD)分解后,在原始信号的低频区易产生虚假的内禀模态函数(intrinsic mode function,IMF)分量,干扰特征信息的提取,引发误判。该文提出基于能量波动的改进HHT方法及其判别条件。该方法利用分量信号能量递减原则并设定判别阈值来跟踪筛选虚假分量。通过仿真信号对该方法进行了有效性验证,并以原型水轮机非最优工况下动态信号为例,进行了应用检验。结果表明,该方法具有良好的虚假分量识别能力,提取真实的水轮机特征参量,更加适合分析复杂而特殊的水轮机动态特征信息。
The conventional Hilbert-Huang transform (HHT) method can well analyze turbine shaft signals, but after decomposed by empirical mode decomposition (EMD), the low frequency region of the original signal A false intrinsic mode function (IMF) component is generated to disturb the extraction of feature information and lead to misjudgment. This paper presents an improved HHT method based on energy fluctuation and its discriminant conditions. The method uses the principle of component signal energy decrement and sets the threshold value to track and filter the false component. The method is validated by the simulation signal, and the application test is carried out by taking the dynamic signal under the non-optimal condition of the prototype turbine as an example. The results show that this method has a good ability to identify false components and extract the actual turbine characteristic parameters, which is more suitable for the analysis of complex and special turbine dynamic characteristics.