基于动态时间规整和MSB分析的风电齿轮箱故障特征提取方法研究

Aiming to reduce the influences of planetary gear faults on the performance of wind turbine, this research will focus on the study of frequency and amplitude modulation characteristics, fault signal coupling modulation mechanism and the variation regularity of fault feature under different operating conditions in order to develop an effective signal demodulation and accurate faults quantification method for the condition monitoring and fault diagnosis of planetary gearbox. In this research, the methods of dynamic modeling, dynamic time warping (DTW) and modulation signal bispectrum analysis (MSB) will be applied for the dynamic characteristic analysis of the system and the signal processing algorithm developments. A planetary gearbox dynamics model will be built to simulate dynamic transmission process to study the frequency and amplitude modulation and nonlinear coupling characteristics caused by the gear faults. It aims to investigate the coupling modulation mechanism and its variation regularity under different faults and operating conditions. Considering the low signal-to-noise ratio (SNR) will affect the accuracy of the diagnosis result seriously, a novel SNR enhancement and signal stationary processing method is developed based on the DTW and phase compensation techniques combined with the modeling analysis results. Moreover, according to the mapping relationship between the frequency-amplitude modulation components and fault features, the development of fault feature extraction algorithms will be carried out based on the MSB analysis in order to extract the modulation signal components and fault feature parameters. And then to develop a quantitative diagnostic method for the typical planetary gear faults detection in real time. This research will pave a way for improving the reliability and effectiveness of planetary gearbox fault diagnosis and condition monitoring technology.

为了减少行星齿轮故障对风电机组效能的影响，本课题采用动力学建模分析及基于动态时间规整（DTW）和调制信号双谱分析（MSB）方法，重点研究风电类行星齿轮系统的幅频调制特性、故障信号耦合调制机理及其故障特征变化规律，建立有效的特征信号解调及故障准确量化方法。通过建立故障行星齿轮动力学模型对其传动过程进行理论分析，研究因故障引起的振动幅频调制过程及其非线性特性，揭示故障引起的振动响应之间的相互耦合调制机理和规律。考虑到测试信号的低信噪比将会严重影响诊断精度，结合理论参数模型和仿真分析，开展基于DTW和相位补偿技术的信噪比增强及信号稳态化处理方法研究。同时，研究信号的幅频调制成分与故障特征的映射关系，开展基于MSB分析的故障特征提取方法研究。提取确定表征故障特征的调制信号成分和故障特征参数，建立定量表达行星齿轮典型故障的诊断新方法，从而为提高行星齿轮箱故障诊断技术的可靠性和有效性提供理论基础和技术

风电齿轮传动系统动态响应的调制特性及故障耦合机制是机械故障诊断领域较为重要的研究问题，国内外对这类问题的研究很重视。项目针对行星齿轮传动系统的幅频调制特性、故障信号耦合调制机理及其故障特征变化规律，进行了深入的研究。首先，通过建立系统动力学模型，研究了因故障引起的系统动态响应的幅频调制过程及其非线性特性，揭示了故障引起的动态响应之间的相互耦合机理和规律；然后，利用理论研究成果，结合测试信号的低信噪比特性，提出了基于动态时间规整和相位预估补偿的信噪比增强和信号稳态化处理方法；最后，研究了动态响应的幅频调制成分与故障特征的映射关系，提出了基于调制信号双谱分析的信号解调和故障特征提取方法，建立了定量表达传动系统典型故障的诊断方法。项目取得的一系列研究成果，有效的提高了风电类齿轮传动系统故障诊断的精度和可靠性。