风力发电机齿轮箱故障诊断技术研究

发布时间：2018-04-20 09:40

本文选题：风力发电机 + 齿轮箱　；参考：《湖南大学》2011年硕士论文

【摘要】：当前,环境污染和能源问题日益突出,风力发电己成为世界各国更加重视和重点开发的能源之一。随着大型风力发电机组装机容量的增加,其系统结构也日趋复杂,当机组发生故障时,不仅会造成停电,而且会产生严重的安全事故,造成巨大的经济损失。风力发电机的电气故障信号信噪比较大,采用常规的方法就能较好地识别故障。而其机械系统是一个复杂的时变非线性系统,且受到风场不稳定气流等因素的影响,其故障振动信号是非平稳的,使用常规方法难以有效提取故障特征。研究使用先进的诊断技术对其进行故障诊断,具有缩短维修时间、降低维修成本,避免发生严重事故的重要意义。多尺度线调频基稀疏信号分解是一种新的信号处理方法,经验证在时频分析方面效果优越;广义解调可以将时频分布呈曲线变化的多分量非平稳信号转化为时频分布平行于时间轴的平稳信号,因此非平稳信号经广义解调后满足傅里叶分析对平稳性的要求;包络阶次谱方法采用等角度采样,采样信号具有角度域的平稳性,因此也适用于变转速工况下的风力发电机旋转部件故障振动信号分析。本文在国家高技术研究发展计划(863计划)课题“大型风力发电机组状态监控与故障诊断技术研究(项目批准号:2009AA04Z414)”和国家自然科学基金项目“多尺度线调频基稀疏信号分解方法及其在机械故障诊断中的应用(项目批准号:50875078)”资助下,围绕多尺度线调频基稀疏信号分解方法及其在风力发电机齿轮箱故障诊断中的应用进行了理论研究和实验分析。论文主要研究工作和创新性研究成果有: (1)根据风力发电机的结构以及风场的环境特点,研究了风力发电机齿轮箱的故障产生机理。基于风力发电机组的变转速运行特点,建立了风力发电机故障模拟实验系统,并通过数据分析表明,准确的提取、识别振动信号的调制特征是对风力发电机的齿轮箱进行故障诊断的关键。 (2)针对变转速工况下风力发电机齿轮箱振动信号的瞬时频率提取困难问题,将基于多尺度线调频基的稀疏信号分解方法引入到风力发电机齿轮箱振动信号分析。介绍了基于多尺度线调频基的稀疏信号分解原理,并通过仿真算例和实验信号分析证明了该方法在处理瞬时频率连续变化的信号时具有较强的抗噪能力和时频聚集性,非常适合于变转速工况下风力发电机齿轮箱振动信号的分析处理。 (3)针对广义解调时频分析方法中对频率呈曲线变化的多分量非平稳信号相位函数提取困难的问题,将基于多尺度线调频基稀疏信号分解方法引入广义解调,有效解决了相位函数提取问题。仿真算例和应用实例表明,该方法能准确提取相位函数,将多分量非平稳信号解调为平稳信号,适用于风力发电机变转速工况下的齿轮箱故障诊断。 (4)针对包络阶次分析中转速提取困难和基于瞬时频率的转速估计方法精度不理想问题,采用基于多尺度线调频基的稀疏信号分解方法来获取转速信号,在此基础上再对风力发电机齿轮箱振动信号进行包络阶次分析。仿真算例和应用实例表明,该方法能准确提取瞬时转频,解决了传统阶次分析方法中转速测量设备安装困难问题,节约了诊断成本。多尺度线调频基稀疏信号分解算法在分解瞬时频率呈曲线变化的多分量非平稳信号方面效果明显,本文将其引入非平稳振动信号的广义解调分析、包络阶次分析能有效应用于风力发电机齿轮箱系统的故障诊断。
[Abstract]:At present, the problem of environmental pollution and energy is becoming more and more prominent. Wind power has become one of the more important energy sources in the world. With the increase of the capacity of the large wind turbine assembly machine, its system structure is becoming more and more complex. When the unit fails, it will not only make a power outage, but also cause serious safety accidents, causing huge accidents. Big economic losses.
The signal to noise of the electrical fault signal of the wind turbine is relatively large, and the fault is well recognized by the conventional method, and its mechanical system is a complex time-varying nonlinear system, which is influenced by the unstable air flow of the wind field and other factors. The fault vibration signal is nonstationary, and it is difficult to extract the fault features effectively by the conventional method. The use of advanced diagnostic technology for fault diagnosis has the important significance of shortening maintenance time, reducing maintenance cost and avoiding serious accidents.
Multiscale linear frequency modulation based sparse signal decomposition is a new signal processing method. It is proved to be superior in time-frequency analysis. The generalized demodulation can transform the time-frequency distribution multicomponent nonstationary signal into a stationary signal parallel to the time axis, which satisfies Fu Li after the nonstationary signal is demodulated by the generalized demodulation. The blade analysis is required for stability, and the envelope order spectrum method adopts equal angle sampling, and the sampling signal has the stability of the angle domain. Therefore, it is also suitable for the analysis of the fault vibration signal of the rotating parts of the wind generator under the variable speed condition. This paper is in the state high technology research and development plan (863 plan) "large wind turbine state". Research on monitoring and Fault Diagnosis Technology (project approval number: 2009AA04Z414) and National Natural Science Fund Project "multi scale linear frequency modulation based sparse signal decomposition method and its application in mechanical fault diagnosis (project approval number: 50875078)" supported by multi scale linear frequency modulation based sparse signal decomposition method and its wind power generation The application of machine gearbox fault diagnosis is studied theoretically and experimentally.
The main research work and innovative research results are as follows:
(1) according to the structure of the wind turbine and the environment characteristics of the wind field, the mechanism of the fault generation of the gear box of the wind generator is studied. Based on the variable speed running characteristics of the wind turbine, the fault simulation experiment system of the wind turbine is set up, and the accurate extraction and identification of the modulation characteristics of the vibration signal are shown by the data analysis. The key of fault diagnosis is the gearbox of wind turbine.
(2) in view of the difficulty in extracting the instantaneous frequency of the vibration signals of the gearbox in the wind generator, the sparse signal decomposition method based on the multi scale linear frequency modulation (FM) base is introduced to the analysis of the vibration signal of the gear box of the wind turbine. The analysis of signal analysis shows that the method has strong anti noise ability and time frequency aggregation when dealing with continuous change of instantaneous frequency. It is very suitable for the analysis and processing of the vibration signal of the gear box of wind generator in variable speed condition.
(3) aiming at the difficulty of extracting the phase function of the multicomponent nonstationary signal with varying frequency in the generalized demodulation time frequency analysis method, the multi scale linear frequency modulation based sparse signal decomposition method is introduced into the generalized demodulation, which effectively solves the problem of phase function extraction. The simulation calculation example and the application example show that the method can be extracted accurately. The phase function demodulates multi-component non-stationary signals into stationary signals, which is suitable for gearbox fault diagnosis of wind turbines under variable speed conditions.
(4) aiming at the difficulty of speed extraction and the accuracy of speed estimation based on instantaneous frequency in the envelope order analysis, the sparse signal decomposition method based on multi scale linear frequency modulation (FM) base is used to obtain the speed signal. On this basis, the envelope order analysis of the vibration signals of the gear box of the wind turbine is analyzed. The simulation example and Application The example shows that the method can accurately extract instantaneous frequency conversion, and solves the difficulty of installation of rotating speed measuring equipment in traditional order analysis method, and saves the cost of diagnosis.
The multiscale linear frequency modulation (FM) sparse signal decomposition algorithm is effective in decomposing the multi component nonstationary signal with a curve changing instantaneous frequency. This paper introduces it to the generalized demodulation analysis of the non-stationary vibration signal. The envelope order analysis can be used in the fault diagnosis of the gear box system of the wind turbine.

【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3;TM315

【参考文献】