作者：李洋¹, 安平¹, 李志强¹, 刘帅², 马良玉², 刘卫亮²

作者单位：1. 新天绿色能源股份有限公司，河北 石家庄 050000;
2. 华北电力大学自动化系，河北 保定 071003

关键词：非线性状态估计;信息熵;故障预警;归一化;SCADA数据

摘要：

为降低风电机组因机械故障引起的修复成本与风力资源的浪费，提出一种结合非线性状态估计（NSET）与信息熵理论的故障预警算法，使用系统监测数据完成关键设备故障预警，降低设备停机时长。将目标监测参数的前一时刻也作为特征参数之一，并以固定步距挑选历史正常监测数据，组成非线性状态估计算法的记忆矩阵；将改进的信息熵使用范畴进一步限定，并提出递进式故障预警等级，有助于直观了解设备衰退阶段。以风电机组SCADA数据作为数据源，预警发电机驱动端轴承温度高于上限值故障，并探讨不同归一化方法对所提算法的影响，故障算例显示所提算法能够提前预警潜在故障。

Research on fault warning level based on NSET and information entropy
LI Yang¹, AN Ping¹, LI Zhiqiang¹, LIU Shuai², MA Liangyu², LIU Weiliang²
1. China Suntien Green Energy Corporation Limited, Shijiazhuang 050000, China;
2. Department of Automation, North China Electric Power University, Baoding 071003, China
Abstract: In order to reduce the repair costs and waste of wind resources caused by mechanical failure of wind turbines, a fault warning algorithm combining nonlinear state estimation (NSET) and information entropy theory is proposed. The system monitoring data is used to complete the critical equipment fault warning and reduce the equipment downtime. The pre-time of the target monitoring parameter is also taken as one of the characteristic parameters, and the historical normal monitoring data is selected by a fixed step to form a memory matrix of the nonlinear state estimation algorithm; the usage scope of improved information entropy is further limited to complete the fault warning task, and a progressive failure warning level is proposed, which helps to understand the equipment deterioration stage intuitively. Taking the wind turbine SCADA data as the data source, and the fault that the generator drive end bearing temperature is higher than the upper limit value is predicted. The influence of different normalization methods on the proposed algorithm is discussed. The fault study shows that the proposed algorithm can provide early warning aiming at potential failure.
Keywords: nonlinear state estimation;information entropy;fault warning;normalization;SCADA data
2020, 46(7):153-158  收稿日期: 2019-12-21;收到修改稿日期: 2020-02-11
基金项目:
作者简介: 李洋(1989-),男,河北保定市人,工程师,硕士,从事风电机组故障诊断及预警系统研究
参考文献
[1] 兰忠成. 中国风能资源的地理分布及风电开发利用初步评价[D]. 兰州: 兰州大学, 2015.
[2] 刘帅, 刘长良, 甄成刚. 基于数据分类重建的风电机组故障预警方法[J]. 仪器仪表学报, 2019, 40(8): 1-11
[3] CHEN X, YAN R, LIU Y. Wind turbine condition monitoring and fault diagnosis in China[J]. IEEE Instrumentation & Measurement Magazine, 2016, 19(2): 22-28
[4] SINGER R M, GROSS K C, HERZOG J P, et al. Model-based nuclear power plant monitoring and fault detection: Theoretical foundations[R]. Argonne National Lab, IL (United States), 1997.
[5] 郭鹏, 徐明, 白楠, 等. 基于SCADA运行数据的风电机组塔架振动建模与监测[J]. 中国电机工程学报, 2013, 33(5): 128-135
[6] 郭鹏, DAVID I, 杨锡运. 风电机组齿轮箱温度趋势状态监测及分析方法[J]. 中国电机工程学报, 2011, 31(32): 129-136
[7] 郭鹏, 姜漫利. 基于邻比模型分析的风电机组传感器监测研究[J]. 太阳能学报, 2018, 39(5): 1402-1407
[8] 尹诗, 余忠源, 孟凯峰, 等. 基于非线性状态估计的风电机组变桨控制系统故障识别[J]. 中国电机工程学报, 2014, 34(S1): 160-165
[9] 王梓齐, 刘长良, 刘帅. 基于集成NSET和模糊软聚类的风电机组齿轮箱状态监测[J]. 仪器仪表学报, 2019, 40(7): 138-146
[10] 苏连成, 孙伟, 董金国. 基于非线性状态估计的风电机组振动建模研究[J]. 燕山大学学报, 2018, 42(4): 331-341
[11] 李大中, 常城, 许炳坤. 基于样本优化的风电机组齿轮箱轴承温度预测[J]. 系统仿真学报, 2017, 29(2): 374-380
[12] 刘帅, 刘长良, 甄成刚, 靳昊凡. 基于群体多维相似性的风机齿轮箱预警策略[J]. 仪器仪表学报, 2018, 39(1): 180-189
[13] LIU X, XU X, JIANG Z. Application of the state deterioration evolution based on bi-spectrum entropy and HMM in wind turbine[J]. Chaos, Solitons & Fractals, 2016, 89: 160-168