基于统计学分析的齿轮箱油温异常预警

Abnormal Warning of Gearbox Oil Temperature Based on Statistical Analysis

齿轮箱作为风电机组的变速机构,因其结构设计复杂紧凑、部件之间耦合性较强,且长期受到交变载荷和冲击载荷的作用,容易受齿轮磨损、点蚀和轴承表面损伤等故障影响,而齿轮箱油温可以反映齿轮箱各部件的健康状况。本文提出一种基于统计学分析的齿轮箱油温异常预警方法,对工况相似的风机的齿轮箱油温进行对比,可以对齿轮箱存在潜在故障的机组进行预警。首先,获取工况相似的邻近机组;其次,对需要的SCADA特征数据进行数据清洗;最后,通过统计学的计算和分析,对齿轮箱油温异常的机组进行预警并将预警等级划分为高、中2个等级。实验结果表明,本文所提出的基于统计学分析的齿轮箱油温异常预警方法在风电场具有良好的适用性且准确率较高,可以有效预警出风机齿轮箱油温异常,对现场检修有一定的指导作用。

As the variable speed mechanism of wind turbines,gearbox is easily affected by gear wear,pitting corrosion and bearing surface damage due to its complex and compact structural design,strong coupling between components,and long-term alternating load and shock load,and gearbox oil temperature can reflect the health status of each component of the gearbox.In this paper,an early warning method for gearbox oil temperature abnormalities based on statistical analysis is proposed,which compares the gearbox oil temperature of wind turbines with similar working conditions,which can warn the units with potential faults in the gearboxes.Firstly,obtain neighboring units with similar working conditions;Secondly,the required SCADA characteristic data is cleansed.Finally,through statistical calculation and analysis,the unit with abnormal oil temperature of the gearbox is warned and the warning level is divided into two levels:high and medium.The experimental results show that the proposed early warning method of abnormal oil temperature of gearbox based on statistical analysis has good applicability and high accuracy in wind farms,which can effectively warn of abnormal oil temperature of wind turbine gearboxes and has a certain guiding effect on on-site maintenance.