Research On Fault Monitoring Technology Of Thermal Power Fan Equipment Based On Data Mining

The maintenance technology of power plant auxiliary equipment is an important guarantee for the safe and stable operation of the generator set,and the traditional equipment regular maintenance method in the power station is increasingly showing its limitations.The shutdown accident caused by equipment failure has brought huge losses to the operation of the power station.The way of equipment maintenance is more and more important to improve the safety and economy of equipment.With the widespread application of automation technology,massive amounts of data to be analyzed are generated during the operation of thermal power stations,which has enabled the application of equipment condition monitoring technology based on data mining.By analyzing and calculating the data,the operation mode of the equipment can be obtained,which can help spot potential failures in time,change from passive to active,and provide early warning and maintenance diagnosis for the equipment to ensure the safe and stable operation of the power station.As a boiler auxiliary equipment of the power station,the fan system provides the air required for combustion and exhausts the combustion gas in time,which provides a guarantee for the safe and stable combustion of the fuel in the furnace.Based on the historical operating data of a power station in Fujian,this paper combines data mining methods to study the fault monitoring technology of wind turbine equipment,and proposes a data mining-based wind turbine fault monitoring technology implementation scheme.First,this article studies association rules and their two major applications in discovering fan operating modes.The first is to introduce the classic association rule algorithm and apply it to the analysis of common faults of wind turbines to discover the characteristic patterns of faults of wind turbine equipment.The second is to explain the limitations of the classic association rules,introduce a clustering algorithm to improve it,and apply the improved method to the normal operation data of the fan to discover the operation mode of the fan under normal operating conditions.The discovered fan mode is saved in the mode library for reference by field personnel,which can guide the on-site operation and fault diagnosis of the fan.Secondly,this paper studies the application of non-linear state estimation in fan equipment failure early warning.Based on the historical normal data of the wind turbine,a non-linear state estimation algorithm is used to estimate the current operating mode of the wind turbine,and the similarity theory is used to determine the operating status of the equipment.Validated by actual fault data,the early warning model can effectively capture the potential failure of the wind turbine and realize early warning of the failure of the wind turbine equipment.