Modeling And Fault Diagnosis Of High Pressure Feedwater Heater For Large-Scale Thermal Power Unit

The operating state of the high-pressure feedwater heater in the regenerative system has an important influence on the safety and economy of the thermal power unit.The establishment of online performance monitoring and fault diagnosis model of the high-pressure feedwater heater is the core of equipment condition maintenance,and has a significant effect on the safe and economic operation of the unit.This paper takes the horizontal high-pressure feedwater heater of large-scale thermal power unit as the research object,and proposes a discreted numerical thermal calculation model of horizontal three-stage high-pressure feedwater heater that can be used for fault simulation and analysis.The online performance of high-pressure feedwater heater is monitored with analytical redundancy method.The operating characteristics of normal and abnormal conditions is analyzed.Based on the model,the operating characteristics of highpressure feedwater heater are studied and the anomalies such as fouling and leakage of highadjusted tube bundles under different conditions are simulated.Using the large number of fault simulation data generated in the process,a high-pressure feed water heating fault diagnosis framework is designed based on the neural network.The main research content includes:Firstly,according to the heat transfer mechanism and the physical structure of the heater,a thermal calculation model of a horizontal three-stage high-pressure feedwater heater is proposed considering the influence of the water level on the distribution of heat transfer area.and is used.The example model of No.1 high-pressure feedwater heater in a 600 MW steam turbine system is establishes.The model has been verified by the operation data to prove its accuracy and adaptability to variance conditions,which provides support for the following chapters on the study of high-performance operating characteristics and fault characteristics.Secondly,based on the thermal calculation model,a series of operational characteristics of high-pressure feedwater heater are analyzed.The effect of inlet boundary parameters on the heat transfer performance under normal conditions is studied.It is found that the inlet temperature and the extraction steam pressure of the feedwater have the most obvious effect on heat transfer performance,followed by the feedwater mass flow and extraction steam temperature,and the feedwater inlet pressure has minimal effect on the heat transfer performance.Based on the study of high-pressure feedwater heater operating characteristics under abnormal conditions of plugging and abnormal water level,the relationship between the high heat transfer performance and the plugging rate and the optimal operating water level range are obtained.Thirdly,based on the thermal calculation model,fault simulation models are established and validated for the common faults of tube bundle fouling and leakage.The characteristics of the faults are studied using simulation data.A fault diagnosis framework for high-pressure feedwater heater is designed.A large number of fault samples are generated using the fault simulation model.A high-pressure feedwater heater fault diagnosis model is established based on BP neural network,and cross-validation and comparison of similar modeling algorithms is performed.Using the fault time series data collected from the power plant,the diagnosis effect of the high fault diagnosis model is verified.Finally,based on.Net framework,SQL relational database and EXA real-time database,the on-line monitoring and fault diagnosis platform and off-line analysis platform for highpressure feed water heaters are designed and developed.