Prediction Of Remaining Life Of Steam Turbine Cylinder Structure Based On Monitoring Data

As the main equipment of the steam-fired combined cycle generator set,large steam turbine units need to meet the requirements of flexible operation and long service life due to frequent start-stop and variable load operation.The service life of durable parts of steam turbines such as cylinders is one of the limiting factors that affect the long-term flexible and safe operation of steam turbines.During the peak-shaving process of the unit,low cycle fatigue is the main factor affecting the service life of the steam turbine cylinder.Frequent heat exchange during operation causes a large temperature difference in the cylinder,and the resulting alternating thermal stress will increase over time.Accumulation leads to low-cycle fatigue of metal materials.In severe cases,it will cause huge economic losses and potential safety hazards.Therefore,in the operation and use stage of large steam turbines,it is hoped that the long-term safe and reliable operation of the steam turbine will be ensured through life prediction,life monitoring and operation maintenance.At the same time,the current industrial production widely uses advanced data collection technology and sensor IoT technology.Large-scale equipment groups,energy-efficient sensors and advanced data collection systems will generate massive equipment monitoring data.Data monitoring and analysis have become mechanical equipment An important means of health status management.In this paper,the low-cycle fatigue remaining life of the steam turbine cylinder structure based on monitoring data is carried out.Taking into account the actual operating conditions of the steam turbine and the complex operating conditions of the cylinder structure,this study combines monitoring data mining and finite element simulation of the failure mechanism,and developed a data physics hybrid modeling method to predict the low cycle fatigue remaining life of the steam turbine cylinder structure.The specific research work includes:(1)Collecting and sorting the turbine operation data and steam turbine cylinder metal wall temperature data,and preprocessing the monitoring data through advanced data cleaning technology to solve the data defects such as sensor abnormalities and noise interference in the monitoring data.The follow-up method application provides accurate data support;(2)Through the long-term operation and maintenance data of the steam turbine unit,the start-up conditions experienced by the unit are discriminated and analyzed,and the start-stop conditions of the unit during operation and use are analyzed;(3)The finite element analysis method is used to accurately analyze the temperature field and thermal stress field of the steam turbine cylinder structure under each typical working condition,and the low-cyclefatigue life loss of the cylinder structure under each typical working condition is obtained;(4)The working condition statistical results of the monitoring data and the calculated life loss situation establish the remaining life prediction model of the steam turbine cylinder structure,and embed the model into the monitoring data management and analysis platform to realize the real-time prediction of the remaining life of the steam turbine cylinder structure.