Research On The Early Warning Method Of Thermal Bending Failure Of Steam Turbine Generator Set

Steam turbine generator set is the core equipment of thermal power plant, as the high speed rotating machinery, the vibration of the turbine is inevitable. When the vibration exceeds the limit, it will affect the safe and stable operation of the unit. Thermal bending is a kind of common vibration fault, which usually occurs when the unit load is running, which may cause the vibration of the unit to be forced to stop.In this paper, a large number of thermal bending faults of turbine generator sets are collected, and the thermal bending faults are studied. Firstly, the cause of thermal bending failure of steam turbine rotor is summarized, and the corresponding mechanism is analyzed, and the characteristics of thermal bending failure are analyzed in terms of amplitude, frequency and vibration. Based on the above research results, this paper proposes a BP neural network iterative multi-step forecasting method and applied to engineering practice, the actual operation data of the power plant is analyzed by using the wavelet denoising method based on EEMD decomposition. The results show that the BP neural network iterative multi-step forecasting method has higher prediction accuracy, better trend and more long-term forecasting results than the general multi-step forecasting method.