Research On Fault Diagnosis Of Advanced Adiabatic Compressed Air Energy Storage Unit

Rotor system fault diagnosis is of great significance for the evaluation of the operating status of advanced adiabatic compressed air energy storage units and the determination of equipment maintenance plans.Based on the analysis of the vibration signals of the rotor system,the dissertation studies the fault diagnosis of the rotor system imbalance in the advanced adiabatic compressed air energy storage unit.This paper mainly studies the application of holographic spectrum technology in the diagnosis of temporary thermal bending faults of air compressor rotors,the initial bending and unbalanced vibration response of rotors based on isogeometric analysis,and the diagnostic methods of rotor unbalanced faults based on holographic spectrum and random forest algorithm.The vibration signal of the rotor system contains a lot of information that reveals the operating status of the rotor system.In order to better characterize the running status of the rotor and diagnose the fault,the paper introduces the principle of holographic spectrum technology.For the imbalance of the air compressor rotor,by observing the change of the tangent slope of the initial phase point of the rotor holographic spectrum with the speed,it is accurate.The cause of the imbalance of the rotor of the air compressor was judged,indicating that the holographic spectrum contained information about the rotor’s operating status.In actual measurement,it is found that there is a situation where the rotor is difficult to distinguish between the initial bending and the mass imbalance.The paper derives the linear motion differential equation of the rotor system with initial bending in the rotating coordinate system.By analyzing the vibration response of the rotor with initial bending and mass imbalance,the initial bending of the rotor can be regarded as a continuous distribution.Unbalanced quality.For the imbalance fault of the rotor system,a method for diagnosing the imbalance of the rotor based on hologram and random forest is proposed.The fast Fourier transform of the vibration signal is used to obtain the rotor amplitude-frequency characteristics.The input feature range is initially determined.The sliding window is used to ensure the continuity of the input features.The holographic spectrum is used to extract the features.The obtained sine and cosine terms are used.As the input features,the optimal parameters of the random forest algorithm are obtained using grid search and cross-validation after the initial random forest is generated,and the input features are selected based on the importance score of the input features,and the random is established based on the obtained parameters.The forest classification model has formed a holographic spectrum-random forest method,and then diagnoses the imbalance of the rotor system.The analysis of the results proves that the method has better fault diagnosis accuracy and lower fault diagnosis time.