Fault Diagnosis Of Air Compressor Valve Leakage Based On Acoustic Emission Signal

Air compressor is the main power component of many large-scale industrial systems,and its operation status is directly related to the safety and stability of the system.Its equipment is large in scale and expensive in price.Once a failure occurs,it will not only cause huge economic losses,but also bring huge hidden dangers to production safety.Air valve is the key component of air compressor intake and exhaust system,and it is also one of the most vulnerable parts to failure.If air valve leaks,it will cause air compressor cylinder pressure to decrease,output power to decrease,power consumption to increase,seriously affecting the safe operation of air compressor,so it is very necessary to monitor and diagnose its working status.In this paper,the application of acoustic emission monitoring technology and vibration monitoring technology in gas valve leakage monitoring of piston reciprocating air compressor is studied by combining theoretical analysis with experimental analysis.The main contents are as follows:(1)Designed the air compressor valve leakage fault simulation test-bed,designed the experimental acquisition scheme according to the valve leakage fault mode,collected the acoustic emission signal and vibration signal of the valve at different measuring points under different conditions,and provided the data basis for the follow-up work.(2)Acoustic emission and acceleration signals are analyzed and extracted from three aspects: time domain,frequency domain and time-frequency domain,and the differences of signals are studied from the angle of signal similarity.The differences between fault signals and normal signals are measured by Euclidean distance,common subsequence length and KL divergence.(3)The BP neural network fault identification model is established.Different characteristic parameters are extracted for the different characteristics of acoustic emission signal and acceleration signal respectively as input of the model,and the leakage fault state of the valve is identified,which verifies the validity of the model.