Research On Monitoring Signal Analysis Method Of Long Distance Light-Sensitive Oil And Gas Pipeline

With the advanced development of science and technology constantly,many new and clean energy such as solar energy,wind energy,appear in our life,but at this stage of oil and natural gas resources is still needed for supporting industrial development and people's life and production of the main energy,and oil and gas reserves and requirements on the regional distribution of serious imbalance.Looking at the safety accidents of oil and gas pipelines in recent years,it has become a hotspot for researchers to ensure the safety of oil and gas pipelines.Therefore,it is urgent to find a real-time monitoring system for pipeline safety and find appropriate algorithm support.In view of the importance of the monitoring system of the optical fiber oil and gas pipeline,it is urgent to develop a strong algorithm support.In order to obtain the real-time vibration information along the oil and gas pipelines,a signal acquisition platform of the optical fiber monitoring system was built.FPGA and DSP were used as the core components of the hardware platform to complete the data collection of three typical invasion forms of the optical fiber line(mechanical mining vibration,oil stealing vibration by drilling and manual mining vibration).Do raw data support for subsequent signal processing.The characteristics of three typical intrusion signals in time domain,frequency domain and scale domain are extracted for different types of pipeline intrusion.The short-time signal energy and the short-time overlevel rate are selected as parameters to represent the time-domain characteristics.The results show that the over-level ratio has a good effect to distinguish whether there is vibration or not,but it is difficult to recognize different vibration signals.In frequency domain analysis,the original vibration signal is processed in frames,and the processed data is analyzed in spectrum.The results show that this method can clearly identify the corresponding frequency bands of different invasion signals.Finally,the spectrum of wavelet coefficients and the proportion of wavelet packet energy in the scale domain are selected as the parameters to characterize the scale domain features.The analysis results show that different intrusion signals can be identified accurately according to the kurtosis characteristics of wavelet packets.