Research On Telemetry Of Fluid Flow In Pipeline

Flow measurement plays an increasingly important role in nowadays industrial production. Based on techniques of hydroacoustic signal processing, the thesis constructs a technical scheme of remotely measuring the pipeline flow,that is,the signal including flow information is transmitted to a receiver through hydroacoustic channel and then is processed to get the frequency information,finally the flow is obtained in light of Doppler theorem. Main achievements and originalities are listed as follows:1, The thesis proposes a technical scheme of remotely measuring the pipe flow which consists of flow measurement based on ultrasonic Doppler, frequency shift of ultrasonic Doppler signal, signal transmission, detection of frequency shift Doppler signal, extraction of Doppler signal, threshold denoising based on wavelet packet and frequency estimation and so on.2, Detailed investigations are performed on Doppler theorem applied to flow measurement.The thesis pays much attention to the relationship between the flow parameter and the power spectrum of Doppler signal, which is the basis of the model of synthesizing the Doppler signal.The rationality of model is verified and can be used to simulate the flow signal.3,There are several advantages in the Gabor wavelet,which will help analyze hydroacoustic signals by using the wavelet scalogram,thus the basic frequency of transmission and arrival hour of the frequency shift Doppler signal can be detected.It becames the key to realize the remote transmission of flow information.4, The frequency division theorem of wavelet packet decomposition is studied detailedly,which becames the basis of deducing the relationship between the sampling frequency and the number of decomposition levels.It is of great use seeking for the best reconstructing wavelet packets by using infinite norm.Combined with floating thresholds and a new function for threshold processing,the denoising algorithm is proved to be fit for the low signal-to-noise Doppler signal through three evaluation criterions.5, Detailed investigation is executed on five methods of estimating average frequencies.After a series of comparisons on operation quantity and estimation performance,the energy operator algorithm is determined to estimate the average frequencies of denoised Doppler signals.A particular research is devoted to estimating instantaneous frequencies by using the MUSIC algorithm,and simulation results show that this algorithm suits for estimating frequency components of Doppler signals.