| In a lot of manufacture processes , the measurement and control of the fluid is a difficult problem which couldn't be solved for a long time. In recent years, with the development of electronic technology and user's demand to the systematic measuring precision higher and higher, ultrasonic wave is more and more extensive as a newer technical application. Because ultrasonic wave can carry velocity message of fluid while traversing in the liquid flowing, so we could get the corresponding fluid noise signal through dealing with the ultrasonic wave signal that is received . The ultrasonic wave's speed in water relates to temperature of water, impurity and depth of water, the intensity of the ultrasonic wave signal is varying with distance. Mainly for the function of liquid molecule, solid particle suspended substance and bubble ,etc., the energy of the sound wave will be absorbed and reflected and the intensity will be weakened. Moreover with the influence of the noise, the signal which had been received is unstable, so we should adopt high-accuracy spread time measure method in measuring the ultrasonic wave signal . With the development of microelectronics , digital signal process and computing technology , the theory and algorithm of digital signal process can be applied widely . Exactly because the development of FFT and many fast algorithm in digital signal process, it is possible to carry on correlation operation in the frequency domain. We could get some information related to flowing by transforming fluid noise signal to frequency domain at the same time and utilizing various kinds of signal analytical methods ,thus we can improve technology of correlation measurement further . Therefore the research of correlation measurement in the frequency domain has great potential value. The article carried on preliminary experiment to this , designed one correlation measurement system based on cross-high-order cumulant. The correlation measurement system has introduced digital signal process technology. It has dealed with fluid noise signal in the frequency domain . The method is to get the cross-high-order cumulant of fluid noise signal at first, then get the Cross-spectral , construct a function , make it time delays is in the peak value place . Compared with the contact-type flowmeter, this system has not only brought convenience for measurement and installation but also improved the precision of the flowmeter and its stability further. The research content in the paper is on-line measurement of the fluid. It has brought forward a new measurement based on the cross-high-order cumulant. Compared with other methods, the measurement based on the cross-high-order cumulant has more outstanding noise-restrain ability. Fast algorithm in digital signal process make it possible that complicated algorithm could apply to reality. The main research contents of this paper are generalized as follows: 1. The basic principle of correlation measurement flowmeters and its mathematics model. It set up model to correlation measurement systems at first, draw its physical model and mathematics model. 2. Developing a hardware system of amplitude modulation and the signal collection .The key work is improve resolution. It composed of an amplitude modulation board , A/D conversion board and power. 3. Selection of the keynel algorithm. Compared cross-high-order cumulant with correlation measurement, we analysed the difference in noise-restrain ability and precision in estimated value and had selected the key algorithm of the system finally. 4. The test and the realization of the algorithm. While constructing function T (Ï„), get rid of the noise element of fluid signal in the matrix, have improved the precision time delay estimation as well as reduced the calculating amount. In presenting the emulation conclusion, through comparison of many kinds algorithm , under appending the zero data sequences to signal data sequences can mend the cross-high-order cumulant, the time delay estimation can still be accurate when SNR is very low. While spectral peak is approached , introduce the Stat. method , draw the accurate time delay estimation through comparative variance. Compared with general approaches, the methods in this paper have many prominent virtues:...
|
PDF Full Text Request