Signal Processing Of Vortex Flowmeters And Nonlinear Dynamic Modeling Of Sensors

Digital signal processing systems of Vortex flowmeters are developed by the modern digital signal processing technique in order to improve the measurement precision and enhance functions of meters. The Butterworth wavelet filter with good magnitude-frequency characteristics is adopted to process the vortex flow signal, and the fundamental frequency of signal is estimated effectively. The algorithm is implemented in a digital signal processing system based on a TMS320VC5409 DSP chip, and its effectiveness is proved. Using a TMS320LF2407A integrated many peripherals on the chip, another digital signal processing system with small size and low cost is developed by the spectral analysis and correcting algorithm for practical applications. The results of field experiments show that this system has excellent performances and powerful functions.Nonlinear dynamic process of sensors is too complicated to be modeled. The Volterra series can described the nonlinear dynamics process integrally, but it is difficult to identify its nonlinear transfer functions. The two-order Volterra model can be identified with the correlation analysis and frequency-domain analysis when the input is Gaussian White noise. The block-oriented model can decomposed the nonlinear dynamic process into a linear dynamic sub-system and a nonlinear static sub-system in series, which is convenient to correct the nonlinear dynamic process. An identification method for the block-oriented model is proposed in this paper, which requires a group of inputs with different magnitudes. The identification method can be easily realized in experiments.