Software Developments Of Signal Processing Systems For Two Kinds Of Flowmeters

Flow measurement plays an important role in process control. This paper studies the digital signal processing methods and software implements for two kinds of flowmeters. One is vortex flowmeter, and the other is coriolis mass flowmeter.Two key problems in applications of digital vortex flowmeter with spectral analysis are studied carefully in this paper. One is slow response speed, and the other is big output fluctuation when the flow rate is small or the diameter of meter is small. The threshold is set to increase the response speed of the meter. The results of spectral analysis are analyzed for determining the number of zero value in order to solve the output interrupted problem. A series of processing measures, e.g. limiting the amplitude of spectral analysis results, removing the frequency values that exceed the standard deviation, removing the maximal and minimal values, and performing the second average, are proposed to reduce the output fluctuation of meter and improve the stability. Considering large variations of the vortex signal amplitude with the flowrate, a programmable amplifier is built by digital potent- iometers and a low-power operating amplifier to adjust the vortex signal amplitude. The software of a separated type of vortex flowmeter is developed based on the software of an integrative type of meter.The digital vortex flowmeter with the spectrum analysis measures flow-rate according to the signal power dominant principle. In some strong fixed interference circumstances, however, the vortex signal power is not maximal. Therefore a notch filter is designed. Its notch frequency, notch width and notch depth can be adjusted to attenuate the fixed strong disturbance that is within the frequency range of vortex signal, and not to affect the vortex frequency measurement at the same time. The notch filter is used first to remove the fixed strong disturbance, and then the spectrum analysis is employed to estimate the frequency of vortex signal. A dsPIC33FJ256MC710 DSP made by Microchip company is used to build a digital signal processing system for vortex flowmeter, and good experimental results are obtained.Based on our group previous work, i.e. signal modeling and processing for Coriolis mass flowmeter, a digital signal processing system is developed with TMS320F2812 DSP chip, and the adaptive notch filter and sliding Goertzel algorithm are realized to process the time-varying signal in real time. In order to solve the problems of generation and testing of the time varying signal, a testing platform is built with a signal generator, the signal processing system and a person computer. The testing results show that the system and algorithms developed in this paper are effective.