Algorithm Research And Development Of CMP

The Coriolis Mass Flowmeter (CMF) is widely used in industry, because it has many advantages--high measurement accuracy, being able to measure fluid mass directly and measure various parameters of fluid. The original CMF signal processing is based on analog circuits. As analog signal processing system is riddled with defects, several companies and research institutions put in a lot of manpower and resources to study digital signal processing methods and systems. CMF is composed of primary instrument and secondary instrument. Primary instrument includes vibrating tube, exciter and sensor. While secondary instrument is to deal with signals from the sensors, which is used to detect the vibration of the tube. The mass flowrate is proportional to the phase difference of the signals from primary instrument.The purpose of this paper is to develop a new generation and top-level Digital Coriolis flowmeter with high performance and accuracy. In this paper, the implementation of embedded system on the Second Instrument (based on counters) and the Digital Signal Processing methods for it have been studyed. In the aspect of system development, the paper introduces the software design of embedded systems and key technologies in implementation. A method of incorporating UML into embedded system development is described, leading to the whole system is modeled clearly and efficiently. Technologies, such as shared memory, semaphore and scheduling algorithms are used. In the aspect of algorithm, details of traditional DFT approach applied on Coriolis flowmeter signals are introduced. And then several new signal processing algorithms for Coriolis flowmeter are introduced. Chip-Z algorithm is used on the signal frequency measurement initialization and the frequency tracking. The method of intensive sampling only in Narrowband ensures greater accuracy improvement and reduces computational time; At the Measuring stages of phase difference, method of grouped filters and vector inner product are used, firstly filtering the two signals with grouped filters, and then using vector inner product on the two signals for the phase difference calculation.The simulation in MATLAB environment indicates that we can obtain high frequency tracking accuracy and reduce computational time through Chip-Z algorithm, and it is feasible to use grouped filters algorithm to deal with the noise and vector inner product algorithm to calculate the phase difference between the two signals.