| The accurate measurement of energy is the basis of energy efficiency goals. Coriolis mass flow meter (CMF) is widely used in many industries where energy needs to be measured, such as oil, natural gas, chemical and pharmaceuticals because it can measure the true mass flowrate directly and accurately. At present, CMF has been successfully applied to the application of continuous steady flow measurement. In real industrial application, CMFs also need to deal with the batch flow and gas-liquid two-phase flow measurement. In the case of batch flow, the flow will change rapidly, and the traditional signal processing method cannot track this change because of slow response speed. In the case of gas-liquid two-phase flow condition, the damp ratio of the flowtube will increase high, the traditional analog drive method cannot maintain the vibration of flowtube because of its limited drive gain and inflexible control method, thus induce the failure of measurement. What’s more, in gas-liquid two-phase condition, the disturbance of entrained gas will lead to large measurement errors and the measurement results will have no sense unless they are corrected. For these reasons, the digital signal processing method, digital drive method and correction method for two-phase flow measurement errors are studied in this paper.The exsiting lattice adaptive line enhancer (ALE) combined with DTFT algorithm has a modest dynamic response and cannot be applied to the batch flow and gas-liquid two-phase flow measurement. To solve this problem, the mutation signal model is utilized to describe the sensor signal under batch flow condition and the ARMA signal model is proposed to describe the sensor signal under gas-liquid two-phase flow condition. Using these two models, the characteristic of ALE combined wih DTFT is assessed. By simulation, the lattice ALE and DTFT processing method is improved through the aspects of DTFT initial interval, frequency updating interval and lattice ALE convergence time. And both the response speed and measurement accuracy when measuring batch flow and gas-liquid two-phase flow have been improved.To solve the problem of micro-bend type and straight type flowtube’s high intrinsic frequency, a digital zero-crossing detection algorithm using Lagrange interpolation method is proposed to deal with the sensor signal. This method possesses the ability of small amount of calculation, fast response speed and high calculating accuracy. Thus it can be used to process the sensor signal under batch flow and gas-liquid two-phase condition.To solve the problem of flowtube stalling in gas-liquid two-phase flow case when using traditional analog drive method, the digital drive method based on MDAC and DDS chip is adopted. The frequency and phase of the drive signal is controlled by a DSP chip through the DDS chip, and the amplitude is controlled through the MDAC chip. The DDS chip can synthesize the sine wave with required frequency and phase information automatically, so the large amount of computation to synthesize the sine waves and the large memory to buffer the sine waves can be saved which ensure that the whole drive method can be realized based on DSP chip. The drive algorithms containing start-up method as well as frequency, phase and amplitude tracking method are also given in this paper.The digital Coriolis mass flow transmitter based on TMS320F28335 chip is developed to realize the digital signal processing method and digital drive method in real time. The start-up test and water flow calibrations are performed. The start-up results show that when connecting the developed transmitter with CMF025 type flowtube, the proposed digital drive method can start-up the flowtube reliablely and shorten the start-up time from 15s when using analog drive method to 2s. The water flow calibration results show that the measurement accuracy is better than 0.1%, the repeatability is better than 0.05% and the turn-down ratio is 15:1 for CMF025 type flowtube and 20:1 for a 15mm micro-bend type flowtube.The experimental study on dynamic performance of CMF under batch flow and gas-liquid two-phase flow condition has been performed to validate the dynamic characteristics. Both the rapidity of signal processing and the timeliness of drive control are studied to improve the dynamic characteristic of CMF. By connecting the developed transmitter with CMF025 type flowtube, batch flow measurement and drive control under gas-liquid two-phase flow condition are tested. The experimental results show that, the response time of the transmitter is less than 0.5s, the accuracy is better than 0.15% for batch flow measurement and 0.1% for steady flow measurement. Meanwhile, the transmitter can track the sensor signal well when gas-liquid two-phase condition happens, thus the reduction of sensor signal is improved. Also, when gas-liquid two-phase condition ends, the flowtube can recover to normal vibration quickly and the problem of large sensor signal overshoot is solved.To reduce the large measurement errors in gas-liquid two-phase case, the BP artificial neural network is proposed to correct the liquid mass flow and mixture density errors.The measurement errors are obtained through gas-liquid two-phase flow experiments. By comparing the characteristics among different artificial neural network, BP neural network has been chosen to model the measurement errors and correct them on-line. The results of on-line correction for two-phase flow measurement show that, for CMF025 type flow, when the liquid flowrate varies from 3 kg/min to 15 kg/min and the GVF varies from 0 to 30%, the corrected mass flow errors can be kept within ±3.5%, while the corrected density errors can be kept within ±2%. |
PDF Full Text Request