| Capacitance coupled electrical resistance tomography(CCERT)is a novel electrical tomography technology that combines resistance tomography(ERT)and non-contact conductivity measurement(C4D)methods for multiphase flow measurement.By measuring conductivity through capacitive coupling,CCERT technology avoids the problems associated with direct contact between metal measuring electrodes and the solution being measured in traditional ERT technology.This results in higher reliability and stability,making CCERT a more promising technology for multiphase flow measurement.Despite these advantages,there is still limited research on CCERT,which has not fully exploited its potential and is yet to be industrialized.Therefore,to accelerate the development and practical application of this technology,this research aims to focus on the following contents and innovations of CCERT.(1)Optimized the design of CCERT sensor.By analyzing the background noise affecting the measurement of CCERT sensor,the connection between the structural parameters of the sensor and the measurement performance was explored by means of finite element simulation.The designed CCERT sensor uses a pipe material of rigid PVC with an outer diameter of 55 mm,a thickness of 2.2 mm,a metal measuring electrode of 180 mm long and a tension angle of 26°.(2)A 12-electrode CCERT measurement system was developed.The developed measurement system uses 500KHz AC voltage signal excitation and has an acquisition rate of more than 160 fps.The hardware and software units including quadrature signal generator module,excitation and detection switching module,analog phase sensitive demodulation module,analog-to-digital conversion module,high-speed USB transmission module,and power supply module were designed and fabricated.The performance of the developed system prototype is verified.The verification results show that the system has a high acquisition rate,and the equivalent resistance measurement data and system imaging can respond well to the change of the medium conductivity in the pipe.(3)Image reconstruction and void fraction measurement of gas-liquid two-phase flow in CCERT are completed by combining compressed sensing(CS)theory.Research on the current CCERT image reconstruction algorithm is relatively scarce,and there are problems such as low resolution of reconstructed images and difficulty in balancing image reconstruction quality and speed.This paper selects suitable sparse bases,optimizes the sensitivity matrix using three methods,and constructs a CCERT image reconstruction model under the framework of compressive sensing theory.Then,three algorithms under the compressive sensing theory are used to solve the model,and two typical algorithms are selected as comparisons.The results show that the GPSR-BB algorithm performs well in terms of reconstructed image quality and noise resistance,and can meet certain real-time requirements.Finally,the GPSR-BB algorithm combined with the optimal sensitivity matrix optimization method is used to complete the CCERT gas-liquid two-phase flow void fraction measurement experiment,and the maximum absolute error in the 100 experimental data of stratified flow and homogeneous flow is 6.59%.(4)The flow type identification of CCERT gas-liquid two-phase flow is investigated based on the contribution of the overcomplete dictionary.First,a total of 5200 sets of simulation samples and 1900 sets of experimental samples are obtained for all flow types including atypical flow patterns.Secondly,the data-driven K-SVD and compression-aware OMP algorithms are used to construct an over-complete dictionary of all the flow patterns,and the flow type identification algorithm is studied based on the idea of contribution of the over-complete dictionary.Finally,the simulation and static experiments of flow type identification for gas-liquid two-phase flows are completed,and the experimental results show that the flow patterns have higher than 90%recognition accuracy except for two bubble flow type. |
PDF Full Text Request