Study On Electrical Resistance Tomography Of Flowrate Measurement For Gas-liquid Two-phase Flow

Flowrate measurement of gas-liquid two-phase flow is widely applied in many industriessuch as power, chemical engineering, petroleum and thermal. It is very important to the controlprocess and energy saving management. Since the room is narrow and environment iscomplex in some special engineering applications, it is very difficult to measure flowrate indynamical thermal equilibrium system. Meanwhile, it is also necessary to measurement flowpattern for the system design. Electrical resistance tomography (ERT) is widely used in theflowrate measurement area because of many advantages, such as its miniaturization, non-intrusive, visible, small flow field disturbances and little environment affection. And it isalways noticed in recent twenty years. However, due to sensitive soft field, ill-posed,reconstruction speed, non-rigid motion and the fluid temperature and pressure factors in theimage reconstruction process, the flowrate measurement precision is low, which seriouslylimits its application. This dissertation purpose is to solve some problems in Electricalresistance tomography, such as sensitive soft field, ill-posed, motion estimation and thetemperature and pressure problems.The main contents of the thesis are as follows:（1）Since the sensing field exists soft-field characteristic, and different electrodestructure and different measure models lead to different electrical distributions. The paperwill study on the method of electrode structural optimization design based on the differentmeasure models through three-dimensional simulation technology. Finally, the electrodestructure parameters in the ERT sensor are provided to realize flowrate measurement.（2）Considering the ill-posed and fast reconstruction problems, it is taken into accountstatistical information including the prior probability of the permittivity distribution and thenoise information. And based on the statistical prior iterative reconstruction algorithm, a fast regularization statistical prior image reconstruction algorithm is proposed. The results showthis algorithm gets better effect in the precision and speed than other iteration methods.（3）Considering the questions of parameter uncertainties and external disturbances in themeasurement system, a construction algorithm based on H infinity filter is researched, whichhas strong anti-disturbance ability and great robustness. Meanwhile, aiming at the weak realtime and computational complexity of interference factor in the traditional H infinity filteringalgorithm, the adaptive H infinity filtering reconstruction algorithm is proposed. It canadaptive update interference factor based on signal to noise ratio. And the article puts alsoforward another new adaptive H infinity filtering reconstruction method. Forgetting factor isintroduced to update state variables in the method. It can improve the correction function ofreal-time measurement values and suppress error propagation. The results show thereconstructed images obtained by two methods have good effect in reconstruction quality,noise suppression and artifacts removal.（4）Considering the dynamic non-rigid motion estimation question of gas-liquid two-phase flow, a construction algorithm based on regularized H infinity particle filter is proposed.Because the method has superior noise suppression and robustness, the particles of theparticle filter are allocated based on the prediction state estimated from H infinity filter. Atthe same time, to overcome the sample impoverishment problem of particle filter, theobservation vector of H infinity filter is regularized. Experimental results show that thealgorithm is feasibility and effectiveness, and it accurately track the centroid positions in thedynamic reconstruction process which can improve the estimation precision.（5）Building measurement platform of gas-liquid two-phase flow in pipeline andcombining the density, void fraction, mass quality and flow measurement results in thedifferent temperature and pressure conditions, the flowrate correlation and its performanceare optimized and compared. Finally, the flowrate equation is established through experimentthe optimization results. Experiments show that it is feasible and stable for flowratemeasurement, which improves the measurement precision.