ECT Image Reconstruction And Sensor Optimization Of Arc Object Distribution

In the actual industrial production process,we often encounter two or more kinds of objects mixed flow conditions,this phenomenon is generally known as multiphase flow in academia and industry.Process Tomography(PT)is an important technique for Process detection in two-phase flow.Electrical Capacitance Tomography(ECT)is a kind of process Tomography technology.It has the advantages of non-invasive,low cost,simple structure,etc.It can be applied as long as the dielectric constant of each phase is different between the media.Its principle is to use a set of sensors installed on the device under test array electrodes,formed from different angles analyte scanning space of sensitive field,according to the measured object in the different medium has different dielectric constant of the properties,and the implementation of the distribution of the field of incomplete information extraction,using the appropriate image reconstruction algorithm,each phase of analyte pipe field distribution in the form of a visual image.This paper mainly improves the quality of reconstructed images by improving the sensitive field,and jointly uses ANSYS Maxwell,C++and MATLAB for research.The main research achievements of this paper are as follows:(1)The principle of ECT is introduced briefly,and the composition of ECT system as well as the positive and negative problems in the mathematical principle are explained mathematically.A 36-electrode ring capacitance sensor is designed for a specific model.In this paper,the traditional imaging algorithms are introduced,and the typical algorithms are analyzed for image reconstruction under general manifold.(2)Since the sensitive field has the characteristic of "soft field",in the actual environment,the sensitive field will change with the change of the medium,but the sensitive field is not expected to be changed in theory.Therefore,this paper measures the capacitance value of each position by changing the placement position of the medium in the measured area.Through the relationship between capacitance value,sensitivity matrix and distribution matrix of dielectric constant,the fixed sensitive field under ideal state is obtained,and based on the sensitivity field generated by the old and new methods,the advantages and disadvantages of the reconstructed image are analyzed.(3)It is proposed that when the sensor is long,the measured area will be extreme,such as too long,too narrow,irregular and so on.It is also based on the fact that the extremely low signal-to-noise ratio of the electrode signal across both ends of the measured area will reduce the overall measurement accuracy,resulting in poor quality signals in part of the measured area,In this paper,the sensitivity field of the 36-pole sensor is optimized,and the image of the ring object is reconstructed based on the sensitivity field before and after optimization,and the error analysis is used to compare the advantages and disadvantages of the sensor.