Superconductor Quench Detection System Based On Infrared Thermal Imaging Technology

In recent years,research on superconductors has become more and more mature,and the application of superconductors in various fields has provided a good opportunity for energy upgrading and high-tech development.With the increasing demand for superconducting devices,the quench phenomenon existing in superconductors hinders the development of superconducting devices.The traditional superconductor quench detection uses electrical methods,which are easily disturbed by electromagnetic noise during the detection procedure and cannot accurately determine whether there is a quench phenomenon in the superconductor in time.Therefore,an auxiliary detection method is needed to determine whether there is a quench phenomenon in the superconducting equipment.As a non-contact and real-time detection method,the infrared thermal imaging technology does not need to be connected to the superconducting device,which has certain advantages compared with the traditional quench detection method.In this thesis,the infrared thermal imaging technology is used to design a superconductor quench detection system with good accuracy and real-time performance.The host computer software of the quench detection system is the core of the system,so that the main research content of this thesis is the main function modules of the host computer software in the system.Firstly,according to the actual needs,a suitable infrared thermal camera is selected as the infrared detection equipment and the overall design of the superconductor quench detection system is completed.In the view of the infrared images with high noise and low contrast,corresponding image enhancement algorithms and pseudo-color processing algorithms are proposed.Secondly,a corresponding pattern recognition algorithm is designed according to the infrared images obtained in the simulation experiment.The pattern recognition algorithm extracts the eigenvalues of the preprocessed images,and then imports the eigenvalue data into the BP neural network model optimized by the genetic algorithm to determine the state of the superconductor.After optimization of the threshold and weight in the BP neural network by genetic algorithm optimizes,the error is reduced from the initial 0.1 to 0.05.Finally,according to the software requirements of the superconductor quench detection system,the host computer software of the superconductor quench detection system is designed and implemented.The software consists of an image acquisition and display module,an image processing and feature extraction module,a data processing and recognition module,an early warning module.After the simulation experiment,it is calculated that the quench detection system takes 500 ms to complete one detection,which can meet the needs of an actual scene.In addition,the data obtained through the host computer software facilitates subsequent research on the quench mechanism.