Study On Plasma Image Data Processing Technology On EAST Device

High-power,high-parameter,long-pulse steady-state plasma operation are the main goals of the Tokamak device discharge experiments.Under those conditions,it is necessary to observe the discharge and measure the temperature of the target plate of the divertor which is frequently bombarded by high-energy particles in real time.It is essential to provide real-time plasma boundary information and accurate temperature of the target area of the divertor,which makes a great sense to achieve the goal of the Tokamak discharge experiments.In the EAST device,the visible/infrared endoscope diagnostic system is mainly used to observe the movement state of the plasma and to measure the surface temperature of the plasma facing components.At present,this diagnostic system has the ability of preliminary image data acquisition and temperature calculation,but it does not have the functions of plasma image boundary tracking,temperature data jitter correction and high-speed data sharing.In order to solve abovementioned problems,this paper mainly studies the following three aspects from the perspective of image data processing:Firstly,a boundary detection algorithm suitable for plasma visible images is proposed.The algorithm is based on the principle of wavelet transform,and the final image boundary is output by detecting and fusing the boundary information at different scales.The actual detection results show that the proposed algorithm is better than the traditional Canny detection operator in the boundary recognition of plasma images,which not only has good noise suppression ability but also retains more boundary details.Secondly,a jitter correction algorithm suitable for plasma infrared images is proposed.The algorithm utilizes the idea of the plasma boundary detection,by detecting the offset of the relevant structure of the divertor target area in the infrared image data,which effectively avoids the poor imaging quality of the infrared camera and the low signal-to-noise ratio of the image in not suitable for using the feature point recognition algorithm to calculate the jitter.With the interpolation calculation,the temperature data can be corrected.The calculation results show that the proposed algorithm provides a practical way for refining temperature data.Thirdly,a system for high-speed transmission and sharing of plasma image data based on reflective memory networks was constructed and implemented.During the discharge experiment of Tokamak,the visible/infrared endoscope diagnostic system will collect a large amount of image data,which needs to be shared with each terminal for processing.When using Gigabit network cable for transmission,it often encounters slow and unstable transmission speed.The high-speed image data transmission and sharing system based on the reflective memory network can provide a stable offline transmission speed of not less than 80 Mbyte/s,and a rich expansion interface,which can meet the data transmission capability requirements of the Tokamak device discharge experiments.In this paper,an effective solution is provided and successfully applied in practice to solve the problems existing in the visible/infrared endoscope diagnostic system in the EAST Tokamak device.However,the data processing algorithms proposed in this paper still need continuous optimization and the accuracy of data calculation results needs to be further improved.