A prototype GUI for the multi-channel sensor data acquisition and monitoring system of KTX

The paper presents a highly responsive GUI (Graphic User Interface) for KTX, through the use of a multi-threaded program architecture. "Keda Torus eXperiment" (KTX) RFP device is a supporting research for ITER in china. Reversed field pinch (RFP) is an important toroidal magnetic confinement device, which has been suggested as one of the attractive paths to fusion reactor. Compared with other toroidal configurations, such as tokamak and stellarator, the distinctive feature of RFP is its weak toroidal magnetic field. This weak magnetic field yields a string of potential reactor advantages, such as normal magnets, high engineering beta and high mass power density. Specifically the KTX program includes the following ingredients: the physical design and basic plasma diagnostic, the mechanical construction, the power supplies, the operating control system and the data acquisition system. The KTX reversed field pinch program will provide a significant opportunity to advance magnetic confinement fusion. The data acquisition and monitoring system of KTX deals with about 600 input channels. The maximum sampling rate of each channel can be achieved 2 MSPS/s. Existing solutions to this task are plagued with time-consuming image processing and data monitoring steps. In order to improve the efficiency of the whole process, we designed a parallel architecture for image processing. Because drawing space of a single channel is very limited, the data used in the drawing process would be less than the collected data. To improve the graphics quality and the drawing speed, a method to extract data before drawing is present in this paper. And the tasks are evenly assigned to several threads. This technique is called uniform partition. Necessary inter-thread communication ensures the coordination of the various threads, and results are place in shared memory. Pipeline is an important parallel technique. It is used in the data monitoring module. The major task is divided into a series of small tasks in the data monitoring module. In the pipeline structure, if a task is completed, then the subsequent task can start immediately. Each level of the pipeline can accept new task after finishing the current task, so the pipeline can increase efficiency. Reasonable framework makes the program run highly efficient and stable. At the same time, it can make the full and efficient use of the multi-core processor s. As the result, the GUI presented in this paper can meet the high demand for