Temperature Sequential Data Fusion Algorithm Based on Cluster Hierarchical Sensor Networks

The process of extracting gold by biological oxidation involves oxidizing the refractory high-sulfur and high-arsenic ore with the help of bacteria to decompose the wrapping material of gold to extract the gold. Therefore, maximizing the activity of bacteria will directly affect the efficiency of gold extraction, for which it is particularly important to maintain the pulp temperature in the oxidation tank at the optimal bacteria breeding temperature. However, gold mines are generally located in mountainous areas, and the large temperature difference between day and night in winter, coupled with the influence of wind and snow, creates variations in the temperature in the oxidation tank. The traditional temperature measurement method cannot fully reflect the temperature change of the oxidation tank. As a multi-field application method, sensor information fusion can effectively address the problem of pulp temperature measurement. First, we analyzed the heat transfer principle inside the oxidation tank, and designed the cluster hierarchical sensor network according to the spatial position of each oxidation tank and the environmental interference factors. The network structure is divided into three layers; the bottom of the sensor to collect pulp temperature data shows a spiral distribution in the inner wall of the oxidation tank. Each cluster head node sensor is used as an intermediate layer to complete local measurement fusion estimation. Finally, the fusion center is taken as the upper layer to realize the global state fusion estimation. Secondly, in the data processing of the bottom temperature sensor, the traditional unscented Kalman filter (UKF) algorithm is improved and the fading memory matrix is added to improve the identification of nonlinear modeling errors. The sequential observation fusion estimator (SOFE) algorithm is embedded in the measurement update to improve the performance of local measurement fusion. Finally, in the global state fusion estimation, the sequential analysis is combined with the inverse covariance intersection, and the sequential analysis and inverse covariance intersection-global state fusion estimation (SICI-GSFE) algorithm is proposed. Through calculation and simulation, the results show that the external interference can be reduced by combining all the temperature state estimations, and the accuracy of the best global temperature state estimation is improved.

Advanced Helical Plasma Research towards a Steady-State Fusion Reactor by Deuterium Experiments in Large Helical Device

The Large Helical Device (LHD) is one of the world’s largest superconducting helical system fusion-experiment devices. Since the start of experiments in 1998, it has expanded its parameter regime. It has also demonstrated world-leading steady-state operation. Based on this progress, the LHD has moved on to the advanced research phase, that is, deuterium experiment, which started in March 2017. During the first deuterium experiment campaign, an ion temperature of 10 keV was achieved. This was a milestone in helical systems research: demonstrating one of the conditions for fusion. All of this progress and increased understanding have provided the basis for designing an LHD-type steady-state helical fusion reactor. Moreover, LHD plasmas have been utilized not only for fusion research, but also for diagnostics development and applications in wide-ranging plasma research. A few examples of such contributions of LHD plasmas (spectroscopic study and the development of a new type of interferometer) are introduced in this paper.