Implementation of spectral clustering algorithm for automatic identification of plasma coherence patterns

The number of data accumulated by controllable nuclear fusion devices is too large, and a large have not been fully exploited. In such big processing machine learning can play an important role. Therefore, in this work the spectral clustering method used to realize automatic data, which easily quickly find pattern information contained data. discovery these patterns great significance improving plasma confinement understanding physics. addition, applied electron cyclotron emission imaging (ECEI), one-dimensional diagnostic system emissiometer, magnetic probe, soft X-ray, fast radiation (fast bolometer) other different systems on EAST tokamak device. sawtooth identified, migration verified, problems poor supervised need rely labeled are solved. Finally, work, ECEI probe discover possible new mode time domain frequency respectively, provides idea for exploring modes.