Working conditions classification of ball mill pulverizing system based on quantum cat swarm optimization clustering algorithm

Ball mill pulverizing system (BMPS) is an important equipment in the thermal power plant and working conditions classification is the premise of realizing the control optimization. This paper proposes working conditions classification of ball mill pulverizing system based on quantum cat swarm optimization clustering algorithm. Firstly, this algorithm creates N cats, representing N kinds of data classifications respectively, to initialize the cat swarm. The position of each cat is coded by the clustering centers. And then the cats are randomly divided into the searching mode and the tracing mode to find their optimal positions. For the searching mode, the simulated annealing method is used to help the cats resist the lure of local optimal position. For the tracing mode, the quantum particle swarm optimization method is applied to enhance the global searching ability of the cats. Moreover, the sum of squared error is used as the fitness function. The proposed method is performed on the artificial dataset, UCI real dataset and the field working conditions dataset obtained from BMPS. Experiments verify that the proposed method is much more suitable for the data set of BMPS and performs much better.