Research of industrial data fusion based on rough set and evidence theory

With the development of information technology and the arrival of the era of big data, the form of data is more and more complex and the scale is more and more large. An improved data fusion method based on rough set and evidence theory was proposed to solve the problem of insufficient data fusion and the difficulty in giving full play to the value of data in the manufacturing process of enterprises. Firstly, the continuous attributes in rough set are discretized and the importance of attributes is measured. Secondly, attribute reduction is realized by improving particle swarm optimization algorithm, and attribute reduction flow chart is constructed. Then, the reduced data set is taken as evidence to establish a new belief function to enable multi-source data fusion. Finally, the validity and feasibility of the algorithm are verified by using the data of pelletizing process as the data source.