Research of uniformity evaluation model based on entropy clustering in the microwave heating processes

摘要

This paper proposes a uniformity evaluation method based on Spectral Clustering and Maximum Information Entropy (ECUEM) for clustering the simulation results in the microwave heating system. The proposed method can effectively evaluate the dataset of the electric field E, the magnetic field H, the temperature field T, and analyze the non-uniformity phenomenon in the microwave heating processes. Compared with other clustering algorithms, the ECUEM can get better clustering results for the dataset in simulation of microwave heating. In particular, in the resonant cavity, the experimental results show that the minimum the evaluation results, the better the materials heating uniformity. In addition, when the ECUEM method is used to analyze the experiment of waveguide moving, the best position (0, 11/20*do, 3/14*ho) of waveguide can be obtained; at the same time, the uniformity or efficiency of materials microwave heating is the best. Moreover, other rules have been obtained in the microwave heating processes. Thus, the proposed method would provide a new method to guide the researchers who are working in the area of dataset clustering in the microwave heating. (C) 2015 Published by Elsevier B.V.