A bran-new performance evaluation model of coal mill based on GA-IFCM-IDHGF method
Wentao Xu,Yaji Huang,Siheng Song,Gehan Cao,Mengzhu Yu,Haoqiang Cheng,Zhicheng Zhu,Sheng Wang,Ligang Xu,Qiubai Li
DOI: https://doi.org/10.1016/j.measurement.2022.110954
IF: 5.6
2022-01-01
Measurement
Abstract:Adopting a more scientific and precise assessment method to evaluate the current running performance of coal mills in practice work is essential to maintain the security and reliability of the coal-fired power plant. However, traditional assessment methods have ignored an important problem that the importance among variables varies with the current operation data of coal mill, which has a serious influence on performance evaluation of coal mill. In this paper, a bran-new GA-IFCM-IDHGF assessment method is proposed. Genetic algorithm (GA) is first applied to optimize initial parameters, which is fundamental and significant step to obtain more accurate the number of clusters. Secondly, intuitionistic fuzzy clustering model (IFCM) is adopted to identify the running modes of the coal mill. According to the Delphi, analytic hierarchy process, gray relations analysis and fuzzy integrated evaluation (DHGF), the security range of variables under the normal observed samples from different running modes is acquired, which is helpful to determine whether the current operation data is normal data. Subsequently, an improved Delphi, analytic hierarchy process, gray relations analysis and fuzzy integrated evaluation (IDHGF) method is used to assess the current running performance of coal mill under the abnormal monitoring data. The performance of the proposed performance assessment method is verified through its application in a self-defined step fault system and an actual industrial cases of coal mill. Compared with the traditional methods, the experimental results demonstrate the effectiveness of the proposed method.