Abstract: Aiming at the problems that the comprehensive performance evaluation of combustion detection instruments of thermal power units was difficult to achieve due to the complex structure, numerous indexes and difficult to quantify, a comprehensive performance evaluation method was proposed based on information entropy-grey fuzzy fusion model, and an information entropy-grey fuzzy evaluation model was established. The proposed method was compared and verified by 3 detection methods of the engineering examples of carbon content in fly ash of generator sets. The fuzzy mathematics was used to quantify the 3 types of language scales, such as instrument performance, applicable performance and risk performance, and the model inputs were established by normalizing 11 indexes from the optimal membership degree. The coupling correlation degree between the inputs was comprehensively considered by the grey correlation coefficient matrix, and the entropy weight method was used to replace the traditional expert method, eliminating the subjective influences of the weighting processes. The comparison results show that the secondary index weights of 3 schemes account for 14.63%, 58.51% and 26.86% of the total respectively under the eliminating human interference environments. The membership degrees of the optimal result 1 are as 0.4265, 0.6642 and 0.9673 respectively. Compared with the single factor analysis method, the model has consistency in the evaluation results, and shows better comprehensive evaluation efficiency in multi data analysis and eliminating human interference.

Key words: grey fuzzy, entropy weight method, comprehensive performance evaluation, carbon content in fly ash, information entropy

摘要： 针对火力发电机组因结构复杂、指标众多且难以量化导致的燃烧检测仪器综合性能评价难以实现的问题，提出了一种基于信息熵-灰色模糊融合模型的火电机组燃烧检测仪器综合性能评价方法，建立了信息熵-灰色模糊评价模型。以发电机组飞灰含碳量工程范例的3种检测方式对比验证所提方法，采用模糊数学量化仪器性能、适用性能、风险性能3类语言标度，以从优隶属度规范化11项指标建立模型输入，用灰色关联系数矩阵综合考量输入量之间的耦合关联度，以熵权法替代传统专家法，剔除赋权过程的主观影响。对比结果表明，3种方案的二级指标权重在去人为干扰环境下，分别占总体的14.63％、58.51％、26.86％，对最优结果1的隶属度分别达到0.4265、0.6642和0.9673。与单因素分析方法相比，该模型在评价结果上具有一致性，而在多数据分析和去人为干扰中表现出更好的综合评价效能。

关键词: 灰色模糊, 熵权法, 综合性能评价, 飞灰含碳量, 信息熵