Abstract: Aimed at the defects of incomplete evaluation indicators and unreasonable evaluation results in the evaluation of existing power infrastructure projects, this paper proposes an improved TOPSIS-based power infrastructure project quality deviation evaluation model. By analyzing the factors that affect the rationality of the quality evaluation of power infrastructure projects, an evaluation index system for project quality deviations is constructed. The combination weighting method is adopted to effectively balance the subjective and objective weight values and complete the reasonable selection of weight combination coefficients. Using cosine similarity and gray correlation analysis, the improvement of the traditional TOPSIS method is completed, based on which, the improved TOPSIS method is applied to the quality evaluation of power infrastructure projects, so that the accuracy of the evaluation results is further improved. The results of calculation examples show that the model can effectively improve the project quality deviation evaluation indicators, make the evaluation results more in line with the actual situation, and provide effective support and decision-making for the evaluation of power infrastructure projects.

Key words: quality deviation, index system, power infrastructure project, grey relational analysis, cosine similarity

摘要： 针对现有电力基建项目评审时存在的测评指标不全面、评价结果不合理等缺陷，提出基于改进TOPSIS的电力基建项目质量偏差度评价模型。通过对影响电力基建项目质量评审合理性的因素进行分析，实现项目质量偏差度评价指标体系的构建。采用组合赋权法，使得主客观权重值得到有效的平衡，完成权重组合系数的合理选取。利用余弦相似度和灰色关联分析，完成对传统TOPSIS法的改进。在此基础上，将改进TOPSIS法运用于电力基建项目质量评价工作中，使评审结果的准确性得到进一步的提高。算例结果表明，该模型可有效完善项目质量偏差度评价指标，使得评价结果更加符合实际情况，为电力基建项目的评审工作提供有效支持和决策。

关键词: 质量偏差度, 指标体系, 电力基建项目, 灰色关联分析, 余弦相似度