Abstract: In order to effectively quantitatively analyze and evaluate the wear risk grade of filling pipeline，the attribute recognition theory is introduced，and a comprehensive evaluation model of filling pipeline wear risk based on the attribute recognition of improved CRITIC method is established. Firstly，12 representative influencing factors including weighted average particle size of filling aggregate were selected from four aspects of slurry characteristics，pipeline characteristics，construction technology and filling technology， and a multi-level comprehensive evaluation index system was established. Secondly，the method of improvement CRITIC is adopted to determine the objective weight of the indicator. According to the attribute characteristics of the indicator and the grading criteria，the measure function is established，and the single indicator attribute measure and comprehensive attribute measure values are calculated respectively. Under the principle of attribute recognition，the risk level is comprehensively evaluated. Finally，taking Jinchuan Longshou Mine，Baoshan Lead-zinc Mine and Xincheng Gold Mine as examples，relevant calculation was carried out to judge the pipeline wear risk status of each mine. By comparing with the evaluation results of unascertained measure theory and variable fuzzy set theory，the accuracy of the established attribute recognition model based on Improvement CRITIC method in the analysis and evaluation of wear risk was verified. The results show that the wear risk levels of filling pipeline in three mines are respectively level Ⅱ，level Ⅲ and level Ⅲ. The results are consistent with the engineering practice，and the evaluation results of the three comprehensive evaluation models are the same，which provides a new method for the wear risk assessment of filling pipeline.

Key words: backfill pipeline, attribute recognition, improved CRITIC method, wear risk