Abstract: In view of problems of low accuracy of online monitoring method of mine cable using single parameter, and that monitoring method using multi-parameter does not carry out weight distribution and information fusion based on expert experience, a mine cable state maintenance method based on information fusion was proposed. Distributed sensors are used to real-timely monitor surface temperature, grounding current of cable sheath and partial discharge during cable operation. Preliminary fusion diagnosis for multi-source information is carried out according to fuzzy theory. The fusion results are used as identification framework and evidence body of D-S evidence theory.The weight of each operating parameter is determined based on expert experience, and reasonable reliability function allocation value of the evidence theory is determined combined with the preliminary diagnosis results, then the working state of the tested cable is judged according to the maximum reliability criterion. The example analysis result proves that the method can effectively integrate multi-source information of mine cable and improve accuracy of fault diagnosis.