Abstract: It’s particularly important to recognize timely the electrical connection looseness fault of coal-mine bolted cable joints. Lots of looseness fault experiments under different loosening state,current and load conditions were car- ried out with self-developed experimental platform. The temperature characteristics,contact voltage and current charac- teristics of the loosening bolted cable joint under different conditions were studied. A new looseness fault identification method based on current energy entropy and Probabilistic Neural Network (PNN) was proposed. The multi-resolution analysis of current signal was conducted by using wavelet transform,and the current energy entropy used as typical fea- ture parameter of electrical connection looseness fault was extracted. Then the current energy entropy was put into the PNN fault identification model. The PNN fault identification model was established by using newpnn function with Mat- lab software and the spread parameter S was optimized by using Loop optimization method. The relationships between the identification accuracy and both the number of training samples and high frequency electromagnetic noise were also discussed. Lots of testing results show that the suggested method can identify the electrical connection looseness fault of coal-mine bolted cable joint effectively.