Abstract: In view of the problem that the supporting capacity of the hydraulic support cannot be effectively exerted due to the poor working conditions of the hydraulic support and it is easy to induce roof disasters, in order to achieve effective monitoring and early warning of the support working conditions, the main influencing factors of the working condition of hydraulic support wereanalyzed,andseveral key indicators reflecting the working conditions of the hydraulic support were obtained through the analysis of the working resistance of the hydraulic support, and six main indicators for the comprehensive evaluation and real-time early warning of the support working conditions were determined,namely: unqualified ratio of setting force of hydraulic support, high reporting ratio of working resistance, opening ratio of safety valve, ratio of support not maintaining pressure, unbalanced ratio, unqualified working resistance ratio.According to the fuzzy mathematics membership comprehensive evaluation method, the multi-parameter comprehensive evaluation and early warning model of support working condition was established.The support working condition was divided into four grades according to good, medium, poor and very poor,andthe membership degree interval of hydraulic support working condition grade and the membership degree function corresponding to each early warning index parameter wasgiven.Based on the variable weight theory, the variable weight coefficient of each early warning index was determined,and the impact of the change of the state quantity of the early warning index on the overall working condition of the support was highlighted. Finally, the comprehensive evaluation and early warning level of the working condition of the support were obtained, and the early warning information of the working condition of the support was displayed in the form of a radar chart in the early warning interface of the system. The early warning model has been developed and applied in KJ21 roof disaster monitoring and early warning system, which verified the rationality of the support working condition evaluation system and effectively improved the support quality of hydraulic support in fully mechanized mining face.