Abstract: The security situation of survivable system is becoming more and more serious. The premise of enhancing survivability is to recognize the survival situation of the target system. A survival situation awareness model is constructed based on survival cluster recognition and prediction. Firstly, the improved Ward clustering method is used to realize the classification and recognition of different service levels. Secondly, the autoregressive integrated moving average (ARIMA) model is used to predict the future trend of the target system's survival situation and the residuals of the prediction results are corrected. Finally, the survival situation of survivable system is realized by the combination of pre-recognition and post-prediction. Simulation results show that the proposed model has better recognition effect and higher prediction accuracy.