Abstract: Empirical data were extracted from three corpses in one historical war,and the time series models were set up with Box-Jenkins method.we have also quantified several Factors that may have important influence in casualty forecasting,such as adversary weapon system parity,terrain,weather,etc.Especially we choose several social behavior factors by expert consultation,determined the weights of these factors by means of group AHP(analytic hierarchy process),and then constructed a quantitative indexes system of social behavior factors on battle casualties.After that,we set up the relative computer algorithm about these factors to adjust the computer simulation outcome based on the empirical data.

Key words: army corps, mountain offensive operation, casualty forecasting, model