Abstract: In order to guarantee the rationality of demand prediction for emergency materials, post-disaster uncertain factors and makers' subjective factors were examined fully, and a PT based model was built for dynamic adjusting emergency material demand. Firstly, the decision maker's perceived values were divided into two sorts, a valid sort of perceived value (in scheme) and an invalid sort, with the decision maker's psychological expectations as a dynamic reference point. Then, by determining the value function and the probability weight of the scenario, the comprehensive prospect value of the scheme can be calculated, and the prediction scheme of the emergency material demand corresponding to the minimum prospect value can be obtained. Finally, a sensitivity analysis was carried out, of the prospect value of the scheme to risk preference, risk aversion coefficient and the perceived probability coefficients of both benefit and loss. The effectiveness of the model was demonstrated by the data on a certain seismic disaster. The results show that the model based on the dynamic change of disaster situation is more suitable for use in a situation after disaster, which helps to find out the best prediction scheme, and can improve the decision-making quality and rescue efficiency .

Key words: emergency rescue, prospect theory(PT), emergency material demand, dynamic reference point, project adjustment