In the face of major public health emergencies, the epidemic prevention resources of individual cities are always limited. However, it is effective for improving the efficiency of emergency rescue to utilize different city infection peaks and cooperatively dispatch different epidemic prevention personnel and materials within the urban cluster. Based on this, this paper constructs a cross regional collaborative scheduling optimization model for epidemic emergency resources from the perspective of infection staggered peaks, covering emergency hospital site selection, cross regional transportation of emergency materials, and cross regional support of medical staff. This mixed integer programming model fully considers multiple heterogeneity and multi decision coupling effects in resource scheduling processes. For the convenience of model solving, this article proposes two new logical inequalities to add cutting based on the characteristics of the staggered collaborative scheduling problem. The results indicate that in the face of major public health emergencies, the satisfaction rate of patients" needs is the primary consideration factor for decision-makers. In the process of implementing graded diagnosis and treatment, a relaxed matching strategy for mild symptoms should be chosen instead of the optimal perfect matching strategy. At the same time, decision-makers also note that important parameters such as the amount of emergency supplies, the number of medical staff, and the capacity of emergency hospitals all have significant threshold effects. These research conclusions can provide specific decision-making support for cross regional collaborative scheduling of emergency resources for major public health emergencies.