This paper proposes a hierarchical co-construction mechanism and a digital resource sharing management framework, and realizes the integration and dynamic scheduling of educational resources in various regions with the help of cloud computing platform. Aiming at the uncertainty and multi-objective conflict problem of resource allocation, Fuzzy Interval Linear Programming (FILP) and Chance Constrained Programming (CCP) are introduced to deal with parameter ambiguity and constraint probability relaxation. Playing the key role of dynamic programming (DP), the problem of educational resource allocation is decomposed into a recursive optimization process through multi-stage decision-making models and state transfer equations, so as to realize intelligent allocation that maximizes the overall benefits. Relying on the cloud computing platform to count the status quo of educational resource allocation in higher education institutions, the model in this paper is utilized to practice dynamic resource allocation. The results show that in the dynamic allocation of educational resources in 2 categories of teachers and funding, the satisfaction degree of vocational higher education institutions is 81.82% and 81.67% respectively, and the satisfaction degree of general higher education institutions is 83.33% and 86.00% respectively. The accuracy of this paper’s model is greater than or equal to 90% in 6 times of the resource allocation test, and the running time is no more than 30s.
