Abstract: With more and more park integrated energy systems participating in cooperative games in the future, it is of great significance to allocate the cooperative benefits of multi-park integrated energy systems reasonably and efficiently. Therefore, this article proposes a collaborative optimization operation method for multi park comprehensive energy systems based on cooperative game theory and matrix semi-tensor product. A cooperative game model is developed for multi-park integrated energy systems with different park integrated energy system as participant. Considering the feasibility analysis of cooperation, this paper uses matrix semi-tensor product to improve the Shapley value method and proposes an improved benefit allocation method based on matrix semi-tensor product. Finally, this paper carries out the simulation experiment of the integrated energy system of 2—6 parks. The results show that the solution rate of the proposed method is 3.912%, 12.3967%, 24.9259%, 27.9451% and 35.7561 % higher than that of the traditional Shapley value method. It can be verified that the method proposed in this paper can effectively improve the efficiency of profit allocation in cooperative alliances and demonstrates promotional advantages in the future scenario where multiple parks continuously join cooperative alliances.