There are problems of high energy consumption and low efficiency of power data center task scheduling in the existing cloud environment. Thus, based on power cloud architecture, this paper proposes a cloud data center task scheduling based on the stochastic Petri net model, and considers the time, load, and energy consumption constraints to improve the ant colony algorithm to solve the problem presented by the model. The advantages of this method are verified by comparing and analyzing the running time, energy consumption, average waiting time and system load. The results show that the improved ant colony algorithm can effectively reduce the energy consumption of the data center and guarantee performance. It provides a reference for the development of a task scheduling strategy in a power data center. This work is supported by the Science and Technology Project of State Grid Corporation of China (No. 520234190006).