Aiming at the problem of long average delay of traditional UAV emergency communication, a performance optimization method of UAV communication link based on emergency communication environment is proposed. The target tasks in the emergency communication channel are divided into different target task groups. According to the distance between the target tasks in the target task group, the symmetric matrix is constructed, all elements in the symmetric matrix are classified, and the classification results are calculated by ant colony algorithm to construct the emergency communication task scheduling model of UAV. Combined with the emergency communication task scheduling model and the demand, priority, time window, and communication ability of UAV emergency scheduling task, the time delay control of UAV emergency communication is realized with the goal of maximizing the efficiency of UAV task execution, maximizing the completion rate of UAV task execution and minimizing the risk of UAV delay. The experimental results show that the average delay of the three methods also increases. When the system load is 50%, the average delay of common methods is 23 ms, the average delay of traditional methods is 76 ms, the average delay of the method proposed in this paper is 26 ms, and the common average delay is the shortest. When the system load is 99%, the common methods cannot realize emergency communication scheduling due to system load overshoot, and the corresponding average delay cannot be considered, indicating that the common methods are unstable. The average delay of the method proposed in this paper is stable under different system loads and always remains below 40 ms. It is proved that compared with the traditional methods, the method proposed in this paper takes less time for UAV emergency communication and has less average delay under different system loads, so it has better application value.