Wind farms must promptly respond to grid dispatch instructions. Delays in monitoring data transmission at the millisecond level can cause wind turbines to disconnect from the grid. Furthermore, given the diverse data types, varying reliability needs, and unpredictable packet loss or bandwidth fluctuations, traditional congestion control mechanisms are unable to meet these stringent requirements. Therefore, a communication congestion control algorithm for wind power dispatch operation monitoring network is proposed. The congestion level value is calculated, and based on the data integrity requirements of different application scenarios, the weights and thresholds are dynamically adjusted to ascertain the congestion status. During congestion, the algorithm coordinates four stages: slow start, congestion avoidance, fast retransmission, and fast recovery, to ensure stable and efficient network communication. Experimental results demonstrate that the proposed method significantly reduces monitoring data transmission delays and wind turbine disconnection rates across various congestion levels. It also enhances the average business utility of the wind power scheduling and operation monitoring network, offering notable advantages in response speed, bandwidth stability, and bandwidth supply. Consequently, it ensures stable and efficient network communication while improving the utilization rate of communication links.