Abstract: This project addresses issues such as significant fluctuations in valve inlet temperature and frequent start-stop operations of air cooler fans in valve cooling systems of VSC-HVDC projects under harsh environments. Based on actual engineering data, a refined dynamic simulation model integrating fluid transport delay and valve tower thermal effects was developed using the MATLAB/Simulink platform. Three optimized control strategies for air cooler fan grouping and start-stop logic were proposed, with emphasis on increasing the proportion of variable-frequency fans and optimizing PID tuning parameters. Simulation results ultimately identified Strategy II (grouping variable-frequency fans into four major sets with staggered arrangement) as the optimal approach. This strategy demonstrated excellent performance under low-load and low-temperature conditions, effectively suppressing frequent fan operations, significantly smoothing valve inlet temperature fluctuations without sharp spikes or drops, and markedly enhancing system thermal stability and control quality. It provides important theoretical support and practical guidance for the design of valve cooling systems in VSC-HVDC projects under complex environmental conditions.