Abstract: This paper addresses the issue of the Proportional-Integral-Derivative（PID） controller in high-speed trains being unable to adaptively adjust parameters during operation, proposing a distributed cooperative composite controller based on the improved adaptive genetic algorithm（IAGA）-PID algorithm.First, the high-speed train is modeled as a multi-agent system, considering information coupling to construct a realistic operational model.Second, an IAGA-PID controller is designed using inter-train speed and displacement data, innovatively applying the artificial potential field method to adjust the required inter-train distance based on speed, while a sliding mode observer is designed to compensate for uncertain disturbances during operation.Finally, simulations using actual high-speed railway parameters demonstrate, through comparisons of different PID tuning methods and with/without the observer, that the proposed method significantly outperforms others in speed tracking control, enabling all trains to quickly and accurately follow the target profile, maintain safe tracking distances, and exhibit strong adaptability to complex operational environments.