To enhance the adaptive capabilities of PID controllers for service-oriented logistics robots under diverse operating conditions, this paper proposes an online PID parameter tuning strategy based on the Adaptive Particle Swarm Optimization (APSO) algorithm. This strategy is applied to a double closed-loop feedforward control system with environmental perception feedback, referred to as the APSO-PID system. First, an embedded controller performs online parameter identification, updating the robot’s velocity and position transfer functions in real time, providing an accurate dynamic model for the controller. Second, the APSO algorithm enhances the standard Particle Swarm Optimization by incorporating adaptive learning factors and a variable inertia weight strategy. This enables real-time online optimization of PID controller parameters. Finally, comparative experiments against the traditional Cohen-Coon and Ziegler-Nichols methods demonstrate that the proposed APSO-PID control strategy offers significant advantages in terms of dynamic response speed and trajectory tracking accuracy.