Addressing the scenario in which each robot can only acquire the azimuth information of adjacent robots in dynamic formations with incomplete constraints, this paper proposes a distributed PID formation control algorithm based solely on azimuth information. With considering that the pilot robot is susceptible to disturbances such as wind direction or road surface irregularities, which may disrupt formation maintenance, the algorithm introduces relative position and velocity feedback of the follower robots. This approach effectively eliminates steady-state error, suppresses the influence of disturbances, improves system dynamic performance, and ensures global system stability. Then, the Routh-Hurwitz stability criterion is used for stability analysis, verifying the global stability of the formation system. Finally, simulation experiments compare the performance of the proposed control law with control laws based on pure proportional and proportional-integral strategies in terms of convergence speed and disturbance rejection. The results show that the proposed control law enables the formation to recover after disturbances and achieve rapid trajectory tracking of the leader, with the relative maximum deviation of the total azimuth error reduced by 5.4%.