This paper presents a collaborative control architecture designed for wheeled pipeline robots. Initially, the establishment of models for pipeline robots is conducted, including kinematic models and differential steering models. The paper analyzes the motion patterns within pipelines and cooperative control strategies for pipeline robots. Subsequently, research is conducted on motion control algorithms, presenting a cooperative control system for pipeline robots. This includes the development of a local path planning algorithm based on Reinforcement Learning and a hierarchical controller suitable for speed and steering control of the driving wheels of pipeline robots using Fuzzy-PID algorithms. Finally, simulation analyses are performed on the straight-line and turning motions of the pipeline robot.