This study focuses on optimizing the formation topology in multi-robot systems using the leader–follower method. The primary objective is to enhance system observability and navigation accuracy. First, the chapter introduces the importance of observability in leader–follower systems, particularly analyzing how particle filter algorithms can control the positioning error of the leader robot. The Fisher Information Matrix (FIM) and Cramér-Rao Lower Bound (CRLB) are employed to evaluate the formation topology for single and multiple follower robots. By constructing an evaluation function based on these metrics, the optimal formation topology is identified to minimize localization errors and improve overall system performance. Simulation experiments are conducted to verify the effectiveness of the proposed optimization strategies, demonstrating significant improvements in obstacle avoidance and navigation precision. This chapter provides a solid foundation for improving the efficiency and reliability of multi-robot formation control in complex environments.