Abstract: To address the issue of privacy preserving for nodes in second-order continuous-time multi-agent systems, we propose a consensus control method based on state decomposition. First, we randomly decompose each node into two subnodes, with internode communication for exchanging information from only one of the subnodes, thereby preserving the privacy of the individual nodes. Second, we present dynamic and static consensus control protocols for second-order systems with directed communication topologies. Based on the matrix theory, we analyze the stability and convergence of the system and extend the conclusions to systems with undirected communication topologies and leader-follower configurations. The simulation results demonstrate that the proposed method achieves accurate consensus control while preserving system privacy.