Abstract: The railway passenger service and production management and control platform (referred to as the management and control platform) is an important platform for railway passenger stations (referred to as passenger stations) to provide passenger services and production operations. In recent years, with the rapid development of management and control platforms, the endogenous security risks caused by the increase in internal attack surface and unknown external attacks have been constantly increasing. To solve the network security problems in the development process of management and control platforms, this paper modeled and analyzed based on game theory, and proposed a Dynamic Heterogeneous Redundancy (DHR) architecture for APT (Advanced Persistent Threat) attacks adapted to management and control platforms, as well as a scheduling strategy adapted to this architecture. The paper introduced the State Cycle Weight Scheduling (SCWS) algorithm based on redundant body scheduling, and conducted simulation experiments on the performance of SCWS algorithm, compared the scheduling dynamics and ability to resist APT attacks of this algorithm with related algorithms under limited resource conditions. The experiment shows that the SCWS algorithm has significantly better scheduling cycles and resistance to APT attacks than other algorithms under this constraint.