Abstract: In response to the concealment requirements for road transportation under the threat of multi-satellite surveillance, an intelligent route planning method integrated with road risk value estimation is proposed. At the risk assessment level, a road risk assessment model based on satellite scanning patterns is constructed. By simulating the orbital operation and scanning time sequence of multiple satellites, dynamic risk values are assigned to the road network, accurately depicting the evolutionary law of risks in the time dimension. At the route planning stage, an innovative Q-Learning-based selective node waiting strategy is designed, which breaks through the limitation of traditional methods that only rely on spatial avoidance. Through the time scheduling of key nodes, dynamic risk avoidance is achieved. Relying on the dynamic simulation of satellite monitoring risks, this method realizes transportation route planning that balances real-time performance and concealment through the joint optimization of Dijkstra and Q-Learning algorithms. Experimental results show that the proposed method can effectively reduce the satellite monitoring risk during road transportation and significantly improve the transportation survival capability in complex environments.