Aiming at the problems of lagging state identification and the lack of a response mechanism of buried oil and gas pipelines in mining areas under disturbed conditions, a monitoring system with continuous identification and strategy linkage is constructed. The system is divided into four layers of state sensing, data processing, strategy execution, and coordinated control. Additionally, it adopts multi-parameter node deployment and sliding window modeling to extract state features and construct an integrated intelligent monitoring system with hierarchical structure deployment, sensing node coordination, multi-parameter fusion discrimination, and anomaly intervention feedback, supporting non-stationary process identification and closed-loop control. The project is deployed to cover faults and buried depth mutation segments, with a system identification error of 0.3‰ to 0.9‰, a fault segment response delay of 583.99 ms, and a strategy calling success rate of 92.96%. In five months of continuous operation, the manual maintenance frequency is decreased by 57.14%, the average working time is decreased to 2.1 h, and the inspection intervention ratio is decreased by 32.8%. The results show that the system can adapt to disturbance evolution and dynamic scheduling, meeting the identification and control needs in the non-homogeneous loading environment of mining areas, with an improved ability to self-evolve path control.