The invention discloses a low-frequency oscillation online control strategy based on mode matching, and relates to the technical field of power system stabilization and control. The low-frequency oscillation online control strategy comprises the step of applying a dynamic stability control strategy generated by small interference calculation and sensitivity analysis to low-frequency oscillation online based on mode matching. According to improved Prony, QR and IRA algorithms and a characteristic value sensitivity principle, a criterion of mode matching and an online modal calculation method are provided, advantages of the time domain improved Prony algorithm, the frequency domain QR algorithm and the IRA algorithm are combined to match a small interference calculation result with mode information obtained in real time by the on-line improved Prony algorithm, when a vibration mode calculated by small interference real-time scanning is mismatched, a power system stabilizer PSS based on aparticle swarm optimization technology is adopted for restraining low-frequency oscillation, the low-frequency oscillation recognition speed of a large power grid is improved on the whole, meanwhile,the real-time performance and the accuracy of discovery and solution are met, parameter optimization is conducted on the PSS through the particle swarm optimization technology, and the robustness ofthe system is improved.