Abstract: In the low-inertia power system with a high proportion of renewable energy sources, the frequency of the system is characterized by serious fundamental frequency excursions and rapid frequency changes, which leads to the increase of the synchronous benchmark signal detection errors in the grid-connected tests and the support performance evaluations of the new energy stations, threatening the safe and stable operation of the system. A synchronous reference signal detection algorithm based on the dynamic double DFT spectral line interpolation is proposed. By analyzing the frequency dynamic response of the low inertia power system after the power disturbance, the static and dynamic characteristics of the fundamental frequency are revealed, and on this basis, the static and dynamic power grid signal analysis model is established. Then, the error of the traditional DFT harmonic detection algorithm is quantified in the fundamental frequency offset scenario, and the failure mechanism of the spectral line interpolation DFT algorithm in the dynamic fundamental frequency scenario is illustrated. The HHT algorithm is further used to identify the dynamic fundamental frequency, and combined with the double DFT spectral line interpolation method the amplitudes of the fundamental frequency and harmonic signals are corrected so as to improve the detection accuracy. In order to improve the detection efficiency, an algorithm selection criterion is also added. The test results show that the detection algorithm proposed in this paper is able to accurately measure the reference quantity in the strong dynamic grid signals. It improves the evaluation accuracy of the grid connection performance, such as the synchronous support of the new energy stations and the power quality. Moreover it provides a new alternative scheme for the synchronous reference signal detection in the power systems with a high proportion of renewable energy.