The real-time digital measurement technique is one of the most important issues for the smart power substation. To enhance the ability of substations to measure electric power and to address the challenges of substation metering data not being able to be centrally computed and the difficulty of handling massive multi-time frequency data computation at the power substation. A new digital power measurement method is proposed based on the Hanning window function to reduce the spectral leakage. The signal is converted from the time domain to the frequency domain using a fully-phase fast Fourier transform, and harmonic parameters are obtained through polynomial approximation-based single-spectrum interpolation. The calculation accuracy is enhanced by using a variable forgetting factor Recursive Least Squares algorithm, with coefficients updated through an adaptive linear combiner, significantly improving the speed and accuracy of the measurement for power substation.