Research on spectral analysis method of load characteristics in smart grid

With promoting of demand response mechanism, injection of large-scale new energy into the distribution network in smart grid and rapid change of economic situation as well as industrial structure, time-varying of electric loads intensifies. Conventional indexes of power load characteristics such as load factor and peak-valley difference factor have difficulties in estimating electric load peak and valley characters comprehensively and effectively. This paper puts forward spectral analysis method to analyze power load characteristics and pulls out corresponding indexes to describe peak and valley fluctuation characters of electric load curves, which transfers research on load characteristics from time domain to frequency domain and makes up the deficiency in study of electric load characteristics in time domain. This method divides the time series of power load into superposition of periodic components with different amplitude, phase and frequency and finds out main-cycles components contained in power load. Through analyzing fluctuation characteristics of main-cycles components, we can obtain the essence of load curves´ fluctuation more effectively. Finally, the paper establishes frequency domain analysis indexes system of load characteristics to support interactive construction of users and power networks in smart grid. The example has confirmed the effectiveness of the indexes system.