The recently proposed ambient signal-based load modeling approach offers an important and effective idea to study the time-varying and distributed characteristics of power loads. Meanwhile, it also brings new problems. Since the load model parameters of power loads can be obtained in real-time for each load bus, the numerous identified parameters make parameter application difficult. In order to obtain the parameters suitable for off-line applications, load model parameter selection (LMPS) is first introduced in this paper. Meanwhile, the convolution neural network (CNN) is adopted to achieve the selection purpose from the perspective of short-term voltage stability. To begin with, the field phasor measurement unit (PMU) data from China Southern Power Grid are obtained for load model parameter identification, and the identification results of different substations during different times indicate the necessity of LMPS. Meanwhile, the simulation case of Guangdong Power Grid shows the process of LMPS, and the results from the CNN-based LMPS confirm its effectiveness.

The identification of load model parameters from practical measurement data has become an essential method to build load models for power system simulation, analysis and control. In practical situations, the accuracy of the load model parameters identification results is impacted by data quality and measurement accuracy, which leads to the problem of identifiability. In this paper, an identifiability analysis methodology of load model parameters, by estimating the confidential intervals (CIs) of the parameters, is proposed. The load model structure and the combined optimization and regression method to identify the parameters are first introduced. Then, the definition and analysis method of identifiability are discussed. The CIs of the parameters are estimated through the profile likelihood method, based on which a practical identifiability index (PⅡ) is defined to quantitatively evaluate identifiability. Finally, the effectiveness of the proposed analysis approach is validated by the case study results in a practical provincial power grid. The results show that the impact of various disturbance magnitudes, measurement errors and data length can all be reflected by the proposed PⅡ. Furthermore, the proposed PⅡ can provide guidance in data length selection in practical load model identification situations.