Extracting the dynamic feature information of the pressure pulsation signal in the draft tube of the pump turbine (PT) and accurately identifying the strengths of vortex rope are the research focuses in the engineering field of the PT in recent years. Based on ensemble empirical mode decomposition (EEMD) and index energy of mode, the features of the pressure pulsation signals in the draft tube of a PT at different load conditions in the generating mode are extracted in the present paper, and the following conclusions are obtained. Firstly, the index energies of modes based on EEMD can effectively reflect the energy distribution in the signal. Secondly, in the process of the increment of the strength of vortex rope, the maximum index energy of mode based on EEMD increases continuously, indicating that the flow status in the draft tube become more complex and the feature information of vortex rope is more abundant. Finally, the eigenvector constructed using the maximum and mean index energies can accurately reflect different strengths of vortex ropes in the draft tube. And it can be adopted as the input eigenvector of the intelligent classifier, which is conducive to further recognition and diagnosis, and has important engineering significance.