| With the development trend of large-scale regional interconnection of power grids,UHV construction and investment of new energy units,while meeting the demand for electricity load,it also poses new challenges for the safety and stability control of power systems.The traditional offline control mode is difficult to meet the timeliness and accuracy of the current grid control requirements.Transient stability control is gradually shifting to the ’ real-time decision-making,real-time control mode.With the development of wide-area measurement system(WAMS),the real-time control mode based on short-term response information has become the focus of current scholars.In the existing research,the extended equal area criterion(EEAC)or the transient energy function method based on the two-group instability mode is usually used,combined with the trajectory prediction method to achieve real-time emergency control.This kind of method cannot reasonably distribute the control quantity because the operation state of the unit in the group cannot be observed after equivalence;moreover,in the existing feature trajectory prediction schemes combined with artificial intelligence methods,there is a lack of efficient use of key information,and there is still room for further research and optimization.In view of the above problems,this paper studies the emergency control strategy of power system transient stability based on individual machine energy function and real-time response information.The work of this paper can be summarized as follows:(1)The basic theory of individual machine energy function of local analysis method is introduced.From the perspective of transient energy consumption,the analysis method of individual-machine equal area criterion(IMEAC)based on the stability of critical units to judge the stability of multi-machine system is studied,and the identification method of critical units based on short-term response information under each disturbance is proposed.The focus of system stability is shifted to the critical units with serious disturbance,which is used as a means of system stability judgment in the overall emergency control system to improve the efficiency of stable state identification and provide a starting basis for control.(2)Aiming at the prediction of transient stability disturbed trajectory,based on a single long short term memory neural networks(LSTM),a feature trajectory prediction model combined with attention mechanism(AM)is proposed to predict the power characteristic curve of critical units,and highlight the characteristics that have the greatest impact on the trajectory prediction results.The ability of the model to mine the internal correlation between key information and sequence is enhanced,and the training efficiency and prediction accuracy of the model are improved.It provides data support for system stability assessment and emergency control.(3)Combined with the individual machine energy function and IMEAC method,a calculation method of critical unit cutting control quantity based on residual area is proposed.Moreover,the stability of each critical units in the system is based on the correlation of the inertia center.Considering that the cutting operation will lead to the deviation of the inertia center,a control quantity correction method is proposed.Finally,combined with the trajectory prediction model and the IMEAC method,a real-time emergency control strategy system is proposed to realize the stability of the system at the advanced instability time and take effective cutting control.In this paper,the New England 10-machine 39-bus system is used to verify the effectiveness of the proposed trajectory prediction method and emergency control strategy.Compared with the emergency control scheme proposed by the global analysis method,the advantages and accuracy of the proposed emergency control system are proved.It provides a feasible strategy for the power system security prevention and control system and consolidates the second defense line. |
