754 / 2019-04-27 15:44:39

Static Security Risk Assessment and Prediction of Power Grid under Planned Maintenance

Static risk assessment and prediction; Multi-source heterogeneous; Information processing; Deep belief neural network

Aiming at the current problems of single index, highly subjective and unable to comprehensively measure the impact on the system under the planned maintenance power grid risk assessment, a power grid static security risk assessment and prediction method based on multi-source heterogeneous information fusion is proposed in this paper. Firstly, the multi-source and multi-level factors reflecting the static security of power grid are analyzed from the aspects of equipment failure rate, danger degree and topological structure. Secondly, multi-source heterogeneous information is pre-processed, such as identification, cleaning and conversion, etc. Finally, a risk assessment and prediction model is established by using the Deep Belief Neural Network (DBNN) algorithm. Through repeated training and parameter tuning, the accuracy of the model is the best. The proposed method uses deep learning theory to fully mine the effective historical information, realizes the assessment of static security risk points of power grid, assists dispatchers and operators to adjust the operation mode of power grid in an orderly manner, and minimizes the incidence of system failure. The simulation results on New England 10-machine 39-bus system show that Simulation results on New England 10-machine 39-node system show that, the model proposed in this paper is more reliable than a single index risk assessment; compared with the traditional DBN and DNN algorithms, the accuracy of DBNN is high; the validity and feasibility of the research in this paper is fully proved.