Research Of Visually Early Warning In Smart Distribution Network

With the vigorous development of China’s economy, the requirement of power quality are increasing, the safe operation of the distribution network becomes very important. Visually early warning in smart distribution network can reflect real-time dynamic distribution network to grid operation and management personnel, to prevent the deterioration of the abnormal state, protection of smart distribution network security, stable and reliable operation.According to the distribution network SVG file, this thesis parse and reconstruction of the spatial grid structure of distribution network, then uses the build a distribution network’s element early warning assessment method. This thesis introduces the association rules mining method in relevancy assessment of the distribution network, optimizes the AprioriTid algorithm according to the need of distribution network early warning, using the optimized algorithm to obtain the correlation parameters of components. Considering the great influence of disaster climate on distribution network components, this thesis combines disaster climate data and distribution network operation data, using the same method to mining association rules, then calculate the probability of failure. Synthesis of these two parameters and reliability assessment parameters, get the results of components’ early warning assessment results. On the basis of the relevancy assessment, uses Quantum immune algorithm and immune algorithm based on information entropy and association rule mining to derived early warning results. This thesis gives the immune algorithm specific processes in distribution network’s early warning, and conducted feasibility analysis of algorithm. By the end of the thesis with the actual distribution network as an example, using software to verify the early warning method proposed in this thesis. Verified by the experiment, this thesis puts forward the early warning method of early warning results generally consistent with the simulation results, the warning effect is good.The distribution network early warning method and parse and reconstruction of the spatial grid structure of distribution network method have been applied in the software of Northeastern University and Liaoning Electric Power Co., National Grid cooperation project "The key technology research to improve the distribution network fault handling capacity".