Research On Power Grid Risk Transmission And Prevention And Control Based On Complex Network

Nowadays,there have been several major blackouts caused by the collapse of the power system in the world,which have brought huge economic losses to the society and may also cause casualties.The nature of blackout accidents is caused by the transmission of grid risks.Therefore,the research on the transmission of grid risks and prevention and control is a topic that has attracted much attention.From the perspective of complex network theory,this paper studies the risk propagation process of power grids,and on this basis,proposes risk prevention and control methods for transmission and distribution networks.The specific research work is as follows:The SIR model is used to describe the process of power grid risk propagation.The power grid is a complex system,which can be constructed into a complex network of nodes and edges.Analogous to the infectious disease model,the operating state of the grid nodes is divided into three states: S(risk attack state),I(risk attack state),and R(risk removal state).By setting the risk source and transmission rate and the risk removal rate to realize the simulation of the grid risk propagation process.Using IEEE39-bus system and IEEE118-bus system as transmission network examples,and IEEE33-bus system as distribution network examples for simulation,it is concluded that most of the nodes in the system will be affected without any prevention and control measures.Taking into account the general existence of nodes in the transmission grid and the uneven distribution of clustering coefficients,it is used as an indicator of the importance of network nodes,and the prevention and control of risks in the transmission grid can be achieved by immunizing important nodes.This paper adopts a global immune strategy based on two rounds of selection mechanism that comprehensively considers the two indicators of degree and clustering coefficient.This strategy gives priority to immunizing nodes with high degree and low clustering coefficient.Through the simulation of the risk propagation of the IEEE39-node system and the IEEE118-node system under the target immunity and the global immunity strategy based on the tworound selection mechanism respectively,it is concluded that the latter has a better immunity effect.Since the distribution network generally operates in a radial pattern,the global immune strategy based on the two-round selection mechanism with the clustering coefficient as the reference index is not suitable for the distribution network.In this paper,the node contraction method is used to evaluate the node importance of the distribution network.This method takes the degree and location of nodes as factors to measure the importance of nodes,and then imposes an immune effect on important nodes.Taking the IEEE33 node power distribution system as an example for simulation,comparing the density of risky attacked nodes before and after immunization,it proves that the adopted immune strategy has a significant effect on the risk prevention and control of the distribution network.Considering that the distribution network generally adopts closed-loop design and open-loop operation,the method of network reconstruction can be used to reduce the risk of distribution network operation.Summarizing the causes of blackout accidents,it can be seen that the chain reaction caused by line overload and tripping is the main cause of blackout accidents.Therefore,this paper uses the load balance rate as the objective function to reconstruct the network in order to achieve the goal of balancing the line load and reduce operational risks as much as possible.Realization of distribution network reconfiguration needs to be based on power flow calculations and an optimization algorithm is used to achieve it.Since the power flow calculation program is frequently called in the network reconstruction,the forward and backward generation method based on node layering is adopted.This method only needs to calculate the layered matrix (60) and the upper node matrix two auxiliary matrices,which greatly improves the speed of calculation.It is pointed out that when the particle is in the optimal position in the standard BPSO algorithm,its velocity is 0but the probability of position change reaches 0.5.Aiming at this shortcoming,the mapping function is improved to enhance the ability of the algorithm to find global optimization in the initial stage and local optimization in the later stage.In addition,considering that a large number of invalid solutions will be generated after reconstruction(resulting in the existence of rings and islands in the distribution network),an algorithm that only needs to process the adjacency matrix to judge the network connectivity is adopted.Finally,with the minimum load balance rate as the objective function,the IEEE33-node distribution system is reconstructed to make the distribution network load more uniform,and at the same time,the node voltage quality and overall efficiency are improved.