Research On Risk Area Prediction And Emergency Method Of Power Cyber Physical System

With the proposal and continuous development of smart grid and energy internet strategies,a large number of electrical equipment,data collection equipment and computing equipment are connected by physical network and cyber network.The traditional power system with physical equipment as the core has gradually evolved into a highly coupled power physical cyber fusion system(Cyber Physical Systems,CPS).This conversion relationship not only provides data support and intelligent decision-making for the power system,but also brings more security risks to the operation of the system.Once a physical network in the CPS network fails,the risk is likely to cause the exchange part of the network to spread and spread to the entire system,causing the operation of the entire system and even causing a large-scale power outage.Therefore,how to accurately predict the risk area at the early stage of the accident and propose targeted emergency measures is of great significance to ensure the safe and stable operation of the electric power CPS system.The main research work here is as follows:Most of the existing power CPS network models are oriented to the network topology,and ignore the load heterogeneity and coupling non-uniformity of specific component nodes,which cannot truly reflect the actual operation of the power CPS system.Therefore,a non-uniform power CPS load and constraint characterization model is constructed.The model first uses the network topology parameters such as clustering coefficients and node degrees of component nodes,and combines the actual operating characteristics to propose a node load and constraint calculation method.Then,according to the "one-to-many" node dependency relationship,the improved ball bin allocation method is adopted to design a non-uniform dual-network coupling method,which accurately reflects the node load capacity and network topology characteristics,and provides accurate and effective models for risk area prediction and emergency method research support.Considering that the dynamic transmission mechanism of power CPS network security risk is different between the cyber layer and the physical layer,it is difficult to perform unified analysis and prediction.A power CPS risk area prediction method based on dependent Markov chain is proposed.First analyze the power CPS security risk propagation process by considering the directionality of the dual network state interaction impact.Then a dependent Markov chain probability framework is proposed,by defining the system's full state space and state transition probability,eonstruct a risk area prediction model.and finally a cross-adaptive gray wolf optimization algorithm improved by adaptive position adjustment strategy and cross-optimal solution strategy is subsequently developed to optimize the prediction model.Due to the high degree of node association in the risk area,once the risk breaks out,it requires extremely high cost to suppress and block the risk.Therefore,in consideration of limited resources,this paper proposes a method of emergency response in the risk area that maximizes returns.First,according to the state of the system risk area,adjust the emergency measures of node injection power and coordinate with load cut to avoid the aggregation of high-load nodes in the local area.Then comprehensively consider the effect and cost after the implementation of emergency measures,put forward economic and safety evaluation indicators,and construct an emergency model of electric power CPS risk area.Finally,the differential evolution algorithm is used to solve the emergency model,and the economy is optimized under the condition of effective risk suppression.