Bidirectional Risk Analysis and Warning Technology for Power Grid-Communication Network

The increase of power system capacity leads to the existence of certain risks and hidden dangers in power grid construction. This paper proposes a bidirectional risk analysis and warning technology for power grid-communication network. First of all, by describing the link risks of power grid-communication network as a complex system composed of input, state, and output attributes, and through the functional relationship among attribute elements, the correlation among events is analyzed from the attribute level. Secondly, the Bayesian network is used as a modeling tool to identify the causal and coupling relationship among concurrent risk events, and construct a concurrent risk events chain model. The prior probability information among network nodes is obtained based on history data of the power grid-communication network, and the evolution analysis of emergency events is implemented by using Bayesian network inference algorithm to realize risk detection in advance and active early warning. Finally, the effectiveness of the proposed method is verified by simulation.