Research On Dispatching Operation Risk Analysis And Intelligent Check Technology

Dispatching and control operation is an important responsibility to achieve safe and economic operation of the power grid.With the continuous development of the power system and the increasing penetration of new energy,the operation mode of the power grid will change significantly in a short period of time.In order to ensure the stable operation of the power grid,regulators need to adjust the status through dispatching and control operation.The traditional regulation system only provides basic operation and monitoring functions,but there is a general lack of logic model between programs,which requires experience knowledge judgment by operators.In addition,there is a lack of standardized operation and automated verification of safety measures,increasing the difficulty of operational decision making and complexity of event handling by regulators and reducing the efficiency.Therefore,there is an urgent need to introduce artificial intelligence technology to achieve risk analysis and rapid verification of control strategies and operational processes to improve their disposal efficiency and the speed of safety verification.In this paper,based on typical scenarios,operational processes and refinement rules for dispatching operation risks,a three-dimensional knowledge expression method is used to construct a knowledge base to achieve safety checks for dispatching operations.The main research contents are summarized as follows:(1)The variables to be selected are normalized for different values and units of equipment parameters under grid operation.In this paper,the input features are selected using the improved DBSCAN(Density-Based Spatial Clustering of Applications with Noise)algorithm to find the typical feature variables representing the grid operation mode.Based on the situation that regulators are more concerned about devices that change faster and have a greater impact on the grid,a method for measuring the degree of variation between scenes is proposed based on different device weights.The permutation entropy of different equipment time series is calculated to obtain the degree of regularity of different equipment time series changes,and higher weights are assigned to equipment with lower regularity.According to the above difference measurement method,the typical scenes of operation mode are generated by clustering with improved K-means algorithm,which lays the foundation for the subsequent mining of refinement rules for safety check of dispatching and control operation.(2)The risk sources of dispatching and control operation process are analyzed,and the operation state matrix is generated according to the possible states in the operation process.The probability and risk consequences of each state are analyzed.In response to the problem that the evaluation objectives of the original state before dispatching operation,the operation completion state and the operation process state are different,a unified risk assessment index system for power grid regulation and control operation state is established.The subjective and objective assignment methods are used in this evaluation system and the weights are combined based on deviation minimization.Then the risk theory is used to evaluate the dispatching and control operation process and extract the risk points of dispatching operation process.(3)Risk assessment is achieved by conducting dispatching operation simulations based on different typical scenarios and historical data,thus uncovering key variables that influence the risk indicators of dispatching operations.Since the purpose of this feature selection is not to refine the relationship between different variables but to find out the intrinsic connection between the feature variables and the riskiness of dispatching operation,the improved DBSCAN algorithm is not applicable,so the XG-Boost(e Xtreme Gradient Boosting)algorithm is used to find out the key features.After discretizing the data for the feature variables and dispatching operation risk indicators,the FP-growth(Frequent Pattern Growth)algorithm is used to dig into the rules between dispatching operation and safety check,which are difficult to discover by experience.A knowledge base of safety check is constructed for dispatching operations based on rules and knowledge matching is used for intelligent security check.