Research On Intelligent Aided Diagnosis For Power System Fault Based On Information From SCADA System

With the expansion of the power system scale,the investment of intelligent systems and the access of new energy generation,the complexity and uncertainty of the power grid structure continue to increase.Maintaining the reliability of power system operation is the common goal of current research in the field of electric power.Once the power system fails and the protective action circuit breaker trips,there will be a huge amount of message information pouring into the regulation system,so the effective filtering of the massive realtime fault data in the monitoring system and the use of intelligent fault diagnosis and analysis methods to quickly find the fault source are of great research significance for improving the efficiency of fault post-processing,reducing the power outage time,and improving the reliability of power system operation.Aiming at the problems of incorrect action of protection circuit breaker,incorrect transmission of message information,information redundancy under multiple complex faults and difficulty in quickly locating faulty components during the operation of the power grid,this thesis proposes a multi-population genetic algorithm power grid fault diagnosis model based on based on information from SCADA system.Establishing a information detection and processing model to screen the initial information data.According to the correlation between the protection action information and the related circuit breaker action information,the validity and reliability of the information will be determined,and then used for the diagnosis model.To reduce the computational effort of the fault diagnosis model,Establish a rapid identification model for fault areas.By using the effective data after processing,the related components are identified by comparing with the system topological structure,and the fault area composed of the set of related components is used as the solution object of the fault diagnosis model.According to the current action logic of relay protection,the traditional fault diagnosis model is modified.When constructing the objective function,the influence of the reclosing action on the component fault hypothesis is regarded as a constraint condition,which can better improve the practicability of the fault diagnosis model.In order to improve the shortcomings of premature convergence and slow convergence speed of traditional genetic algorithm,an improved multi-population genetic algorithm was proposed,and use examples to verify the optimization ability of the improved algorithm.The fault diagnosis model proposed in this thesis is constructed based on the MATLAB platform,and the IEEE 9 node system is used for simulation analysis.Simulate Three different fault states respectively,such as simple faults,protection devices and circuit breakers refusal to operate and multiple secondary faults.Using classical genetic algorithm and multi-population genetic algorithm to solve each case,then by comparing the optimization simulation results under different algorithms,the effectiveness of the proposed method in power grid fault diagnosis is verified.