Research On Relay Protection Value Setting And Management Based On Risk Decision

With the continuous development of distributed energy, as the main direction of future power development, smart grid topology will become increasingly complex.The high requirements of relay protection in terms of scope, diagnostic accuracy and accuracy make the existing fault diagnosis methods face severe challenges.Comprehensive analyzing the actual operation mode, the external environment and other risk factors, setting value for the distribution network, logically setting the value to meet the protection requirements of the relay protection is the key to protecting the safe of power system.In the paper,a fault diagnosis method with high correctness and fault tolerance is proposed, and the influence of external risk on relay protection is analyzed synthetically to set and manage the protection settings. The main work of this paper is as follows:The paper introduces the current situation of relay protection custom management, power grid fault diagnosis and optimal management of relay protection and risk assessment, and elaborates the setting process of relay protection setting in smart grid.Aiming at the problem of data surplus as well as lacking of information and the difficulty of obtaining representative fault diagnosis model of current fault information system, researching on the method of feature extraction and fault diagnosis based on random forest, removing the most of redundant eigenvalues from great capacity of data. The number of optimal decision trees and the most suitable feature quantities are obtained by using the stochastic forest algorithm, so as to reduce the dimension of the input vector space and shorten the calculation time. Combining with Knowledge of automation, extracts the rules, and then the formation of knowledge, to build online revision of the cycle decision model.The stability of the power system involves many uncertain factors, analyzes the safety index system of the fixed value, and studies the probability risk of Bayesian reasoning. Considering the system operation mode, the external environment risk, the fault type line failure rate, the probability distribution to protect the value of decision-making management, for different expert advice and system analysis to provide a variety of information sources, based on multi-information fusion collaborative decision-making mechanism, a comprehensive analysis of the distribution of information and importance, reduce the amount of data makes the decision more comprehensive, and improve the comprehensive management of risk management.Combined with the actual operation of the grid, the IEEE 14 node system as a research case, to obtain the best economic optimization program; research to protect the value of the service test system, test the value of the revised group, through the experiment to verify the value of the set of service experimental system online modify the availability of the settings, and correct the design of the protection settings.