Open AccessBad data identification method of power system based on co‐evolutionary symmetric loss function
Author(s) -
Yu Lijie,
Sun Weifu,
Yang Zhongqin
Publication year - 2021
Publication title -
international transactions on electrical energy systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.428
H-Index - 42
ISSN - 2050-7038
DOI - 10.1002/2050-7038.12589
Subject(s) - population , electric power system , computer science , evolutionary algorithm , data loss , power (physics) , statistics , control theory (sociology) , algorithm , mathematical optimization , mathematics , artificial intelligence , computer network , physics , demography , control (management) , quantum mechanics , sociology
Summary With the rapid development of power system, the structure and operation mode of power grid are more and more complex, and the identification of bad data in power system is an important basis to judge the operation status and fault of power system. Therefore, an evolutionary symmetric loss function based identification method for power system bad data is proposed. The Karl Pearson coefficient of variation is used to represent the Poisson distribution of power system data. Based on the prior distribution parameters of any symmetric loss function as posterior mean, the coefficient of variation of power system is estimated. Considering coevolutionary coevolution model of coef power system & variational coefficient method, the data based on population density is constructed, and the disturbance is adjusted ‐ the optimization term is added to the model. Based on the optimized coevolutionary algorithm and framework, combined with local evolution adaptive mutation strategy, the population evolution is completed, and the output solution is the result of identification of bad data in power system. The experimental results show that the method has high recognition rate, known bad data and unknown bad data, low packet loss rate, high false alarm rate, and alarm rate in the process of identifying bad data in power system. It shows that the method has small error and high recognition rate.