| With the development of China's power system reform and energy transformation,Renewable energy such as wind power and electric vehicles are connected to the power grid.The impact of the external environment on the power grid is increasing.Modern power systems has developed into a complex system with strong uncertainties.The comprehensive use of big data,domain knowledge and artificial intelligence technology to develop artificial intelligence systems for the power industry is of great significance for maintaining the safe,economic and stable operation of the power system.However,considering the existing technical conditions,power supply reliability and equipment safety,it is unrealistic to test the artificial intelligence system directly in the actual power system.Most experiments are inseparable from the simulation system.Therefore,it is of great significance to study the simulation design of a complex power system with strong uncertainty and to simulate the actual operating state of the power system as much as possible.Based on the existing power system simulation model,simulation technology and simulation algorithm,this paper proposes a software integration idea and framework structure,and integrates the existing results to realize the construction of electric digital digital system for power industry AI research and development.On the basis of this,look forward to the future development direction of this technology and the research mainly focuses on:(1)The multi-agent technology is used to realize the integration of various heterogeneous models of the power system.A flexible and easy-to-expand simulation platform can be established with using multi-agent technology,which can comprehensively consider the influence of various factors of the power system.The simulation platform can be used to simulate the dynamic characteristics of the power system under the combined action of uncertainty factors,wide-area control,and relay protection.(2)The uncertainty model of power system is studied.The stochastic process,stochastic differential algebraic equation and probability function are used to establish uncertainty models such as load,wind power output,transmission line probability fault and communication delay.In addition,in order to simulate the influence of power system relay protection and control on power system transient and medium-and long-term dynamics,a wide-area PSS control,AGC control,and logic-based relay protection model are established.The validity of the models built are verified in small cases.(3)The correctness of the communication interaction of the multi-agent simulation platform Agent built in this paper is verified on the IEEE 14-node test system.By comparing with the Monte Carlo(MC)method,it is verified that the strong uncertainty simulation system designed in this paper can consider the uncertain factors nature of stochastic volatility more than MC method.The simulation results are more comprehensive than the MC method that only considers the initial state uncertainty.The preliminary empirical research on the 36-node test system shows that the strong uncertainty simulation system designed in this paper can comprehensively reflect the influence of each agent on the power system,and can realize various data sources as well as “big data” and “situational awareness” scenarios.Simulation provides comprehensive simulation data for the study of new intelligent dispatch decision support systems. |
PDF Full Text Request