| With the continuous development of the power system,the interconnection degree of power grids is gradually increasing.At the same time,the complexity of the power system network environment continues to raise due to the large number of new energy access,applications of measurement devices and other factors.The traditional weighted least square method state estimation(WLS-SE)as a classical method in the actual power system will face greater challenges,and improving its convergence performance and estimated quality in complex network environments will become quite important.The step size factor of the traditional WLS-SE state correction expression is usually fixed to 1,and the weight remains unchanged in the iterative process,which cannot resist the adverse impact of bad data in the measurement set on the estimated value of state variables,and cannot obtain good iterative performance and estimated result when the numerical conditions of the power system network tend to deteriorate.In order to promote the convergence performance and improve estimated quality of power system WLS-SE in the above case,this paper has done the following studies on the step size factor and the measurement weight:Firstly,the traditional power system WLS-SE and the detection and identification of bad data are introduced,which provides a method basis for the following research work.Based on the ideal iterative principle,the logic function widely used in neural network is selected,and the smart step size adjustment function is constructed by a series of steps to the logic function such as transformation and variable substitution.By introducing algorithm control parameters into the smart step size adjustment function makes the shape of function more controllable.At the same time,the graph of the algorithm function is analyzed to clarify the role of each parameter.The iterative expressions of parameters p and α are proposed,and parameters p and α can be adjusted automatically with the change of state correction during iteration,so the step size factor can be adjusted smartly according to the state correction and control parameters.In order to ensure the adaptability of the proposed low model dependent state estimation method on smart step size adjustment in different node systems,a specific adjustment strategy of the state estimation algorithm control parameters is proposed.The excellent performance of the proposed smart step size adjustment algorithm is verified by the simulation under the conditions of bad data and ill-conditioned.Secondly,the deficiencies of common equivalent weight function are pointed out.This paper introduces the IGG(Institute of Geodesy and Geophysics,Chinese Academy of Sciences)robust method and its geometric significance,modifies it according to the shortcomings of the IGG robust method to put forward the dynamic weight adjustment algorithm,which can not only make full use of the measurement information in the measurement set to ensure the observability of the system,but also adjust the measurement weight dynamically through the measurement residuals in the iterative process,so as to improve the estimated quality of WLS-SE in the complex power grid environment.The effectiveness of the proposed dynamic weight adjustment algorithm is verified by testing in a real grid.Finally,a state estimation algorithm based on double loop feedback control is proposed by combining the smart step size adjustment algorithm and the dynamic weight adjustment algorithm.Based on IEEE30 node system,simulation analysis is carried out under the conditions of bad data measurement and quasi ill-conditioned and ill-conditioned.By comparing the convergence performance,estimated quality and stability performance of each algorithm,the better performance of the proposed algorithm is verified.Based on IEEE14,30,118,and 2383 node systems,the adaptability of the proposed algorithm in different node systems is verified under the conditions of bad data measurement and quasi ill-conditioned and ill-conditioned. |
PDF Full Text Request