Research On Power Grid Information Observation Methord Based On Kalman Filter Algorithm

The developing of the power system topology and the large-scale access to distributed energy have all placed higher demands on modern power system information measurement systems.The fast,accurate,real-time and comprehensive information acquisition of the grid is a reliable guarantee for power system state estimation,operation monitoring,system control and optimization management.The grid information observations mentioned here mainly include the observation of the grid-connected point and the observation of the grid network information.Existing grid information observation methods have some limitations.Most of them rely on communication systems and cannot meet the complex and changing power grid environment.This poses certain hidden troubles for the safety and stable operation of power systems.In view of the above problems,this article has carried out the following research work on the grid information observation method:(1)There kinds of grid information observation methods which are widely used at this stage are listed: Extended Kalman filter,generalized second-order integrator frequency-locked loop and Adaptive Vectorial Filter.This paper analyzes the implementation principles of the three synchronization algorithms and makes a comparison by simulation model,which is the basis for subsequent research in this paper.There is a limitation in the scope of application of the grid information observation algorithm in the occurrence phase based on the three typical synchronization algorithms.It cannot estimate the DC offset in the voltage signal,and it is difficult to take into account the system time domain performance as well as statistical indicators.(2)To aim at the problems existing in the grid information observation algorithm in the current stage,a grid-connected observation algorithm based on Kalman filter is proposed on the basis of existing research.It can separate the DC offset,positive/negative sequence components and harmonic components in the gridconnected point voltage signal.By combining the linear Kalman filter with a frequency-locked loop structure,the proposed grid-connected observation algorithm can avoid the complex operation of the nonlinear Kalman filter,and can also take into account its noise suppression capability and improve the dynamic response at the same time.Then,according to the characteristic that the steady-state gain of the linear Kalman filter is fixed,a filtering parameter optimization method is designed based on the concept of the state observer in modern control theory,which can take into account the time-domain performance and statistical indicators.Finally,the simulation comparison with the generalized second-order integrator frequency-locked loop and the Adaptive Vectorial Filter verifies its effectiveness and superiority under the complex grid conditions,and it is verified on the DSP hardware platform.(3)To aim at the stability problems caused by the interference of the communication system to the condition monitoring and system control of the grid,a Kalman filter-based observation method for no-communication information is proposed.Taking the power network with distributed power supply as the research object,small-signal models for distributed power,network topology,and loads are established respectively.Then,the network information observation is designed based on Kalman filter algorithm.By measuring the local voltage and current information of the distributed power,the status of the power grid is observed under no-communication conditions,including other distributed power local information and system node voltages.It can eliminate the delay and packet loss caused by the communication system and improve system stability.Finally,the validity of the proposed network voltage observer is verified by a three DGs simulation model.