Research On The Trajectory Based Trajectory Sensitivity Analysis And Validation Of Power System Digital Simulation

Digital simulation is an important tool in analyzing power system dynamics, and the validity of simulation output is with vital importance to power system dynamic anlysis and control. Several postmortem analysis show that the simulation output based on the given models and parameters is with great difference to the measured trajectory. Though how to recognize the erroneous models and paramters based on the simulation error, and correct the erroneous models and parameters to improve the validity of digital simulation and ensure the secure and economic operating of power system is an important issue to researchers.Trajectory sensitivity is an important tool in power systems digital simulation validation, and the conventional trajectory sensitivity calculation method is based on power system models and parameters to get the computations models, and then numerical integration is used to obtain trajectory sensitivities.The computation cost of trajectory sensitivity is great, and it depends on the accuracy of models and parameters.In this paper the higher order models of power system is concerned, and a convolution method based on trajectory to calculate trajectory sensitivity in single machine power system and multi-machine power system is proposed The sensitivity of trajectory to component parameters and to fault clearing time can be obtained only by trajectory, and the numerical intergration of power system to obtain trajectory sensitivity is avoided, and the relation between trajectory sensitivity and trajectory is extracted. The influence of erroneous parameters can be avoided, and the computation cost can be reduced, and the convolution method is simple in calculating and provides a new way for analyzing power system dynamic.The difficulties of model validation of digital simulation are concerned, and a waveform-dividing method to recognize the oscillation mode is proposed, and a set of indexes to assess the accuracy of simulation is proposed based on the trajectory performance. An error source recognition method based on trajectory sensitivity analysis is presented. Case studies of a two machine and an infinite bus system in dynamic laboratory validation show that the error index and error source recognition method are valid.The idea of dividing power system into several subsystems which can validate the subsystem separately is used and a hybrid dynamic simulation based power system decoupling method is concerned, and a measured voltage based hybrid dynamic simulation method is proposed to fulfil the decoupling of power system and subsystem model validation. By the real power system model and parameter correction and New England 10 machine power system model validation the validity of the proposed method is verified。In this paper a Matlab based power system model validation software of digital simulation is developed to improve the efficiency of bulk power system model validation.The software use the dyamic analysis funciton of Power System Analysis Toolbox(PSAT), and can fulfil the power system decoupling and recognize the error source and correct the erroneous parameters automatically, and the software is a powerful tool for model validation of power system digital simulation.