A Model Order Reduction Method for Nonlinear Multi-time Scale Systems

Based on singular perturbation theory and quadratic approximation method of the stability region boundary of nonlinear system,a stability law was derived for nonlinear multi-time scale system,whose order was reduced by neglecting fast dynamics.Then,the quantitative prediction expression of stability margin index of reduced system was presented using the high-order Taylor series expansion of stability margin index of the original system.Subsequently,a model order reduction method for nonlinear multi-scale systems was proposed,which can ensure stability consistency between original and reduced systems.Finally,according to the proposed reduction method,the model order of an AC/DC power system was reduced by neglecting fast dynamics.Time-domain simulation results demonstrate the proposed model reduction method is effective and applicable to provide simplified model for dynamic characteristic analysis of the AC/DC power system.