Parameter Identification Method for Converter Controller of DFIG Based on Measurement Signal Disturbance

JIN Yuqing,JU Ping,LIU Weihang,PAN Xueping

DOI: https://doi.org/10.7500/AEPS20150313003

2016-01-01

Abstract:The converter controller model is the key part of a doubly-fed induction generator (DFIG) model.Because of the large number of controller parameters and the series connection between the PI controllers of inner and outer control loops,the traditional identification methods have some trouble with identifiability or engineering realization.A parameter identification method is proposed based on the disturbances made on the measurement signals of secondary side,which features multi-step identification and interactive iteration.The method separately makes disturbances on the measurement signals which are used as input of PI controllers in the converter control system.These disturbances can enlarge the effect of the triggered PI controller on the response of a DFIG.Therefore,parameters of the triggered PI controller can be identified by using its measurable control target as the observation.For the series connected PI controllers,the measurement signal disturbance will be made on the latter PI controller while shielding the change of the input measurement signal of the former PI controller. Thus,the decoupling identification can be carried out on the parameters of series connected PI controllers.Using the proposed method,the two parameters of each PI controller in the converter control system can be identified separately without any identifiability problem under the certain measurement signal disturbance.And the interactive iteration of identification can further improve the identification accuracy of total parameters.