Abstract: A new compensating method is presented to eliminate nonlinear dynamic measurement error of virtual instrument system. In this method, according to the static and dynamic calibrated data of virtual instrument system, the parameters of static and dynamic compensation link models are identified on the basis of the functional link artificial neural networks (FLANN) that is constructed by the least square-support vector machine (LS-SVM), then both the compensation link models are connected in series to the original virtual instrument system so as to correct its nonlinear characteristics and improve its dynamic performance, thus obtaining the desired input/output characteristics of the system. Test results revealed that this method is effective and superior to compensating nonlinear dynamic error for virtual instrument system.