N4SID based subspace identification method for normal temperature continuous transonic wind tunnel system

In this paper, considering the features of normal temperature continuous transonic wind tunnel which are nonlinear, strong coupling, fast time-varying and uncertain, meanwhile, combination the advantages of subspace identification method which not only could accomplish the system identification merely need the input and output data of the identified system, but also directly obtain the state space description, the form is favor to do the deep analysis for system. Now, we propose the idea that utilize N4SID algorithm to establish math modeling for the wind tunnel steady state. Through simulation experiments show that the identified systems by N4SID and PEM algorithms have satisfied approximation for the real wind tunnel system steady state, about 90%, even the confidence level of the identified result reach to 0.95. However, the accuracy of identification model from N4SID is better than PEM's.