Mach Number Prediction for 0.6 m and 2.4 m Continuous Transonic Wind Tunnels

With the development of design technology, more and advanced diverse wind tunnels have been constructed to match complex requirements. However, it is hard a precise physical model tunnel that can be controlled. In addition, if new designed, experimental data may insufficient build controlling model. This article reports research on following two models: (1) for 0.6 m continuous transonic supported by large amount historical data, false nearest neighbor (FNN) algorithm was adopted calculate order input variables, nonlinear auto-regressive with exogenous inputs–backpropagation network (NARX-BP) proposed its Mach number prediction model; (2) 2.4 only small method migration, output slope/bias correction–particle swarm optimization (IOSBC-PSO) algorithm, developed convert old into tunnel, so could conducted. Through simulation experiments, found introducing NARX-BP model, root-mean-square error (RMSE) decreased 44.93–77.90%, maximum deviation (MD) 64.05–85.32% compared BP The performance IOSBC-PSO migration also better than non-migration as evidenced 82.06% decrease RMSE value 78.25% MD value. experiments showed effectiveness strategy.