G.G. Dutta, A. Singhal, and A.K. Ghosh (India)
Modelling, Simultion, Parameter Estimation, Neural Networks and Real Flight Data Nomenclature a0, a1, a2 = constants of series approximation CD = drag coefficient d = diameter , , dx dy dz dt dt dt = time derivatives of trajectory coordinates, X, Y, Z g = acceleration due to gravity M = Mach number m = mass of projectile s = reference area of the shell m = mass of projectile s = reference area of the shell Wx = wind velocity components in x, direction X, Y, Z = trajectory co-or
The proposed method to estimate drag coefficient from flight data of artillery rocket draws its inspiration from the research work done in the area of aerodynamic modeling and estimation of aerodynamic coefficients using Feed Forward Neural Networks (FFNNs). The universal mapping capability of FFNNs has been used to build flight dynamic model to predict acceleration for a given set of input variables of an artillery rocket in motion. The drag coefficient CD was predicted by suitable interpolation and manipulation of the predicted acceleration for a given set of motion and atmospheric variables. The proposed method is first validated using simulated flight data of an artillery rocket. The final validation has been carried out using five sets of real flight data of a typical artillery rocket in motion. Further, the procedure using Maximum Likelihood (ML) method has also been applied on real flight data of these artillery rockets to estimate the values of drag coefficient at different Mach numbers. The estimated values of CD obtained through the proposed method and the procedure based on ML method have been compared to evaluate the suitability of the proposed method. The results show that the proposed method can advantageously be applied on typical flight data of an artillery rocket in motion to estimate the drag coefficient CD.
Important Links:
Go Back
