Abstract
A modified Newton-Raphson minimization technique for determining aerodynamic coefficients and stability derivatives of spin-stabilized projectiles with a six-degree-of-freedom nonlinear dynamical model was developed. The dynamical model for the projectiles is constructed having process noise in the system, and the instrumentation noise of the system outputs is simulated by a data model statistically similar to the measured data. The state equations of the dynamical system are continuous types while the measurement data are discrete. A continuous-discrete estimation model for the motion of the projectiles is constructed in this paper. The state variables of the system were estimated by the extended Kalman filter, and the system parameters were identified by the modified Newton-Raphson technique based on the maximum likelihood criterion. Research results show that parts of the parameters can be identified under proper noise intensity. However, the accuracy of identification is strongly influenced by both process and measurement noise, Moreover, parameter sensitivity to the system behavior is crucial for the success of identification. Two typical aerodynamic characteristics of projectiles, 105 and 20mm, are imposed to investigate the applicability of state estimation and parameter identification. It is found that the drag coefficient of zero angle-of-attack and the rolling moment derivative and identified with effective accuracy in a wide range of noise levels. On the other hand, other parameters are more difficult to identify, but the causes of deficiency for particular parameters in identification are discussed.
