Parametric Aerodynamic Modeling For Airborne Missile Separation And Trajectory Optimization

In the modern war, fighters usually complete a variety of tactical tasks by carrying airborne weapons. As an important part of the research on compatibility of airborne weapon, design and research of the separation process to ensure the safety of the separation process, has always been the focus of world attention. Present research methods mainly include the nested CFD numerical calculation to trajectory simulation or wind tunnel test, which is of high cost and long time consuming. A parameterized model of pneumatic can visually express the form of all the factors to the influence of aerodynamic force, which is convenient for engineers to understand and design the separation process. A t the same time, parameterized model is advantageous to the research of the sensitivity of the aerodynamic force in separation, simulation of a large number of trajectory, separation trajectory optimization design and controller design of separation process. However, in the airborne missile separation process, the aircraft aerodynamic interference effect on the missile is complex, so the missile’s aerodynamic model in separation is no longer a simple linear model as in uniform flow field. Therefore, how to establish a precise and compact parametric aerodynamic model is the key and difficult point for the study. In this paper, it is mainly discussed that the method of parametric modeling of airborne missile aerodynamics, and how to build the optimal control problem of separation trajectory optimization based on the parametric aerodynamic model, solved using the Gauss pseudospectral method. The main contents are as follows:1. Based on the modeling thought of superposition, the total aerodynamic force in separation can be regarded as "uniform flow model + the disturbance model". The sampled data is constructed as regression model using the multivariate orthogonal function least squares method, and the concrete expression of the uniform flow model and disturbance model is given. Based on statistical test of regression model, the strategy to determine the disturbance model structure is proposed. Considering boundary condition and the variation rule of disturbance model, variable coefficient polynomial model is put forward, and the parameters in this model can be estimated through the hybrid optimization algorithm.2. The core algorithm of parametric modeling is introduced in detail, multivariate orthogonal function least squares method, and this method is compared with the general polynomial regression method through the actual numerical example to illustrate its advantages, such as quickly determining model structure and effectively avoiding matrix singularity in parameter estimation. Also, the statistical testing method of multiple linear regression model is given.3. Full factorial method is adopted to design the test sample points, and parametric aerodynamic models are established based on the CFD database. The validity of the modeling method is verified through the comparison wit h modeling data and CTS wind tunnel data. The separation trajectory in the longitudinal plane without control is calculated using parametric aerodynamic model, and the separation trajectory optimization problem in longitudinal plane is constructed while selecting appropriate objective function and constraint conditions, which can be solved through the direct shooting method and Gauss pseudospectral method to get optimal open-loop control law and the optimal separation trajectory.