基于参数辨识原理的飞机参数分析及算法研究

提出一种在飞机概念设计中基于参数辨识理论的设计参数分析方法。在给定的设计重量和任务剖面要求下，利用基于物理的运动方程和动力学方程模型辨识出飞机气动力参数，并在总体性能评估的基础上进行设计参数分析，给气动设计提供了设计参考；在辨识过程中针对参数可行域离散度很高使得很难收敛到Pareto解的问题，提出了将神经网络预测模型融合到遗传操作过程，使得尽量在可行域内搜索。最后通过对某客机进行算例分析，表明基于辨识理论的参数分析方法和改进的算法是合理和可行的。与一般经验公式方法相比，该方法对布局类型的限制较小，在满足概念设计参数分析准确度要求的条件下能够为下一步的气动设计提供设计指标。

Aircraft Sizing Based on the Parameter Identification Theory and Algorithm Improving

From the parameter identification theory, an aircraft sizing method is put forward in the aircraft conceptual design phase. When the desired weight and mission section had been known, the aerodynamic parameters can be identified based on the physical model of motion and kinetic equations. According to the principle of conceptual evaluation, the aircraft optimum parameters can be chosen including those aerodynamic parameters that could be as a design guide in aerodynamic. As parameter feasible region is very discrete, which result in algorithm is difficult to converge at Pareto points in the identification process, an improved algorithm which merges the neural net forecast model into genetic manipulation process has been proposed to search in the feasible regions more effectively. Through the analysis of a civil transport airplane, the result shows the aircraft sizing method based on parameter identification theory is reasonable and feasible. Compared with those experience formula methods, the provided method in this paper has fewer limits for layout type and could express the aircraft sizing process more exactly, and could also offer quantity index for aerodynamic design further more.