尖锥前体飞行器FADS系统的人工神经网络建模及风洞试验研究

针对一种用于尖锥前体飞行器的嵌入式大气数据传感（Flush Air Data Sensing，FADS）系统的解算模型及精度进行研究。针对尖锥外形特征，首先基于钝头体FADS系统的理论模型确定其测压孔配置；然后对确定的测压孔进行典型状态的风洞试验测试，并比对了数值计算数据与风洞试验数据；最后基于人工神经网络建模技术构建了FADS系统的网络解算模型及算法。结果表明:针对尖锥外形测压孔配置特征，基于人工神经网络建模技术的算法解算精度较好，迎角、侧滑角、静压、马赫数的网络输出值与试验值吻合较好，输出的测试误差（绝对值）分别小于0.1°、0.1°、50.0 Pa及0.01；同时也证实了人工神经网络算法在FADS系统中有进一步发展的空间。

Study on artificial neural network modeling and wind tunnel test for the FADS system applied to the vehicle with sharp nosed fore-bodies

The solving model and its accuracy for the Flush Air Data Sensing (FADS) system applied to the vehicle with sharp nosed fore-bodies are studied. Firstly, the pressure ports configuration for sharp nosed fore-bodies is determined based on the theoretical model of the FADS system applied to the blunt fore-bodies. Secondly, a typical wind tunnel test for the FADS pressure ports is implemented, and CFD results and wind tunnel test are analyzed systematically. Finally, the solving model and algorithm are developed for the FADS system based on the Radial Basis Function (RBF) neural network modeling, and the neural network model solving accuracy for the FADS system is analyzed. The results show that, the solving accuracy of the RBF neural network model for the FADS system is high enough. The neural network outputs agree well with the wind tunnel test data, and the testing error distributions (absolute error distribution) for the angle of attack, angle of sideslip, free stream static pressure and Mach number are less than 0.1°, 0.1°, 50.0 Pa and 0.01, respectively. The results also present that the artificial neural network modeling for the FADS system can be further developed in the future.