For aircraft with dorsal intake,when using the flush air data sensing system(FADS)to calculate the atmo-sphere parameters,different engine working conditions may lead to different intake flow rates,which affect the pres-sure at the pressure measurement point.In this paper,a FADS system algorithm based on radial basis function(RBF)neural network was proposed for the flat head aircraft with dorsal intake.The outlet flow rate of the intake was added as one of the calculation factors for atmosphere parameters.The pressure measurement point configuration was determined based on the flat head characteristics of the aircraft.Three RBF neural networks with single output were established for Mach number,angle of attack and angle of sideslip respectively.The results of the two groups of neural network solutions were compared whether the engine working condition was taken into account.The simulation results showed that the neural network solution algorithm without taking the engine working condition into account had good accuracy,and the solution accuracy of the neural network considering the engine working condition was fur-ther improved,the absolute error of Mach number,the angle of attack and angle of sideslip reduced to 0.001 Ma,0.15° and 0.3° respectively.
关键词
嵌入式大气数据传感系统/背部进气/径向基神经网络
Key words
flush air data sensing system(FADS)/dorsal intake/radial basis neural network
维普
万方数据