Research On Flush Airdata Sensing System Algorithms And Other Key Technology

Flush airdata sensing (FADS) system utilizes the surface pressure distribution in the vehicle to measure the airdata indirectly. It meets many challenges which the traditional one can not satisfy, such as stealth, hypersonic, and high angle of attack. In this paper, several important issues in the FADS system are investigated, including the layout of the flush orifices, estimation algorithm, calibration technology, system error and so on.The aerodynamic model, estimation algorithm, and accuracy of system can be affected by the layout of orifices directly, therefore, it is critical to choose an appropriate layout of orifices. Only the angle of attack and angle of sideslip measurement requirements were considered at first, get the best orifices arrangement through calculating angle of attack error. Then consider the angle of attack, sideslip angle and the Mach number measurement requirements, a cost function to be minimized through some optimization method such as gradient search and approach and genetic algorithm. Find the orifices arrangement cause the cost function value minimum as the optimization layout.This paper proposes a high resolution and real-time estimation and calibration algorithm based on the layout of FADS system used in a vehicle, which operates in a wide range of Mach number. Considering the problems of algorithms in the FADS system, this paper develops novel methods to ensure the uniqueness and convergence of the solution in high Mach number. It is shown that either estimate initial values of airdata before iteration, or use low-pass-filtering at the end of each computational cycle is able to force the initial value converge to the neighborhood of the right solution. The advantage and disadvantage of existing algorithm are discussed in this paper, and some suggestions in for practical applications are given.Study and analyze the error source of FADS system, model the error from pneumatic lag, orifice misalignment, viscous interaction, thermal transpiration, and these models will use to forecast the influences of various error sources, and some methods of eliminating or reducing the error.