Research On Key Technology Of Flush Air Data System Algorithm And Air Data Sensing Information Fusion

Air data system is an important airborne avionics equipment which directly affects the flight control system,further impair the flight quality and combat effectiveness.Traditional air data system adopts intrusive measuring method,thus it is hard to adapt to high pressure and temperature introduced by advanced air vehicles.Under high Mach,large attack angle,trans-atmosphere and other severe conditions,traditional air data system cannot work effectively.Directing at this problem,the air data measuring problem for advanced air vehicles is taken as the research background in this thesis.Aiming to solve the bottleneck problems in air data measuring and fault-tolerant technology,and improve the measuring accuracy,range and reliability of air data system,air data solving algorithms for flush air data system are studied,further more,key technologies of air data sensing based on multi-source information fusing are researched.On the basis of pressure measurements from multiple ports and pressure distribution model,FADS achieve air data through solving method.Although the analytical solving method can guarantee system stability,it brings a strong constraint to the layout of the pressure ports of FADS.Directing at this problem,a new analytical method of FADS with weak constraint on pressure ports distribution is proposed in this thesis.The method can directly obtain analytical solutions of attack angle,sideslip angle,dynamic pressure,static pressure and total pressure by using pressure measurements from four non-coplanar pressure ports.Compared with the existing method,this method is no longer dependent on pressure ports on the vertical line,which greatly reduces the geometric constraints to the pressure ports configuration from the air data solving method,improves the FADS configuration flexibility,makes full use of the multiple pressure ports redundancy,improves the reliability of FADS air data solving,and the analytical calculation of all air data parameters reduces the amount of computation.FADS fault tolerent algorithms generally work after the air data solving.With air data solutions from different combinations of pressure ports,fault detection,isolation and reconstruction are realized through voting and other methods.To fully exploit the redundancy of FADS and enhance FADS's independent fault-tolerant performance,FADS solving method with multiple pressure measurements,fault detection algorithm and independent fault-tolerant algorithm.In view of the FADS analytical calculation method are sensitive to large noise and makes it difficult to detect the system faults,FADS least square method based on the model element reduction is proposed.Thus,the air data calculation problem with multiple pressure measurements is solved,the accracy of air data is improved,and the sensitivity to measuring noise is reduced.Then,Chi-square fault detection method is derived for FADS.In addition,to solve the large amount of calculation brought by the large number of pressure ports combinations,the random sampling consensus algorithm is adopted,FADS independent fault tolerance algorithm with random pressure port combination is proposed.As a result,the fault detection,isolation and reconstruction ability of FADS independent falut tolerant method is improved.When the advanced air vehicle is in state such as high maneuver,the influence of the harsh flow field on air data sensor cannot be eliminated.To further improve the reliability of ADS,the full parameter calculation of air data without relying on the air data sensor needs to be realized.Hence,the method of air data calculation based on multi-source information fusion is studied.Considering the stable and horizontal wind in stratosphere,the air data estimation method fusing navigation and flight control system is established with the kinetic model of the vehicle as bond.At the filter aspect,two filtering models are proposed separately: 1)model considering the air data as state vector and the general navigation parameters as measurements;2)model considering the air data and navigation parameters as state vector and other navigation data as measurements.The indirect all-round measurement of air data without extra hardware system and measuring equipment is realized.The wind field in troposphere and middle layer does not satisfy the stationarity.To provide air data estimation during different flight stages for advanced air vehicle and improve the reliability,the air data estimation problem under changing wind field needs to be solved and extend the application scope of air data estimation method.For air data estimating with known wind model,a new method through two-stage wind estimation is proposed.The better initial wind filtering value is generated through the coarse estimation,then the air data filtering structure is derived with the wind model consists of average wind and turbulence.This method can provide effective air data estimation under the wind field variating in this certain pattern.Furthurmore,air data estimating method with wind variation model unknown is researched.Consider the wind speed as unknown deflection,initialized three-step extended Kalman filter algorithm is derived to realize the air data estimation.This method solves the air data estimating problem with the wind variation model unknown,thus the applicable range of air data estimation method fusing multi-source information is expanded.To verify the above-mentioned air data system algorithms,research platform is designed at last in this thesis.In this platform,the FADS analytical solving method,the FADS independent fault tolerant method,air data estimation method fusing navigation and flight control system,and air data estimation method under variating wind field are realized and analyzed.The effectiveness and benefits of these algorithms are validated,which has important role for support of the mentioned research work in this thesis.