A kind of airborne distributed POS Transfer Alignments based on parameter identification
Technical field
The present invention relates to a kind of airborne distributed POS Transfer Alignments based on parameter identification, can be used for various needs
The strapdown inertial navigation system of position, speed and attitude is determined by transfer alignment technique, is particularly suitable for body and be there is bullet
Property deformation situation.
Background technology
Multitask remote sensing load is one of current airborne earth observation developing direction, such as integrated high-resolution mapping camera,
Imaging spectrometer, big visual field infrared scanner, synthetic aperture radar (Synthetic Aperture Radar, SAR) are in same
The multitask load of carrier aircraft, airborne distributive array antenna SAR and flexible multi-baseline interference SAR etc..For equipment task remote sensing
The high performance turbine remote sensing system of load, needs to realize high-acruracy survey to the kinematic parameter of each load distributed points.
Position and attitude measuring system (Position and Orientation System, POS) is at present airborne to see over the ground
The Main Means that remote sensing load obtains the kinematic parameters such as position, speed, attitude are surveyed, which is mainly constituted including IMU, global positioning system
System, navigational computer and the poster processing soft.It is comprehensively airborne for the high-performance for being equipped with multiple or various observation load to see over the ground
Examining system, as multiple or various observation load are arranged on the diverse location of aircraft, aircraft elastic deformation causes the sky between load
Between relativeness change.Now, traditional single POS obviously cannot meet the high-precision motion of different settlement multi-loads
The demand of parameter measurement.Therefore, it is necessary to it is height to set up the distributed space-time datum system (airborne distributed POS system) of high accuracy
In the airborne earth observation systems of performance synthesis, all load provide high-precision time, spatial information, its typical application load
For distributive array antenna SAR.Distributed SAR antenna is distributed in wing both sides, needs to the position at each antenna, speed, appearance
State and main and sub system baseline carry out high-acruracy survey.
Airborne distributed POS system is typically by the main POS of high accuracy and many sub- Inertial Measurement Unit (Inertial
Measurment Unit, IMU) composition.Main POS is typically mounted in carrier aircraft cabin;The general distributing installation of sub- IMU is in body
Diverse location (includes wing), carries out Transfer Alignment to which by kinematic parameters such as the high precision position of main POS, speed, attitudes
To realize the accurate measurement of movable information at place.As airframe has elastic deformation, except certainty between master subsystem
Lever arm error and fix error angle beyond, also there is time dependent random lever arm error and elastic deformation angle.Wherein with
The random lever arm error of time change and elastic deformation angle, do not simply fail to accurate measurement, and can be in main and sub system Transfer Alignment
It is middle to introduce complicated random error, so that sub- IMU movement parameter measurements precise decreasing.Accordingly, it would be desirable in Transfer Alignment process
In the elastic deformation of aircraft is estimated and is compensated.
Elastic deformation in practical flight has two kinds, and a kind of is due to the deflection deformation produced by fighter maneuver, another kind
It is vibration source or fitful wind etc. make that carrier aircraft is produced inside carrier aircraft vibration deformation, both deformations typically can be regarded as mistake at random
Journey.The existing method compensated to elastic deformation angle mainly has three kinds:First method is to simulate body bullet using software
Property motion model, for example the patent of Publication No. 201210113395 using ANSYS aid in modeling method simulation wing elasticity
Deform, but the model set up by the method changes with the change of aircraft material, and plug-in load is not accounted for aircraft bullet
Property deformation impact.Second method is to adopt to increase Kalman filter process noise matrix to reduce body deformation and vibration
Impact.The method be according to true model analysis of covariance result come determine injection white noise intensity, i.e., by increase
Process noise carrys out the deflection deformation of compensating Modeling.The method can also increase wave filter in addition to it can compensate flex motion
Robustness, but to reduce (Kain J.E., Cloutier J.R..Rapid of the precision of Transfer Alignment as cost
transfer alignment for tactical weapon applications[C].AIAA Guidance,
Navigation and Control Conference.1989:1290-1300.).The third method is empirical modeling method,
And more conventional method, will body elastic deformation angle be idealized as the markoff process of second order or three ranks, and by its
Augmentation for Kalman filtering state variable, by Kalman Filter Estimation go out the elastic deformation angle and compensate (Li Duanchang,
Zhong Maiying, Guo Dingfei. the error detection and compensation [C] in distributed POS Transfer Alignments. the 25th Chinese Control and decision making meeting
Collection of thesis .2013:4194-4199.).But the method exist problem be above-mentioned parameter selection more empirically determine, in reality
The accuracy of the parameter is difficult to ensure that in engineer applied.Additionally, the markoff process parameter at body elastic deformation angle is in aircraft
Can change with extraneous factor in flight course, even if can rule of thumb provide corresponding process ginseng in the filtering incipient stage
Number, cannot guarantee that the accuracy of System State Model and noise statisticses in filtering, and filtering accuracy will decline very
To appearance diverging.
The content of the invention
The present invention technology solve problem be:Overcome the deficiencies in the prior art, propose a kind of based on the airborne of parameter identification
Distributed POS Transfer Alignments, the method can improve the Transfer Alignment essence that carrier aircraft body has distributed POS during elastic deformation
Degree.
The present invention technical solution be:A kind of airborne distributed POS Transfer Alignments based on parameter identification.Its
Comprise the following steps that:
(1) body elastic deformation angle is considered as into second order markoff process, is set up comprising sub- IMU ins errors model and angle
The Transfer Alignment error model of error model;
(2) body elastic deformation angle and elastic deformation angular speed are extended for into state variable, and by high accuracy main POS and son
The difference of the speed and attitude of IMU sets up the Mathematical Modeling of sub- IMU Transfer Alignments as measurement;
(3) t is gone out using Kalman Filter EstimationkThe elastic deformation angle at momentWith elastic deformation angular speedK=0,1,
2 ..., N-1, using what is estimatedWithIn line computation and update description elastic deformation angle second order markoff process ginseng
Number;
(4) using the second order markoff process parameter after renewal as subsequent time tk+1Filtering initial value, using karr
Graceful filtering estimates more accurate tk+1The sub- IMU attitude errors at moment, velocity error and site error, using above-mentioned error
The strapdown calculation result of sub- IMU is corrected, t is obtainedk+1The position of Shi Kezi IMU, speed and attitude.
In above-mentioned steps (1), Transfer Alignment error model includes that the angle between sub- IMU ins errors model and master subsystem is missed
Differential mode type.Specifically Transfer Alignment error modeling step is:
1) set up sub- IMU ins errors model
The definition of coherent reference coordinate system includes:Note i is geocentric inertial coordinate system;E is terrestrial coordinate system;Main POS and son
IMU navigational coordinate systems are northeast day geographic coordinate system, respectively with n and n1Represent;Carrier coordinate system origin be carrier center of gravity, x
Along carrier transverse axis to the right, before carrier Y, along carrier vertical pivot upwards, the coordinate system is fixed on carrier z-axis y-axis axle, claims
For upper carrier coordinate system before the right side, the carrier coordinate system of main POS and sub- IMU is represented respectively with a and b;According to above-mentioned definition, sub- IMU
Ins error model is:
A) the attitude error differential equation:
B) the velocity error differential equation:
C) the site error differential equation:
D) the inertia type instrument error differential equation:
WhereinFor sub- IMU attitudes misalignment, φE、φNAnd φURespectively east orientation, north orientation, day
To misalignment, subscript E, N and U represent respectively east orientation, north orientation and day to;It is the angle speed of relative inertness system for sub- IMU navigation
Degree;ForError angular speed;The attitude battle array of its navigation system is tied to for sub- IMU carriersEstimate;WithRespectively sub- IMU speed and velocity error, wherein VE、VN
And VURespectively east orientation, north orientation and sky orientation speed, δ VE、δVNWith δ VURespectively east orientation, north orientation and sky orientation speed error;It is the specific force of sub- IMU, wherein fE、fNAnd fURespectively east orientation, north orientation and day are to specific force;WithThe angular speed and its error of respectively sub- IMU navigation system terrestrial coordinate system relatively;WithRespectively sub- IMU navigation
The angular speed and its error of the relative terrestrial coordinate system of system;L, λ, H and δ L, δ λ, δ H is respectively sub- IMU latitudes, longitude, height and latitude
Degree error, longitude error, height error;For the first derivative of latitude,For
The first derivative of longitude;RMAnd RNRespectively along meridian circle and the principal radius of curvature of prime vertical;εb=[εx εy εz]TWithRespectively sub- IMU gyroscope constant value drifts count constant value biasing, wherein ε with addingx、εyAnd εzIt is respectively sub
IMU carriers system x-axis, y-axis and z-axis gyroscope constant value drift, whereinWithRespectively sub- IMU carriers system x-axis, y-axis and z
Axle adds meter constant value biasing.
2) set up the angle error model between master subsystem
A) fixedly mount the differential equation of error angle ρ:
Wherein ρ=[ρx ρy ρz]TError angle, ρ are fixedly mounted for sub- IMUx、ρyAnd ρzRespectively sub- IMU carriers system x-axis, y
Axle and z-axis fix error angle.
B) differential equation at elastic deformation angle:
Wherein θjFor the elastic deformation angle on sub- IMU carriers system jth axle, j=x, y, z, θ=[θx θy θz]TFor elasticity change
Shape angle;βj=2.146/ τj, τjFor second order markoff process correlation time;ηjFor zero-mean white noise, its varianceMeet:
σj 2For elastic deformation angle θjVariance, β j andTo describe the ginseng of the second order markoff process of elastic deformation angle θ
Number.
The sub- IMU Transfer Alignments Mathematical Modeling that above-mentioned steps (2) are set up is:
Z=HX+V
Wherein systematic state variable X is:
X=[X1 X2]T
System transfer matrix F can be determined by the Transfer Alignment error model of sub- IMU;System noiseWhereinWithRespectively sub- IMU carriers system x-axis, y-axis, z-axis gyro and sub- IMU carriers system x-axis, y-axis, z-axis acceleration
The random error of meter, not including random constant error;White Gaussian noise of system noise W for zero-mean, its variance matrix Q is by gyro
Constant value drift, plus meter constant value biasing and second order markoff process parameter QηjDetermine;The expression formula of system noise acoustic matrix G is:
WhereinThe pose transformation matrix of navigation system is tied to for sub- IMU carriers.
System measurements variable Z=[δ ψ δ θ δ γ δ V 'E δV′N δV′U]T, wherein δ ψ, δ θ, δ γ and δ V 'E、δV′N、δ
V′UThe course angle of respectively sub- IMU and main POS, the angle of pitch, the difference of roll angle and east orientation, north orientation, the difference of sky orientation speed;Measure
NoiseWherein vδψ、vδθ、vδγPOS course angles based on respectively, the angle of pitch,
The measurement noise of roll angle,POS east orientations, north orientation, the measurement noise of sky orientation speed based on respectively;V is
The white Gaussian noise of zero-mean, its variance matrix R by main POS position and velocity accuracy determine;Measurement matrix H is:
Make main POS attitude matrixsNoteFor matrix TaIn l rows, m row element, l=1,2,3, m=
1,2,3;Then in above formulaWithExpression formula be:
T at the sub- IMU mount points of utilization adopted by above-mentioned steps (3)kThe elastic deformation angular estimation value at momentWith elastic deformation Attitude rate estimator valueOn-line identification describes elastic deformation
The second order markoff process parameter beta at anglejWithConcretely comprise the following steps:
1) determine valid data θjWith
NoteWithValid data number be Loop, initial t0Moment Loop=1;
Work as tkMoment elastic deformation Attitude rate estimator valueMeetWhen,
Loop=Loop+1,Conversely,
Loop=1,Wherein []kMiddle subscript k represents tkMoment, D () represent variance, γ1=2
And γ2=0.1 is respectively upper bound threshold parameter and lower bound threshold parameter;
2) calculate tkMoment parameterWith
Initial t0Moment, βjWithInitial value isWith
A) utilize vector θjWithCalculate varianceAnd covariance
B) giving iterative initial value isWith Newton Algorithm it is as follows with βjFor the nonlinear equation of independent variable,
Second order markoff process parameter is obtainedWith
WhereinNote
3) parameterThe Effective judgement of estimated result
Work as tkMoment parameter betajEstimateIt is unsatisfactory forWhen, order
Using the second order markoff process parameter after renewal as subsequent time t in above-mentioned steps (4)k+1Filtering it is initial
Value, goes out sub- IMU attitude errors, velocity error and site error using Kalman Filter Estimation, finally corrects sub- IMU strapdowns and resolves
As a result the step of is:
1) attitude error, velocity error and the site error of sub- IMU are estimated
By tkThe second order markoff process parameter that moment calculatesWithAs subsequent time tk+1Filtering
Initial value, goes out t using Kalman Filter Estimationk+1The sub- IMU the misaligned angle of the platform φ at momentE、φN、φU, velocity error δ VE、δVN、
δVUWith site error δ L, δ λ, δ H;
2) using the strapdown calculation result of the sub- IMU of above-mentioned error correction, obtain tk+1The sub- IMU positions at moment, speed and appearance
State
A) speed amendment
WhereinWithThe revised east orientation of respectively sub- IMU, north orientation and sky orientation speed;WithRespectively sub- IMU strapdowns resolve east orientation, north orientation and the sky orientation speed for obtaining;δVE、δVNWith δ VURespectively
For tk+1The sub- IMU strapdowns that moment Kalman Filter Estimation goes out resolve east orientation, north orientation and sky orientation speed error;
B) position correction
Lnew=Lold-δL
λnew=λold-δλ
Hnew=Hold-δH
Wherein Lold、λoldAnd HoldRespectively sub- IMU strapdowns resolve latitude, the longitude and altitude for obtaining;Lnew、λnewAnd Hnew
The revised latitude of respectively sub- IMU, longitude and altitude;δ L, δ λ and δ H are respectively tk+1The son that moment Kalman Filter Estimation goes out
IMU strapdowns resolve latitude, longitude and altitude error;
C) attitude rectification
Calculate tk+1Shi Kezi IMU geographic coordinate systems n1With computed geographical coordinates n '1Between transition matrix
Calculate tk+1Shi Kezi IMU carrier coordinate systems b and true geographic coordinate system n1Between transition matrix
WhereinFor tk+1Shi Kezi IMU strapdowns resolve the attitude matrix for obtaining;
By the attitude battle array of the sub- IMU after being updatedCalculate tk+1Course angle ψ of Shi Kezi IMUs, pitching angle thetasAnd roll
Angle γs, willIt is designated as
Wherein TlmFor matrixIn l rows, m row element, l=1,2,3, m=1,2,3;Then sub- IMU course angles ψs、
Pitching angle thetasWith roll angle γsMain value, i.e. ψS master、θS masterAnd γS masterRespectively:
θS master=arcsin (T32)
Due to course angle ψs, pitching angle thetasWith roll angle γsSpan be respectively defined as [0,2 π],[- π ,+π];So, ψs、θsAnd γsTrue value can be determined by following formula:
θs=θS master
It is modified by the speed of antithetical phrase IMU, position and attitude, more accurate sub- IMU mount points can be obtained
Speed, position and attitude information, complete Transfer Alignment.
Present invention advantage compared with prior art is:
For the markoff process unknown parameters and the problem of time-varying of body elastic deformation described in practical application, utilize
In Transfer Alignment the speed and attitude information of the main POS of high accuracy can the advantage that is modified of antithetical phrase IMU state variables, profit
The elastic deformation angle that gone out with Kalman Filter Estimation and elastic deformation angular speed calculating markoff process parameter, so as to realize
The on-line identification of the parameter.Overcome preset parameter method that is traditional, empirically determining and cannot truly describe body elasticity
Deformation cause Transfer Alignment precision low deficiency, and not by carrier aircraft material constraint and whether plug-in load is affected, raising
The accuracy of System State Model and noise statisticses, so as to improve the precision of Transfer Alignment.
Description of the drawings
Fig. 1 is the flow chart of the present invention;
Fig. 2 is the flow chart of the second order markoff process parameter identification method of the present invention;
Specific embodiment
As shown in figure 1, the concrete grammar of the present invention implements as follows:
1st, set up the Transfer Alignment error model comprising angle error model between sub- IMU ins errors model and master subsystem
Specifically Transfer Alignment error modeling step is:
(1) set up sub- IMU ins errors model
The definition of coherent reference coordinate system includes:Note i is geocentric inertial coordinate system;E is terrestrial coordinate system;Main POS and son
IMU navigational coordinate systems are northeast day geographic coordinate system, respectively with n and n1Represent;Carrier coordinate system origin be carrier center of gravity, x
Along carrier transverse axis to the right, before carrier Y, along carrier vertical pivot upwards, the coordinate system is fixed on carrier z-axis y-axis axle, claims
For upper carrier coordinate system before the right side, the carrier coordinate system of main POS and sub- IMU is represented respectively with a and b;According to above-mentioned definition, sub- IMU
Ins error model is:
A) attitude error equation:
B) velocity error equation:
C) site error equation:
D) inertia type instrument error equation:
WhereinFor sub- IMU attitudes misalignment, φE、φNAnd φURespectively east orientation, north orientation, day
To misalignment, subscript E, N and U represent respectively east orientation, north orientation and day to;It is the angle speed of relative inertness system for sub- IMU navigation
Degree;ForError angular speed;The attitude battle array of its navigation system is tied to for sub- IMU carriersEstimate;WithRespectively sub- IMU speed and velocity error, wherein VE、VN
And VURespectively east orientation, north orientation and sky orientation speed, δ VE、δVNWith δ VURespectively east orientation, north orientation and sky orientation speed error;It is the specific force of sub- IMU, wherein fE、fNAnd fURespectively east orientation, north orientation and day are to specific force;WithThe angular speed and its error of respectively sub- IMU navigation system terrestrial coordinate system relatively;WithRespectively sub- IMU navigation
The angular speed and its error of the relative terrestrial coordinate system of system;L, λ, H and δ L, δ λ, δ H is respectively sub- IMU latitudes, longitude, height and latitude
Degree error, longitude error, height error;For the first derivative of latitude,For
The first derivative of longitude;RMAnd RNRespectively along meridian circle and the principal radius of curvature of prime vertical;εb=[εx εy εz]TWithRespectively sub- IMU gyroscope constant value drifts count constant value biasing, wherein ε with addingx、εyAnd εzIt is respectively sub
IMU carriers system x-axis, y-axis and z-axis gyroscope constant value drift, whereinWithRespectively sub- IMU carriers system x-axis, y-axis and z
Axle adds meter constant value biasing.
(2) set up the angle error model between master subsystem
Angle error model between master subsystem fixedly mounts error angle ρ=[ρ by sub- IMUx ρy ρz]TBecome with body elasticity
Shape angle θ=[θx θy θz]TTogether decide on, wherein ρx、ρyAnd ρzRespectively sub- IMU carriers system x-axis, y-axis and z-axis alignment error
Angle, θx、θyAnd θzRespectively sub- IMU carriers system x-axis, y-axis and z-axis elastic deformation angle.
A) fixedly mount the differential equation of error angle ρ:
B) differential equation at elastic deformation angle:
Wherein j=x, y, z, βj=2.146/ τj, τjFor second order markoff process correlation time;ηjFor zero-mean white noise
Sound, its varianceMeet:
σj 2For elastic deformation angle θjVariance.β j andTo describe the second order of elastic deformation angle θ
The parameter of markoff process.
2nd, set up the Mathematical Modeling of sub- IMU Transfer Alignments
According to airborne distributed POS Transfer Alignments error model in 1 by body elastic deformation angle and elastic deformation angular speed
State variable is extended for, and using the difference of the speed and attitude of high accuracy main POS and sub- IMU as measurement, sets up sub- IMU transmission right
Accurate Mathematical Modeling is:
Wherein systematic state variable X is:
X=[X1 X2]T
System transfer matrix F can be determined by the Transfer Alignment error model of sub- IMU;System noiseWhereinWithRespectively sub- IMU carriers system x-axis, y-axis, z-axis gyro and sub- IMU carriers system x-axis, y-axis, z-axis accelerate
The random error of degree meter, not including random constant error;White Gaussian noise of system noise W for zero-mean, its variance matrix Q is by top
Spiral shell constant value drift, plus meter constant value biasing and second order markoff process parameterDetermine;The expression formula of system noise acoustic matrix G is:
WhereinThe pose transformation matrix of navigation system is tied to for sub- IMU carriers.
System measurements variable Z=[δ ψ δ θ δ γ δ V 'E δV′N δV′U]T, wherein δ ψ, δ θ, δ γ and δ V 'E、δV′N、δ
V′UThe course angle of respectively sub- IMU and main POS, the angle of pitch, the difference of roll angle and east orientation, north orientation, the difference of sky orientation speed;Measure
NoiseWherein vδψ、vδθ、vδγPOS course angles based on respectively, the angle of pitch,
The measurement noise of roll angle,POS east orientations, north orientation, the measurement noise of sky orientation speed based on respectively;V is
The white Gaussian noise of zero-mean, its variance matrix R by main POS position and velocity accuracy determine;Measurement matrix H is:
Make main POS attitude matrixsNoteFor matrix TaIn l rows, m row element, l=1,2,3, m=
1,2,3;Then in above formulaWithExpression formula be:
3rd, on-line identification describes the second order markoff process parameter at elastic deformation angle
T at sub- IMU mount points is gone out using Kalman Filter EstimationkThe elastic deformation angle at moment (k=0,1,2 ..., N-1)
With elastic deformation Attitude rate estimator valueWithUsing what is estimatedWithOn-line identification description elasticity becomes
The second order markoff process parameter beta at shape anglejWithConcretely comprise the following steps:
(1) t at sub- IMU mount points is gone out using Kalman Filter EstimationkThe elastic deformation angle at momentWith elastic deformation angle
Velocity estimation valueInitial t0Moment gives Kalman filtering algorithm layout is as follows:
State one-step prediction equation
WhereinΦk,k-1Respectively tkMoment system mode one-step prediction value, tk-1Moment system mode is estimated
Value, tk-1Moment is to tkThe systematic state transfer matrix at moment;
State Estimation accounting equation
WhereinKk、Zk、HkRespectively tkMoment system mode valuation, system gain matrix, measurement vector sum measure square
Battle array;
Filtering gain equation
Wherein Pk/k-1、RkRespectively tkWhen the one-step prediction of etching system covariance matrix, tkMoment system measurements noise matrix;
One-step prediction mean square error equation
Wherein, Pk-1、Qk-1、Γk-1Respectively tk-1When etching system covariance matrix, tk-1Moment system noise matrix, tk-1When
Etching system noise drives matrix;
Estimate mean square error equation
Wherein, PkFor tkMoment system mode covariance matrix, I are unit battle array.
(2) second order markoff process parameter identification
Go out t at sub- IMU mount points using Kalman Filter EstimationkThe elastic deformation angle at moment and elastic deformation angular speed are estimated
EvaluationWithWith on-line identification and the second order markoff process parameter beta for describing elastic deformation angle can be updatedj
WithSpecific parameter distinguishes that calculation procedure is:
A) determine valid data θjWith
In filtering, second order markoff process parameter may occur large change at certain moment, now recycle
Estimate before Parameters variationWithCalculating nonlinear equation coefficient can cause procedure parameter estimate to there is larger error, because
This must add determination in parameter identification methodWithValid data θjWithPart.For according to elastic deformation angle speed
Degree can effectively reflect System State Model parameter betajSentencing for Parameters variation in the characteristics of change design parameter discrimination method
Disconnected part, so that it is determined that valid data θjWithThe setting principle of Rule of judgment can be sketched to give estimateOne is put
Letter is interval, i.e., when the estimate related to parameter belongs to confidential interval, that is, think that the moment relevant parameter does not change,
Otherwise then critical parameter changes.Specifically Rule of judgment is:
Wherein []kMiddle subscript k represents moment tk, D () expression variances, γ1=2 and γ2=0.1 be respectively the upper bound and
Threshold parameter lower bound threshold parameter.
NoteWithValid data number be Loop, initial t0Moment Loop=1.Determine valid data θjWithIt is concrete
Step is:
Work as tkMoment elastic deformation Attitude rate estimator valueMeetWhen,
Loop=Loop+1,Conversely,
Loop=1,
B) calculate tkMoment parameterWith
Due to elastic deformation angle θjSecond order markoff process shown in meeting formula (6), takes variance to (6) the right and left,
Convolution (7) can be obtained:
Wherein
With Newton Algorithm with βjNonlinear equation (20) for independent variable is obtained second order markoff process parameterWithIn practical situations both, aircraft elastic deformation correlation time τj>0 and be more than or equal to 1.Therefore, tkMoment newton
Iterative initial value is given before method iterationInitial t0Moment, βjWithInitial value isWithCalculate tkMoment
ParameterWithConcretely comprise the following steps:
I) utilize vector θjWithCalculate varianceAnd covariance
Ii) giving iterative initial value isWith Newton Algorithm it is as follows with βjFor the nonlinear equation of independent variable
(20), you can obtain second order markoff process parameterWithNote
C) parameterThe Effective judgement of estimated result
As the solution that nonlinear equation Newton method is tried to achieve is the approximate solution near initial value, it is therefore desirable to judge iterative approximation
The validity of solution.Due in practical situations both, aircraft elastic deformation correlation time τj>0 and be more than or equal to 1, therefore for
The judgement of iterative solution validity can provide following Rule of judgment:
WhereinFor tkThe moment parameter beta that moment is calculated by equation (20)jEstimate.
Can sketch m odel validity judgment part:Work as tkMoment parameter betajEstimateIt is unsatisfactory forWhen, orderThe particular flow sheet of second order Markov Parameters discrimination method is as shown in Figure 2.
4th, correct sub- IMU strapdowns resolve obtain speed, position, attitude information
Using the second order markoff process parameter after renewal as subsequent time tk+1Filtering initial value, so as to estimate
More accurate sub- IMU attitude errors, velocity error and site error, are resolved using the strapdown of the sub- IMU of the evaluated error amendment
As a result, more accurate sub- IMU positions, speed and attitude are calculated.Specifically amendment step is:
(1) attitude error, velocity error and the site error of sub- IMU are estimated
By tkThe second order markoff process parameter that moment calculatesWithAs subsequent time (tk+1) filter
Ripple initial value, goes out t using Kalman Filter Estimationk+1The sub- IMU the misaligned angle of the platform (φ at momentE、φN、φU), velocity error (δ
VE、δVN、δVU) and site error (δ L, δ λ, δ H).
A) speed amendment
WhereinWithThe revised east orientation of respectively sub- IMU, north orientation and sky orientation speed;WithRespectively sub- IMU strapdowns resolve east orientation, north orientation and the sky orientation speed for obtaining;δVE、δVNWith δ VURespectively
For tk+1The sub- IMU strapdowns that moment Kalman Filter Estimation goes out resolve east orientation, north orientation and sky orientation speed error.
B) position correction
Wherein Lold、λoldAnd HoldRespectively sub- IMU strapdowns resolve latitude, the longitude and altitude for obtaining;Lnew、λnewAnd Hnew
The revised latitude of respectively sub- IMU, longitude and altitude;δ L, δ λ and δ H are respectively tk+1The son that moment Kalman Filter Estimation goes out
IMU strapdowns resolve latitude, longitude and altitude error.
C) attitude rectification
Calculate tk+1Shi Kezi IMU geographic coordinate systems n1With computed geographical coordinates n '1Between transition matrix
Calculate tk+1Shi Kezi IMU carrier coordinate systems b and true geographic coordinate system n1Between transition matrix
WhereinFor tk+1Shi Kezi IMU strapdowns resolve the attitude matrix for obtaining.
By the attitude battle array of the sub- IMU after being updatedCalculate course angle ψ of sub- IMU mount pointss, pitching angle thetasAnd roll angle
γs.WillIt is designated as
And because
Wherein TlmFor matrixIn l rows, m row element, l=1,2,3, m=1,2,3;Then sub- IMU course angles ψs、
Pitching angle thetasWith roll angle γsMain value, i.e. ψS master、θS masterAnd γS masterRespectively:
Due to course angle ψs, pitching angle thetasWith roll angle γsSpan be respectively defined as [0,2 π],
[- π ,+π].So, ψs、θsAnd γsTrue value can be determined by following formula:
θs=θS master (30)
It is modified by the speed of antithetical phrase IMU, position and attitude, more accurate sub- IMU mount points can be obtained
Speed, position and attitude information, complete Transfer Alignment.
The content not being described in detail in description of the invention belongs to prior art known to professional and technical personnel in the field.