CN102901971B - Parity vector method-based double-satellite failure recognition method - Google Patents

Abstract

The invention relates to global satellite navigation system receiver autonomous integrity monitoring technology and discloses a parity vector method-based double-satellite failure recognition method, aiming at the problems of false positives and false negatives caused by fault deviation offsetting when the parity vector method is used for recognizing two fault satellites. According to the technical scheme, the parity vector method is used for recognizing one fault satellite; with the fault satellite as the basis, four fault-free satellites are found out, and the information of the fault-free satellites is used for roughly locating, so that the fault satellites can be recognized; the recognized fault satellites are removed, and then the position resolution is carried out again, so that the locating accuracy is improved; therefore, the problems of false positives or false negatives caused by fault deviation offsetting can be avoided. The method solves the problem of the fault deviation offsetting caused by parity vector residual error and realizes the detection and the recognition for a plurality of fault satellites. After the method is used for detecting and recognizing satellite failure, the locating accuracy is improved. The method is mainly used for monitoring the autonomous integrity of a global satellite navigation system receiver.

Description

Double star fault recognition method based on parity vector method

Technical field

The present invention relates to satellite navigation and location technology, be particularly related to GPS (Global Position System) (Global Navigation Satellite System, GNSS) receiver autonomous integrity monitoring (Receiver Autonomous Integrity Monitoring, RAIM) method.

Background technology

GNSS receiver, in obtaining navigational solution, must be taken the impact of pseudorange deviation on navigational solution into account.The principal element that causes pseudorange deviation has that larger satellite clock floats, the fault of the incorrect and satellite ingredient of navigation message data, is referred to as satellite failure here.Because system cannot guarantee the reaction time to satellite failure, the fast monitored of satellite failure is only carried out at user side so, thereby has occurred receiver autonomous integrity monitoring method.

RAIM utilizes redundancy distance measuring signal to detect to cause the satellite mistake of larger deviations.Conventionally RAIM comprises two functions: fault detect (Fault Detection) and Fault Identification (Fault Identification).Comparatively conventional RAIM method comprises: Agonists by Distance Comparison Method, least square method and parity vector method.But these methods are all under the supposed premise based on a fault satellites.Along with overhead satellites number constantly increases, the probability that two or multi-satellite break down simultaneously can not be left in the basket again.

Traditional fault satellites recognition methods can not be directly used in many fault satellites identifications, and not perfect about the recognition methods of many fault satellites.For example, parity vector method there will be fault deviation to offset while identification for two fault satellites, and brings thus the situation of failing to judge and judging by accident.

The principle of a fault satellites of parity vector method identification is as follows:

The basic observation equation of receiver positioning calculation is:

y＝Hx+e (1)

In formula (1), y is the residual vector of Pseudo-range Observations, and H is observing matrix, and x is receiver location residual error and receiver clock error to be calculated, the measuring error that e is Pseudo-range Observations.

In the time that non-fault satellite exists, think that measuring error e obeys variance and is

normal distribution; In the time having fault satellites to exist, there is deviation in the pseudorange of fault satellites, and now, e is Normal Distribution no longer.

When application parity vector method is carried out fault satellites detection with identification, first observing matrix H is decomposed into the long-pending of an orthogonal matrix Q and a upper triangular matrix R, even:

H＝QR (2)

By the transposed matrix Q of matrix Q ^tpiecemeal:

$Q^T=\left[\begin{array}{c}{Q}_x\\ Q_p\end{array}\right]---\left(3\right)$

In formula (3), Q _p∈ R ^{(n-4) × n}for odd even space matrix, wherein n is number of satellite, Q _xfor Q ^tfront four lines:

p＝Q _py＝Q _pe (4)

In formula (4), p is parity vector.Pseudorange residual sum of squares (RSS) SSE can construct with parity vector p:

SSE＝p ^Tp (5)

While not there is not fault satellites,

obeying degree of freedom is the centralization card side distribution of (n-4),

while there is fault satellites,

the card side that obeys decentralization distributes,

λ is decentralization parameter.Therefore, can construct fault detect judgement amount

${\widehat{\mathrm{\σ}}}_n=\sqrt{\frac{\mathrm{SSE}}{n-4}}---\left(6\right)$

As overall false-alarm probability α ₀after given, can calculate fault detect thresholding σ _{t, n}for:

$\left\{\begin{array}{c}{\mathrm{\σ}}_{T,n}=\frac{{\mathrm{\σ}}_0{T}_n}{\sqrt{n-4}}\\ T_n=\mathrm{icdf}\left({\mathrm{\χ}}^2(1-{\mathrm{\α}}_0,n-4)\right)\end{array}\right.---\left(7\right)$

In formula (7), icdf is (1-α) fractile that computer card side distributes.

time think have fault satellites to exist, and then identification fault satellites, otherwise, think non-fault satellite, the receiver information of current calculating is accurately.The fault detect judgement amount here

with fault detect thresholding σ _{t, n}, for carrying out fault detect.

From formula (4), known parity vector p is through Q by observational error e _pprojection gets, and can carry out fault satellites identification with the geometric properties between them.Structure Fault Identification judgement amount:

$r_i=\frac{\left|\right|p{Q}_{p,i}\left|\right|}{\left|\right|Q_{p,i}\left|\right|}---\left(8\right)$

When all satellites are during without pseudorange deviation,

therefore work as the false-alarm probability α of single satellite ₁give timing, can calculate Fault Identification thresholding:

$T_{r,n}={\mathrm{\σ}}_0{\mathrm{erf}}^{-1}\left(\frac{{\mathrm{\α}}_1}{2n}\right)---\left(9\right)$

In formula (9), erf (x) is Gauss error function, erf ^-1(x) be its inverse function.Work as r _ibe greater than T _rtime, can judge that i satellite is fault satellites, there is deviation in its pseudorange.After i satellite rejected, utilize all the other satellites again to position and resolve to improve positioning precision.The Fault Identification judgement amount r of formula (8) and formula (9) _iwith Fault Identification thresholding T _{r, n}, for carrying out Fault Identification.

Summary of the invention

Technical matters to be solved by this invention, there will be fault deviation to offset while identification exactly, and bring thus the shortcoming of failing to judge and judging by accident for parity vector method for two fault satellites, and a kind of double star fault recognition method based on parity vector method is provided.

The present invention solve the technical problem, and the technical scheme of employing is that the double star fault recognition method based on parity vector method, comprises step:

(1), from an existing n satellite, reject satellite i, n is receiver visible satellite quantity in global position system, n >=7;

(2), to remaining (n-1) satellite according to fault detect judgement amount with fault detect thresholding σ _{t, (n-1)}carry out fault detect, if

think that remaining non-fault satellite in satellite exists, and detects and finishes; Otherwise, carry out fault satellites identification by parity vector method, according to Fault Identification judgement amount r _imaximal value r _j=max (r ₁, r ₂..., r _n), satellite j is rejected;

(3), to remaining (n-2) satellite according to fault detect judgement amount

with fault detect thresholding σ _{t, (n-2)}carry out fault detect, if

confirm satellite i non-fault, make non-fault number of satellite add one, enter step (4); Otherwise, the combinations of satellites (i that record is rejected _k, j _k) and calculate position and the clock correction x of receiver by least square method _k, forward step (5) to;

(4), judge whether the non-fault number of satellite of current gained reaches 4, if reach 4 rough position and clock correction that calculate receiver

and identify fault satellites from residue (n-4) satellite, and again resolve receiver location and clock correction, detect and finish; Otherwise, enter step (5);

(5), make i=i+1, if i > n, non-fault number of satellite is less than 4, now, from reject combinations of satellites with pseudorange residual error statistic λ _kminimum value λ _min=min (λ ₁, λ ₂..., λ _m) corresponding combinations of satellites is as fault satellites, and calculate the receiver location corresponding with this combinations of satellites and clock correction as positioning result, detect and finish; Otherwise return to step (1).

Concrete, fault detect judgement amount

obtained by following formula:

${\widehat{\mathrm{\σ}}}_{n-1}=\sqrt{\frac{\mathrm{SSE}}{(n-1)-4}}$

Wherein, SSE is pseudorange residual sum of squares (RSS);

Fault detect thresholding σ _{t, (n-1)}obtained by following formula:

$\left\{\begin{array}{c}{\mathrm{\σ}}_{T,(n-1)}=\frac{{\mathrm{\σ}}_0{T}_{n-1}}{\sqrt{(n-1)-4}}\\ T_{n-1}=\mathrm{icdf}\left({\mathrm{\χ}}^2(1-\mathrm{\α},(n-1)-4)\right)\end{array}\right.$

Wherein, icdf is (1-α) fractile that computer card side distributes;

Fault Identification judgement amount r _iobtained by following formula:

$r_i=\frac{\left|\right|p{Q}_{p.i}\left|\right|}{\left|\right|Q_{p.i}\left|\right|}.$

Concrete, detect judgement amount

obtained by following formula:

${\widehat{\mathrm{\σ}}}_{n-2}=\sqrt{\frac{\mathrm{SSE}}{(n-2)-4}}$

Wherein, SSE is pseudorange residual sum of squares (RSS);

Fault detect thresholding σ _{t, (n-2)}obtained by following formula:

$\left\{\begin{array}{c}{\mathrm{\σ}}_{T,(n-2)}=\frac{{\mathrm{\σ}}_0{T}_{n-2}}{\sqrt{(n-2)-4}}\\ T_{n-2}=\mathrm{icdf}\left({\mathrm{\χ}}^2(1-\mathrm{\α},(n-2)-4)\right)\end{array}\right.$

Wherein, icdf is (1-α) fractile that computer card side distributes.

Concrete, rough position and clock correction

obtained by following formula:

$\hat{x}={\left(H_0^T{H}_0\right)}^{-1}{H}_0^T{y}_0$

Wherein, H ₀for 4 observing matrixes that non-fault satellite forms of current gained,

for H ₀transposed matrix, y ₀for the Pseudo-range Observations vector of 4 non-fault satellites of current gained.

Concrete, receiver location and clock correction x _kobtained by following formula:

$x_k={\left(H_k^T{H}_k\right)}^{-1}{H}_k^T{y}_k$ k＝1,2,…,N

Wherein, H _kfor having rejected two satellite (i _k, j _k) after the observing matrix that forms of remaining (n-2) satellite, y _kfor having rejected two satellite (i _k, j _k) after the Pseudo-range Observations vector of remaining (n-2) satellite;

Pseudorange residual error statistic λ _kobtained by following formula:

${\mathrm{\λ}}_k=\underset{\underset{h\≠i,j}{h=1}}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{\ρ}}_h-b_k-\sqrt{{(x_h-x_k)}^2+{(y_h-y_k)}^2+{(z_h-z_k)}^2})}^2.$

The invention has the beneficial effects as follows, solved the fault deviation cancellation problem of parity vector residual error, realized the detection identification of many fault satellites.Detect identification satellite fault by the present invention, can improve positioning precision.

Accompanying drawing explanation

Fig. 1 is process flow diagram of the present invention.

Embodiment

Below in conjunction with drawings and Examples, describe technical scheme of the present invention in detail.

The present invention is take fault satellites of parity vector method identification as basis, find out 4 non-fault satellites, then carry out coarse localization by the information of non-fault satellite, and identify thus fault satellites, after being rejected, re-start again location compute to improve positioning precision, thereby avoided being offset and failing to judge or erroneous judgement problem of bringing by fault deviation.

Of the present inventionly mainly contain two gordian techniquies, the one, the identification of non-fault satellite take fault satellites of parity vector method identification as basis, is found out 4 non-fault satellites in order to coarse localization from an existing n satellite; The 2nd, calculated rough position and the clock correction of receiver by the information of 4 non-fault satellites

wherein

the rough coordinates of receiver,

receiver clock correction.And by

calculate the pseudorange residual delta of all the other (n-4) satellites _i, and determine detection threshold according to the false-alarm probability of fault satellites identification, identify fault satellites.

Elaborate with regard to two gordian techniquies below:

1, non-fault satellite recognition principle

If there are two fault satellites to exist, after successively having rejected two satellites by parity vector method in n satellite, if be still greater than fault detect thresholding according to the fault detect judgement amount of remaining (n-2) satellite calculating gained, that is:

first satellite rejecting must be non-fault satellite, do not have deviation in pseudorange.The principle of identification non-fault satellite can be divided into following two kinds of situations and discuss:

(1) if be still less than fault detect thresholding according to the fault detect judgement amount of remaining (n-2) satellite calculating gained, that is:

time, in remaining (n-2) satellite, may not there is not fault satellites or have two fault satellites, but under the effect of offsetting in fault deviation, make

cannot judge in this case whether first rejected satellite is fault satellites.

(2) if be still greater than fault detect thresholding according to the fault detect judgement amount of remaining (n-2) satellite calculating gained, that is:

time, in remaining (n-2) satellite, may there is a fault satellites; Also now fault deviation counteracting is not obvious, has two fault satellites.No matter be which kind of situation, first satellite of rejecting must be all non-fault satellite, otherwise, in remaining (n-1) satellite, only have a fault satellites to exist, now adopt parity vector method another fault satellites can be identified, thereby

If it should be noted that note: for event A; First satellite of rejecting is that non-fault satellite is event B.

Discuss for following three propositions:

Proposition 1: if B is A.This is a wrong proposition, when second satellite of rejecting is during also for satellite without reason, has two fault satellites, if the deviation that now breaks down is offset in remaining (n-2) satellite therefore, even if first satellite is non-fault satellite, also be likely less than σ _{t (n-2)}.

Proposition 2: if

this proposition is the converse negative proposition of proposition 1, is also therefore wrong, and also just explanation can not be because of for it

just judge that two satellites of rejecting are fault satellites.

Proposition 3: if

even reject first satellite be fault satellites this proposition is set up, in the time that first satellite of rejecting is fault satellites, in remaining (n-1) satellite, only there is a fault satellites, use parity vector method this satellite can be identified, so non-fault satellite exists in remaining (n-2) satellite, therefore ${\widehat{\mathrm{\&sigma;}}}_{n-2}<{\mathrm{\&sigma;}}_{T,(n-2)}.$

Method and proposition 3 that the present invention identifies non-fault satellite are mutually converse negative proposition, thereby the correctness of identification non-fault satellite method has been described.

2, fault satellites recognition principle

Can detect 4 non-fault satellites by method above, the information of 4 non-fault satellites just can calculate by the method for least square rough position and the receiver clock correction of receiver thus

Work as acquisition after can compute pseudo-ranges residual delta _i:

${\mathrm{\&Delta;}}_i=|{\mathrm{\&rho;}}_i-\hat{b}-\sqrt{{(x_i-\hat{x})}^2+{(y_i-\hat{y})}^2+{(z_i-\hat{z})}^2}---\left(10\right)$

In the time there is not pseudorange deviation in satellite i, pseudorange observational error

otherwise, e _ibe not 0 Gaussian distribution by obeying average.Therefore, as single satellite false-alarm probability α ₁have to timing:

$P({\mathrm{\&Delta;}}_i>T_{\mathrm{\&Delta;}})={\&Integral;}_{T_{\mathrm{\&Delta;}}}^{+\&infin;}\frac{1}{\sqrt{2\mathrm{\&pi;}}{\mathrm{\&sigma;}}_0}\exp (-\frac{{{\mathrm{\&Delta;}}_i}^2}{2{\mathrm{\&sigma;}}^2})d{\mathrm{\&Delta;}}_i=1-\frac{{\mathrm{\&alpha;}}_1}{2}---\left(11\right)$

Through type (11) can calculate fault satellites identification thresholding T _Δ.Δ _iwith T _Δall identify for fault satellites.

Work as Δ _i> T _Δtime can judge that satellite i is as fault satellites, can realize thus the identification of two fault satellites.

In addition, when fault satellites is during more than one, due to the cause that fault deviation is offset, the Fault Identification judgement amount r of parity vector method _i(i=1,2 ..., maximal value r n) _j=max (r ₁, r ₂..., r _n), r _maxcorresponding satellite j may not be fault satellites.And from reducing the angle of calculated amount, in the time realizing, the rejecting of first satellite can adopt the method for rejecting one by one, in the time of second fault satellites of identification, re-use parity vector method.

From non-fault satellite recognition principle, the present invention is mainly by rejecting the fault detect judgement amount after two satellites

decide whether first satellite of rejecting is fault satellites.Calculating

time, need to meet n-2>=5, therefore, only have and in the time that number of satellite is more than or equal to 7 (n>=7), just can carry out double star fault detect.But when number of satellite is less and fault deviation is offset when serious, even if all satellites are all searched one time, the non-fault number of satellite obtaining may still be less than 4.Now can not calculate the coarse information of receiver, thereby cannot carry out fault satellites identification.

It is right that non-fault number of satellite is less than 4 o'clock needs time all satellites of rejecting combination (i _k, j _k) test.Corresponding every kind of fault satellites combination, can be calculated receiver coordinate and the clock correction x of one group with residue (n-2) satellite by least square method _k=(x _k, y _k, z _k, b _k) k=1,2 ..., m, m is corresponding all combinations of satellites numbers.

Equally according to x _kcan calculate the pseudorange residual error statistic λ of residue (n-2) satellite _k, as decision statistics:

${\mathrm{\&lambda;}}_k=\underset{\underset{h\&NotEqual;i,j}{h=1}}{\overset{n}{\mathrm{\&Sigma;}}}{({\mathrm{\&rho;}}_h-b_k-\sqrt{{(x_h-x_k)}^2+{(y_h-y_k)}^2+{(z_h-z_k)}^2})}^2---\left(12\right)$

If λ _l=min (λ ₁, λ ₂..., λ _m), (i _l, j _l) for fault satellites combination, after rejecting, position and resolve with residue satellite again, obtain receiver exact position.

Embodiment

Fig. 1 has provided flow process of the present invention.Its concrete grammar is as follows:

(1) when, reject satellite i from an existing n satellite, initial, i is 1;

(2), remaining (n-1) satellite is carried out to fault detect, fault detect judgement amount

${\widehat{\mathrm{\&sigma;}}}_{n-1}=\sqrt{\frac{\mathrm{SSE}}{(n-1)-4}}---\left(13\right)$

Calculate the fault detect thresholding σ of (n-1) satellite _{t, (n-1)}:

$\left\{\begin{array}{c}{\mathrm{\&sigma;}}_{T,(n-1)}=\frac{{\mathrm{\&sigma;}}_0{T}_{n-1}}{\sqrt{(n-1)-4}}\\ T_{n-1}=\mathrm{icdf}\left({\mathrm{\&chi;}}^2(1-\mathrm{\&alpha;},(n-1)-4)\right)\end{array}\right.---\left(14\right)$

If

think that remaining non-fault satellite in satellite exists, and detects and finishes; Otherwise, carry out fault satellites identification by parity vector method, calculate Fault Identification judgement amount r _i:

$r_i=\frac{\left|\right|p{Q}_{p.i}\left|\right|}{\left|\right|Q_{p.i}\left|\right|}---\left(15\right)$

Try to achieve r _i(i=1,2 ..., maximal value r n) _j=max (r ₁, r ₂..., r _n), after being rejected, satellite j enters step (3);

(3), remaining (n-2) satellite is carried out to fault detect, the fault detect judgement amount of calculating (n-2) satellite:

${\widehat{\mathrm{\&sigma;}}}_{n-2}=\sqrt{\frac{\mathrm{SSE}}{(n-2)-4}}---\left(16\right)$

Wherein, SSE is pseudorange residual sum of squares (RSS);

Barrier detection threshold σ _{t, (n-2)}obtained by following formula:

$\left\{\begin{array}{c}{\mathrm{\&sigma;}}_{T,(n-2)}=\frac{{\mathrm{\&sigma;}}_0{T}_{n-2}}{\sqrt{(n-2)-4}}\\ T_{n-2}=\mathrm{icdf}\left({\mathrm{\&chi;}}^2(1-\mathrm{\&alpha;},(n-2)-4)\right)\end{array}\right.---\left(17\right)$

Wherein, icdf is (1-α) fractile that computer card side distributes.

If

confirm satellite i non-fault, make non-fault number of satellite add one, enter step (4); Otherwise, the combinations of satellites (i that record is rejected _k, j _k) and calculate position and the clock correction x of receiver by least square method _k,

k=1,2 ..., N, wherein, H _kfor having rejected two satellite (i _k, j _k) after the observing matrix that forms of remaining (n-2) satellite, y _kfor having rejected two satellite (i _k, j _k) after the Pseudo-range Observations vector of remaining (n-2) satellite.Forward step (5) to;

(4), judge whether the non-fault number of satellite of current gained reaches 4, if reach 4 rough position and clock correction that calculate receiver

$\hat{x}={\left(H_0^T{H}_0\right)}^{-1}{H}_0^T{y}_0---\left(18\right)$

In formula (18), H ₀for 4 observing matrixes that non-fault satellite forms of current gained,

for H ₀transposed matrix, y ₀for the Pseudo-range Observations vector of 4 non-fault satellites of current gained.

Obtain

after, can carry out compute pseudo-ranges residual delta according to formula (10) and formula (11) _iwith fault satellites identification thresholding T _Δ, and identify fault satellites from remaining (n-4) satellite, and again resolve receiver location and clock correction information, detect and finish; Otherwise enter step (5);

(5), make i=i+1, if i > n, non-fault number of satellite is less than 4, now, according to x _kobtain pseudorange residual error statistic λ _k, from reject combinations of satellites with pseudorange residual error statistic λ _kminimum value λ _min=min (λ ₁, λ ₂..., λ _m) corresponding combinations of satellites is as fault satellites, and calculate the receiver location corresponding with this combinations of satellites and clock correction as positioning result, detect and finish; Otherwise return to step (1).

Claims (5)

1. the double star fault recognition method based on parity vector method, comprises step:

(1), from an existing n satellite, reject satellite i, n is receiver visible satellite quantity in global position system, n >=7;

(2), to remaining (n-1) satellite according to fault detect judgement amount

with fault detect thresholding σ _{t, (n-1)}carry out fault detect, if

think that remaining non-fault satellite in satellite exists, and detects and finishes; Otherwise, carry out fault satellites identification by parity vector method, according to Fault Identification judgement amount r _imaximal value r _j=max (r ₁, r ₂, r _n), satellite j is rejected;

(3), to remaining (n-2) satellite according to fault detect judgement amount

with fault detect thresholding σ _{t, (n-2)}carry out fault detect, if

confirm satellite i non-fault, make non-fault number of satellite add one, enter step (4); Otherwise, the combinations of satellites (i that record is rejected _k, j _k) and calculate position and the clock correction x of receiver by least square method _k, forward step (5) to;

(4), judge whether the non-fault number of satellite of current gained reaches 4, if reach 4 rough position and clock correction that calculate receiver

and identify fault satellites from residue (n-4) satellite, and again resolve receiver location and clock correction, detect and finish; Otherwise, enter step (5);

(5), make i=i+1, if i > n, non-fault number of satellite is less than 4, now, from reject combinations of satellites with pseudorange residual error statistic λ _kminimum value λ _min=min (λ ₁, λ ₂..., λ _m) corresponding combinations of satellites is as fault satellites, and calculate the receiver location corresponding with this combinations of satellites and clock correction as positioning result, detect and finish; Otherwise return to step (1).

2. the double star fault recognition method based on parity vector method according to claim 1, is characterized in that fault detect judgement amount

obtained by following formula:

${\widehat{\mathrm{\&sigma;}}}_{n-1}=\sqrt{\frac{\mathrm{SSE}}{(n-1)-4}}$

Wherein, SSE is pseudorange residual sum of squares (RSS);

Fault detect thresholding σ _{t, (n-1)}obtained by following formula:

$\left\{\begin{array}{c}{\mathrm{\&sigma;}}_{T,(n-1)}=\frac{{\mathrm{\&sigma;}}_0{T}_{n-1}}{\sqrt{(n-1)-4}}\\ T_{n-1}=\mathrm{icdf}\left({\mathrm{\&chi;}}^2(1-\mathrm{\&alpha;},(n-1)-4)\right)\end{array}\right.$

Wherein, icdf is (1-α) fractile that computer card side distributes;

Fault Identification judgement amount r _iobtained by following formula:

$r_i=\frac{\left|\right|p{Q}_{p.i}\left|\right|}{\left|\right|Q_{p.i}\left|\right|}.$

3. the double star fault recognition method based on parity vector method according to claim 1, is characterized in that, detects judgement amount

obtained by following formula:

${\widehat{\mathrm{\&sigma;}}}_{n-2}=\sqrt{\frac{\mathrm{SSE}}{(n-2)-4}}$

Wherein, SSE is pseudorange residual sum of squares (RSS);

Fault detect thresholding σ _{t, (n-2)}obtained by following formula:

$\left\{\begin{array}{c}{\mathrm{\&sigma;}}_{T,(n-2)}=\frac{{\mathrm{\&sigma;}}_0{T}_{n-2}}{\sqrt{(n-2)-4}}\\ T_{n-2}=\mathrm{icdf}\left({\mathrm{\&chi;}}^2(1-\mathrm{\&alpha;},(n-2)-4)\right)\end{array}\right.$

Wherein, icdf is (1-α) fractile that computer card side distributes.

4. the double star fault recognition method based on parity vector method according to claim 1, is characterized in that rough position and clock correction

obtained by following formula:

$\hat{x}={\left(H_0^T{H}_0\right)}^{-1}{H}_0^T{y}_0$

Wherein, H ₀for 4 observing matrixes that non-fault satellite forms of current gained,

for H ₀transposed matrix, y ₀for the Pseudo-range Observations vector of 4 non-fault satellites of current gained.

5. the double star fault recognition method based on parity vector method according to claim 1, is characterized in that, receiver location and clock correction x _kobtained by following formula:

$x_k={\left(H_k^T{H}_k\right)}^{-1}{H}_k^T{y}_k$ k＝1,2,…,N

Wherein, H _kfor having rejected two satellite (i _k, j _k) after the observing matrix that forms of remaining (n-2) satellite, y _kfor having rejected two satellite (i _k, j _k) after the Pseudo-range Observations vector of remaining (n-2) satellite;

Pseudorange residual error statistic λ _kobtained by following formula:

${\mathrm{\&lambda;}}_k=\underset{\underset{h\&NotEqual;i,j}{h=1}}{\overset{n}{\mathrm{\&Sigma;}}}{({\mathrm{\&rho;}}_h-b_k-\sqrt{{(x_h-x_k)}^2+{(y_h-y_k)}^2+{(z_h-z_k)}^2})}^2.$

Images