CN101357643A - Accurate train positioning method and system realized by digital trail map and GPS - Google Patents

Abstract

The invention provides a method for realizing accurate train positioning by a GPS with the aid of a digital railway map. The train positioning method essentially comprises two subalgorithms, namely a GPS integrity monitoring algorithm and a train positioning calculating algorithm. The GPS integrity monitoring algorithm uses the establishment of a local coordinate system based on railways and transforms a WGS-84 coordinate to a new local coordinate system based on railways as well as uses geographic information provided by the digital railway map to establish a redundancy equation and a constraint condition so as to monitor the pseudo-range observed quantity of the GPS; the train positioning calculating algorithm is responsible for establishing a train position calculating model and a speed calculating model and integrating the two models to realize a train positioning calculating function. The train positioning method, provided by the invention, can not only effectively improve the precision and the stability of the train positioning when the GPS is complete, but also provide a train positioning result with comparatively high precision when the GPS is not complete.

Description

Numeral track map assistant GPS is realized accurate train positioning method and system

Technical field

The invention belongs to the vehicle positioning technology field, particularly a kind of digital track map assistant GPS realization accurate train positioning method and system of utilizing.

Background technology

Along with the quick growth of global economy, railway transportation is being played the part of more and more important role in growth of the national economic.The railroad train transport capacity is big more, and quantity is many more, and speed is fast more, and is just high more to the security requirement of train operation.And the tracking in real time accurately of train position is just become key issue in the train operation control system, follow the tracks of train position real-time and accurately and be preventing the train conflict, knock into the back and the basis of train scheduling and automatic steering.Global position system GPS (Global Positioning System) is being widely used aspect accurate location of train and the guarantee driving safety with its round-the-clock, continuous, real-time high-precision fixed bit function, and GPS train locating method technology is a great research topic in present vehicle positioning technology field.

In traditional train operation control system, train positioning adopts methods such as track circuit, inquiry response device, induction coil and miles counter to realize mostly, and nearly all there are defectives such as the high and accuracy of positioning of equipment cost is low in these orthodox methods.In order to obtain better locating effect on the basis of GPS location, common way is that the output positioning result with other navigation information and GPS carries out fusion application, thus raising train positioning precision, but there is following vice proper in this method:

(1) defective of GPS and INS and DR fusion application.Because INS (Inertial Navigation System, inertial navigation system) need obtain navigation parameter by acceleration/accel is carried out integral operation, thereby determine carrier positions, after working long hours, can produce accumulated error, therefore, INS need increase other auxiliary positioning information again, and this will cause the total system cost higher.DR (Dead Reckoning, the dead reckoning system) be from known coordinate position, according to the parameters such as course, the speed of a ship or plane and duration of voyage of motion carrier at this point, calculate next coordinate position constantly, therefore, DR needs GPS that the initial position of vehicle is provided, in case and temperature changes, perhaps because drift error, the mechanical shock of inertia device, skid, idle running, locking and wheel footpath factor etc. cause output error, these output errors also can build up with the growth of distance and time.

(2) defective of map match.Numerical map is a kind of ectocine that is not subjected to, the navigation information source that stability is higher, and therefore, map match becomes one of important method that improves train Positioning Technology precision and traffic guiding.Yet because the mutual coupling between the accurate positioning result that this location is high precision road data and GPS by numerical map storage is exported improves accuracy of positioning, so it is to dependence height of the autonomous location of GPS; That is to say that it is the prerequisite of map match that GPS can accurately locate.In addition, the error that produces in the autonomous position fixing process of GPS will directly be brought in the map matching process, in case GPS can't export the accurate in locating result, map match will lose efficacy.

Because railway line is difficult for change, train operation track relative fixed can obtain station track, train place section rapidly by interlock system, so obtaining, extracting of track cartographic information is comparatively convenient in the station.Therefore, study the train locating method that a kind of new cost is lower, stability is higher, the track cartographic information is directly used in train positioning resolves and have important practical significance.

The method of utilizing digital track map to improve locating effect in the existing train Positioning Technology mainly is confined to the method for map match.Yet map-matching method also has neither part nor lot in the positioning calculation process but directly depends on the autonomous positioning result of GPS; In case GPS can't export the accurate in locating result, map-matching method will lose efficacy, thereby limit the precision and the stability of train positioning to a great extent.

Summary of the invention

In order to overcome the above-mentioned deficiency of prior art, the present invention utilizes the advantage of digital track map high precision and high stability, the track cartographic information is directly used in during train positioning resolves, auxiliary GPS integrity monitoring algorithm and the train positioning of integrated use numeral track map resolved algorithm, proposed a kind of digital track map assistant GPS and realized accurate train positioning method and system.The technical solution adopted for the present invention to solve the technical problems is:

A kind of digital track map assistant GPS is realized the method for accurate train location, and it comprises the steps:

Step 1 input system data are carried out data extract and processing;

Step 2 realizes the GPS integrity monitoring algorithm that digital track map is auxiliary, it is by setting up the local system of axes based on track, the geography information of utilizing digital track map to provide makes up remaining equation and constraint condition, the GPS that relatively rejects by GPS observed quantity error and decision threshold resolves the bigger observed quantity of process error of mean squares and guarantees accuracy of positioning, and reduces the requirement of traditional GPS completeness monitoring method to the satellites in view number;

The auxiliary train positioning of the digital track map of step 3 realization is resolved algorithm, information such as its comprehensive application track map datum, gps satellite ephemeris and raw pseudo range observed quantity, set up the auxiliary train position of digital track map and resolve model and velocity calculated model, find the solution the position and the speed of train respectively, final by the positioning calculation result under next comprehensive these the two kinds of models of suitable weight allocation, thus realize accurate train positioning;

The data of step 4 output train location estimation.

Above-mentioned steps also comprises following technical characterictic:

Step 1 is described carries out data extract and handles further comprising: extract from observed datas such as the satellite ephemeris of GPS receiver and pseudoranges; And the digital track map data base that comprises station track accurate geographic information.

The described GPS integrity monitoring of step 2 algorithm comprises: one, and set up a kind of local system of axes, and be train position local new system of axes by the WGS-84 coordinate transformation based on track based on track circuit; Its two, monitoring GPS pseudo range observed quantity.

The new system of axes in described this locality is defined as follows: will be along station track, the train place section direction of having discerned of current of traffic as X-axis; To make progress perpendicular to horizontal surface, with the rectangular direction of X-axis as the Z axle; Will be perpendicular to the horizontal direction of track and Z axle to the right as Y-axis; The origin of coordinates is the starting point of this station track section on current of traffic.Because the running orbit one of train is positioned on the track, so under this new system of axes: Y coordinate and Z coordinate can be considered zero; The X coordinate depends on the actual distance that train moves on this station track section.

Station track section head node coordinate along current of traffic under the WGS-84 system of axes of supposing the train place is (x _Head, y _Head, z _Head), the tail node coordinate is (x _End, y _End, z _End).The three-dimensional location coordinates of train (headstock) under the WGS-84 system of axes is that (z), the three-dimensional location coordinates under new local system of axes is (x ', y ', z ') for x, y.Work as x _Head＜x＜x _End, y _Head＜y＜y _End, z _Head＜z＜z _EndThe time, the coordinate of train (headstock) under the new system of axes in this locality can be expressed as

${\left(\begin{array}{c}{x}^{\′}\\ y^{\′}\\ z^{\′}\end{array}\right)}_{\mathrm{new}}={\left(\begin{array}{c}\sqrt{{{(x-x_{\mathrm{head}})}^2+(y-y_{\mathrm{head}})}^2+{(z-z_{\mathrm{head}})}^2}\\ 0\\ 0\end{array}\right)}_{\mathrm{WGS}-84}$

Described monitoring GPS pseudo range observed quantity further comprises: under the WGS-84 system of axes, carry out the first of train position with all visual GPS pseudo range observed quantity and resolve, and will resolve the result and be converted to new local coordinate; And estimated distance residual error.

Under new local system of axes, utilize constraint condition y=z=0, use the value of method of least square backstepping pseudo range observed quantity:

${\left(\begin{array}{c}{x}^{\′}\\ y^{\′}\\ z^{\′}\\ \mathrm{\Δt}\end{array}\right)}_{\mathrm{new}}=R\×{\left(G^{T}G\right)}^{-1}{G}^T\×\left(\begin{array}{c}{l}_{\mathrm{observed}}^1\\ l_{\mathrm{observed}}^2\\ ...\\ l_{\mathrm{observed}}^n\end{array}\right)$

Wherein, G is the observing matrix of GPS least-squares estimation; l _Observed ⁱBe the pseudorange observed value of i satellite, i=1,2 ..., n.

Make y '=z '=0, then can instead push away obtaining l ⁱ, it is and l _Observed ⁱCorresponding pseudorange amount.

$\left(\begin{array}{c}{l}^1\\ l^2\\ ...\\ l^n\end{array}\right)=G^{\′}{\left[{\left(G^{\′}\right)}^T{G}^{\′}\right]}^{-1}{\left(G^{\′}\right)}^T\×{\left(\begin{array}{c}{x}^{\′}\\ 0\\ 0\\ \mathrm{\Δt}\end{array}\right)}_{\mathrm{new}}$

Wherein, G '=R * G.

In described estimated distance residual error process, relatively the pseudo range observed quantity of each satellite and the deviation of the pseudorange correlative that inverse obtains in the previous step are called it residual distance error vector ω here.

$\mathrm{\ω}={[l_{\mathrm{observed}}^1-l^1,l_{\mathrm{observed}}^2-l^2,...,l_{\mathrm{observed}}^i-l^i,...,l_{\mathrm{observed}}^n-l^n]}^T,i=\mathrm{1,2},...,n$

Define among the present invention ${\mathrm{\Δl}}_i=|l_{\mathrm{observed}}^i-l^i|,$ Use Δ l _iWeigh and defend the quality of observed quantity that asterisk is the satellite of i, and as the foundation of differentiating the fault star, with Δ l _iWith detection threshold ξ _ThresholdCompare, have only error when GPS pseudorange observed value at detection threshold ξ _ThresholdIn, carry out GPS with such observed value and resolve to be only and satisfy accuracy requirement.If the error of GPS pseudorange observed value has exceeded detection threshold ξ _Threshold, such observed quantity is with disallowable.

The described train positioning of step 3 is resolved and is meant: utilize the auxiliary location compute that carries out train of digital track cartographic information; Utilize the auxiliary velocity calculated that carries out train of digital track cartographic information; And according to the comprehensive positioning result of weights.Its specific algorithm is as follows:

The auxiliary location compute that carries out train of the digital track cartographic information of described utilization further comprises:

One, set up conventional GPS positioning calculation equation:

Wherein,

Be the transition matrix between the coordinate under the coordinate under the WGS-84 system of axes and the earth's core geodetic coordinate system.Order:

Its two, in GPS positioning calculation equation, add the cartographic information of track section, the GPS observational equation just can be write as:

In the formula, Δ ρ _iBe the pseudorange deviation of i satellite with respect to the original hypothesis position; λ _kFor k moment longitude resolves the result;

For k moment latitude resolves the result; Δ λ _K+1Be the k+1 longitude correction of resolving constantly;

Be the k+1 latitude correction of resolving constantly; Δ h _K+1The height correction that obtains constantly for k+1; Δ t _K+1Be k+1 receiver clock correction constantly;

Its three, least-squares estimation.Its least-squares estimation is separated and can be expressed as:

Wherein,

Like this, under hypothesis train elevation unmodified situation, just can obtain the latitude and longitude coordinates correction and the receiver clock correction of train position.

The auxiliary velocity calculated that carries out train of the digital track cartographic information of described utilization further comprises:

Resolve at GPS and to add the limiting condition of track section geography information to the train running speed direction in the equation, its velocity calculated model can be described as:

In the formula,

It is the pseudorange rate of change of i satellite; v _{East, k}, v _{North, k}Be respectively k and resolve the train east orientation and the north orientation running velocity of acquisition constantly; v _{East, k+1}, v _{North, k+1}Be respectively k+1 and resolve the train east orientation and the north orientation running velocity of acquisition constantly; Be receiver clock correction rate of change; V is a measurement noise; k _Lat/longBe the ratio of unit latitude with the pairing actual length of unit longitude; G _DopplerIt is observing matrix with GPS pseudorange rate of change, receiver speed and the corresponding observational equation of clock correction rate of change.

In the formula,

For the earth at longitude and latitude

The East West radius of curvature of position, For the earth at longitude and latitude

The curvature of meridian radius of position.

Describedly specifically be meant: resolve model and distribute certain weight to come comprehensive train positioning result by giving above-mentioned two, obtain final train position coordinate according to the comprehensive positioning result of weights.

The result who supposes to utilize the location compute model to position acquisition with mean square distance error (DRMS) size be σ _PosUtilize the velocity calculated model position acquisition the result with the mean square distance error size be σ _VelThe weight size that model distributed depends on the size of each self-align sum of errors HDOP value of two kinds of models.Weight w is expressed as:

$w=\frac{1}{{(1+\mathrm{HDOP})}^2}{\left(\frac{{\mathrm{\σ}}_{\mathrm{vel}}}{{\mathrm{\σ}}_{\mathrm{pos}}}\right)}^2$

Wherein, the computation process of HDOP value is as follows:

Suppose δ L _nBe GPS observed quantity l _nError vector,

Be the user coordinates correction

Error vector, G _n ^TBe GPS positioning calculation observing matrix, then

Covariance matrix be:

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\mathrm{cov}\left(\mathrm{\δ}{L}_n\right){\left[{\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\right]}^T$

Every satellite all has range error separately, if they all are independent of each other, and variance is determined value σ ₀ ²Then

$\mathrm{cov}\left(\mathrm{\δ}{L}_n\right)={\mathrm{\σ}}_0^{2}I$

Wherein, I is n * n rank identity matrix.Therefore:

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{G}_n^T{\mathrm{\σ}}_0^{2}I{\left[{\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\right]}^T$

$={\left(G_n^T{G}_n\right)}^{-1}{\mathrm{\σ}}_0^2$

Order

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{\mathrm{\δ}}_0^2$

$=H$

$=\left[\begin{array}{cccc}{h}_{11}& h_{12}& h_{13}& h_{14}\\ h_{21}& h_{22}& h_{23}& h_{24}\\ h_{31}& h_{32}& h_{33}& h_{34}\\ h_{41}& h_{42}& h_{43}& h_{44}\end{array}\right]$

Then the horizontal dilution of precision of GPS location can be used range error σ ₀Be expressed as:

HDOP＝(h ₁₁+h ₂₂) ^1/2σ ₀

Final train positioning is resolved the result and is defined as:

$S_{k+1}=w\·S_{k+1,\mathrm{pos}}+(1-w)\·(S_k+\frac{v_k+v_{k+1}}{2}\mathrm{\Δ}{t}_{k+1/k})$

In the formula, S _{K, pos}Be the location positioning result of the k moment by the least-squares estimation acquisition; v _kBe k speed positioning result constantly; Δ t _K+1/kBe sampling time interval; S _K+1Be the constantly final positioning result of k+1.

A kind of digital track map assistant GPS is realized the accurate train position fixing system, and it comprises:

Data extract and data processing module are used to extract and handle from observed quantity data such as the satellite ephemeris of GPS receiver and pseudoranges, and the data that comprise the digital track map data base input of station track accurate geographic information;

GPS integrity monitoring module, it is responsible for setting up local system of axes and carries out coordinate transformation, and passes through the comparison of GPS measured error and decision threshold, rejects the bigger observed quantity of error, thus monitoring GPS pseudo range observed quantity;

And train positioning is resolved module, and it is responsible for adding track section in described monitoring GPS pseudo range observed quantity cartographic information utilizes method of least square to carry out the location compute of train, and is responsible for finishing the velocity calculated of train.

Annexation between above-mentioned each module is as follows: the quantitative data input after data extract and data processing module processing is responsible for monitoring mould monitoring GPS pseudo range observed quantity to GPS integrity monitoring module by this GPS integrity module; In this monitoring GPS pseudo range observed quantity, add the cartographic information of track section, resolve location compute and the velocity calculated that module is carried out train by train positioning; Export estimated result at last to train position.

Described GPS integrity monitoring module further comprises: data extract and processing module, it is responsible for gathering the almanac data and the pseudo range observed quantity of satellites in view, the cartographic information that in described monitoring GPS pseudo range observed quantity, adds track section, calculate the position coordinate of satellites in view, and the extraction of responsible track cartographic information and processing; Local system of axes is created and coordinate transferring, and it is responsible for setting up a kind of local system of axes based on track circuit, and is the WGS-84 coordinate transformation of train local coordinate based on track; Error judgement and detection threshold determination module, it is responsible for according to the least-squares estimation algorithm, calculation error criterion and decision threshold, monitoring GPS pseudo range observed quantity; And, pseudo range observed quantity screening and rejecting module, it filters out the GPS pseudo range observed quantity that satisfies error condition by the comparison of GPS measured error and decision threshold, rejects the bigger observed quantity of error.

Described train positioning is resolved module and further comprised: locating data is extracted and processing module, is responsible for adding in described monitoring GPS pseudo range observed quantity the cartographic information of track section; Train position resolves module, and it utilizes method of least square to carry out the location compute of train; Train speed resolves module, finishes the velocity calculated of train; And comprehensive module, it resolves model by non-two and distributes certain weight, thus comprehensive train positioning result obtains final train position coordinate.

The invention has the beneficial effects as follows: the supplementary that the present invention makes full use of digital track map to be provided is improved the train positioning effect, avoid the undue dependence of positioning result, and reduced of the requirement of traditional GPS completeness monitoring method the satellites in view number to GPS output result.Use the auxiliary GPS integrity monitoring algorithm of digital track map provided by the invention can be exactly to GPS observed quantity monitor, count at satellite under the situation of abundance, can reject the bigger GPS observed quantity of error effectively, guarantee the precision of GPS positioning calculation.See table (table 1) as can be known, under the GPS condition for completeness, the present invention under equal conditions has higher accuracy of positioning and better position stability than existing GPS localization method; Under the incomplete condition of GPS, the present invention also can be provided in the positioning result in certain available accuracy rating; And be subjected at gps satellite signal under the environment of serious coverage, it can be used as the important method of the accurate location of train and replenishes.

Table 1 train locating method relatively

Description of drawings

Fig. 1 is the train positioning system construction drawing of digital track map assistant GPS;

Fig. 2 is the auxiliary GPS integrity monitoring algorithm flow chart of digital track map;

Fig. 3 is the sub-block diagram of GPS integrity monitoring modular construction of the present invention;

Fig. 4 is the WGS-84 coordinate and conversion scheme drawing based on the local system of axes of track;

Fig. 5 is the auxiliary sub-block diagram of train positioning modular construction of digital track map of the present invention;

The present invention is further described below in conjunction with the drawings and specific embodiments.

The specific embodiment

Embodiment 1: as shown in Figure 1, the core of the train locating method that the digital track map that the present invention studied is auxiliary is a map assist train location algorithm.This map assist train position fixing system can roughly be divided into data extract and processing, GPS integrity monitoring and train positioning are resolved three modules.Though these three modules structurally are independently, all are mutually related in each computation period.This map assist train position fixing system comprises that digital track map auxiliary GPS integrity monitoring algorithm and train positioning resolve two main subalgorithms of algorithm.The input of this system comprises: from observed quantity data such as the satellite ephemeris of GPS receiver and pseudoranges, and the digital track map data base that comprises station track accurate geographic information.System outlet be estimation to train position.

Embodiment 2 is described in further details the present invention by the specific embodiment as shown in Figure 2.In one embodiment, as follows according to the idiographic flow of the auxiliary GPS integrity monitoring algorithm of digital track map of the present invention:

1. gather the almanac data and the pseudo range observed quantity of satellites in view, and calculate the position coordinate of satellites in view.Particularly, utilize NovAtel GPS receiver to gather almanac data, the pseudo range observed quantity data of satellites in view, and extract the position coordinate of the correlation parameter calculating satellites in view in the satellite ephemeris;

2. the extraction of track cartographic information and processing.In one embodiment, obtain train track occupied information, from the track map data base, extract the cartographic information of corresponding station track, and these cartographic informations are converted to form required in the algorithm proposed by the invention by the station interlocking system; Last feasibility in conjunction with concrete data verification algorithm;

3. set up a kind of local system of axes, and be the WGS-84 coordinate transformation of train local coordinate based on track based on track circuit;

4. according to the least-squares estimation algorithm, calculation error criterion and decision threshold are monitored the GPS pseudo range observed quantity;

5. carry out error judgment according to selected threshold value, reject the observed data that does not conform to accuracy requirement, determine to satisfy the observed quantity of error requirements, so that further carry out the GPS positioning calculation.

Embodiment 3 as shown in Figure 3, GPS integrity monitoring algorithm of the present invention can be summed up and is described as following 4 main submodules: data extract and processing module, data extract and data processing module are used to extract and handle from observed quantity data such as the satellite ephemeris of GPS receiver and pseudoranges, and the data that comprise the digital track map data base input of track accurate geographic information; , and calculate the position coordinate of satellites in view and the extraction and the processing of responsible track cartographic information; Quantitative data input GPS integrity monitoring module after treatment is responsible for setting up a kind of local system of axes based on track circuit by this GPS monitoring modular, and is train position local new system of axes based on track by the WGS-84 coordinate transformation; Error judgement and detection threshold determination module, it is responsible for according to the least-squares estimation algorithm, calculation error criterion and decision threshold, monitoring GPS pseudo range observed quantity; And, monitoring GPS pseudo range observed quantity module, it is responsible for screening and rejects pseudo range observed quantity.

Embodiment 4 as shown in Figure 4, in one embodiment, local new system of axes is defined as follows: station track, the train place section direction of having discerned along current of traffic is an X-axis; Make progress perpendicular to horizontal surface, with the rectangular direction of X-axis be the Z axle; Horizontal direction perpendicular to track and Z axle is to the right a Y-axis; The origin of coordinates is the starting point of this station track section on current of traffic.Because the running orbit one of train is positioned on the track, so under this new system of axes: Y coordinate and Z coordinate can be considered zero; The X coordinate depends on the actual distance that train moves on this station track section.

Station track section head node coordinate along current of traffic under the WGS-84 system of axes of supposing the train place is (x _Head, y _Head, z _Head), the tail node coordinate is (x _End, y _End, z _End).The three-dimensional location coordinates of train (is position reference with the headstock) under the WGS-84 system of axes is for (z), the three-dimensional location coordinates under new local system of axes is (x ', y ', z ') for x, y.Work as x _Head＜x＜x _End, y _Head＜y＜y _End, z _Head＜z＜z _EndThe time, the coordinate of train under the new system of axes in this locality can be expressed as

${\left(\begin{array}{c}{x}^{\′}\\ y^{\′}\\ z^{\′}\end{array}\right)}_{\mathrm{new}}={\left(\begin{array}{c}\sqrt{{(x-x_{\mathrm{head}})}^2+{(y-y_{\mathrm{head}})}^2+{(z-z_{\mathrm{head}})}^2}\\ 0\\ 0\end{array}\right)}_{\mathrm{WGS}-84}$

Embodiment 5 as shown in Figure 2, utilizes method of least square estimation error criterion and decision threshold in one embodiment, monitoring GPS pseudo range observed quantity.Be implemented as follows:

Under the WGS-84 system of axes, carry out the first of train position with all visual GPS pseudo range measurement amounts and resolve, and will resolve the result and be converted to new local coordinate.Under new local system of axes, utilize constraint condition y=z=0, with method of least square backstepping pseudorange.

${\left(\begin{array}{c}{x}^{\′}\\ y^{\′}\\ z^{\′}\\ \mathrm{\Δt}\end{array}\right)}_{\mathrm{new}}=R\×{\left(G^{T}G\right)}^{-1}{G}^T\×\left(\begin{array}{c}{l}_{\mathrm{observed}}^1\\ l_{\mathrm{observed}}^2\\ ...\\ l_{\mathrm{observed}}^n\end{array}\right)$

Wherein, G is the observing matrix of GPS least-squares estimation; l _Observed ⁱBe the pseudorange observed value of i satellite, i=1,2 ..., n.

Make y '=z '=0, then can instead push away obtaining l ⁱ, it is and l _Observed ⁱCorresponding pseudorange amount.

$\left(\begin{array}{c}{l}^1\\ l^2\\ ...\\ l^n\end{array}\right)=G^{\′}{\left[{\left(G^{\′}\right)}^T{G}^{\′}\right]}^{-1}{\left(G^{\′}\right)}^T\×{\left(\begin{array}{c}{x}^{\′}\\ 0\\ 0\\ \mathrm{\Δt}\end{array}\right)}_{\mathrm{new}}$

Wherein, G '=R * G.

Relatively the pseudo range observed quantity of each satellite and the deviation of the pseudorange correlative that inverse obtains in the previous step are called it residual distance error vector ω here.

$\mathrm{\ω}={[l_{\mathrm{observed}}^1-l^1,l_{\mathrm{observed}}^2-l^2,...,l_{\mathrm{observed}}^i-l^i,...,l_{\mathrm{observed}}^m-l^n]}^T,i=\mathrm{1,2},...,n$

Define among the present invention ${\mathrm{\Δl}}_i=|l_{\mathrm{observed}}^i-l^i|,$ Use Δ l _iWeigh and defend the quality of observed quantity that asterisk is the satellite of i, and as the foundation of differentiating the fault star, with Δ l _iWith detection threshold ξ _ThresholdCompare, have only error when GPS pseudorange observed value at detection threshold ξ _ThresholdIn, carry out GPS with such observed value and resolve to be only and satisfy accuracy requirement.If the error of GPS pseudorange observed value has exceeded detection threshold ξ _Threshold, such observed quantity is with disallowable.

Embodiment 6

The train positioning modular construction sub-block diagram auxiliary as shown in Figure 5 according to digital track map of the present invention, determine GPS positioning calculation equation according to gps satellite ephemeris and GPS observed data, train positioning is resolved the information that module receives data extract and data processing module and GPS integrity monitoring module, in described monitoring GPS pseudo range observed quantity, add the cartographic information of track section, and utilize method of least square to carry out the location compute of train; Train positioning is resolved the estimation of module output to train position.The specific embodiment process is as follows:

In the GPS of routine positioning calculation equation,

Wherein,

Be the transition matrix between the coordinate under the seat under the WGS-84 system of axes and the earth's core geodetic coordinate system.Order:

Add the cartographic information of track section in GPS positioning calculation equation, the GPS observational equation just can be write as:

In the formula, Δ ρ _iBe the pseudorange deviation of i satellite with respect to the original hypothesis position; λ _kFor k moment longitude resolves the result;

For k moment latitude resolves the result; Δ λ _K+1Be the k+1 longitude correction of resolving constantly; Be the k+1 latitude correction of resolving constantly; Δ h _K+1The height correction that obtains constantly for k+1; Δ t _K+1Be k+1 receiver clock correction constantly.

Like this, its least-squares estimation is separated and can be expressed as:

Wherein,

Like this, under hypothesis train elevation unmodified situation, just can obtain the latitude and longitude coordinates correction and the receiver clock correction of train position.So-called receiver clock correction, when referring to the clock face of receiver and the difference between during the GPS standard, the high-precision quartz chronometer of the general employing of GPS receiver generally is used as an independently unknown number to each observation receiver clock correction constantly, finds the solution in the lump with the location parameter of receiver.

Embodiment 7 finishes the velocity calculated of train referring to accompanying drawing 5 according to the present invention.Resolve at GPS and to add the limiting condition of track section geography information to the train running speed direction in the equation, its velocity calculated model can be described as:

In the formula,

It is the pseudorange rate of change of i satellite; v _{East, k}, v _{North, k}Be respectively k and resolve the train east orientation and the north orientation running velocity of acquisition constantly; v _{East, k+1}, v _{North, k+1}Be respectively k+1 and resolve the train east orientation and the north orientation running velocity of acquisition constantly; Be receiver clock correction rate of change; V is a measurement noise; k _Lat/longBe the ratio of unit latitude with the pairing actual length of unit longitude; G _DopplerIt is observing matrix with GPS pseudorange rate of change, receiver speed and the corresponding observational equation of clock correction rate of change.

In the formula,

For the earth at longitude and latitude The East West radius of curvature of position,

For the earth at longitude and latitude

The curvature of meridian radius of position.

Embodiment 8 give two of the foregoing description 6 and embodiment 7 to resolve the certain weight of model distribution, thereby comprehensive train positioning result obtains final train position coordinate, referring to accompanying drawing 5 according to algorithm model provided by the invention.

The result who supposes to utilize the location compute model to position acquisition with mean square distance error (DRMS) size be σ _PosUtilize the velocity calculated model position acquisition the result with the mean square distance error size be σ _VelThe weight size that model distributed depends on the size of each self-align sum of errors HDOP value of two kinds of models.Weight w is expressed as:

$w=\frac{1}{{(1+\mathrm{HDOP})}^2}{\left(\frac{{\mathrm{\σ}}_{\mathrm{vel}}}{{\mathrm{\σ}}_{\mathrm{pos}}}\right)}^2$

Wherein, the computation process of HDOP value is as follows:

Suppose δ L _nBe GPS observed quantity L _nError vector,

Be the user coordinates correction

Error vector, G _n ^TBe GPS positioning calculation observing matrix, then

Covariance matrix be:

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\mathrm{cov}\left(\mathrm{\δ}{L}_n\right){\left[{\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\right]}^T$

Every satellite all has range error separately, if they all are independent of each other, and variance is determined value σ ₀ ²Then

$\mathrm{cov}\left(\mathrm{\δ}{L}_n\right)={\mathrm{\σ}}_0^{2}I$

Wherein, I is n * n rank identity matrix.Therefore:

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{G}_n^T{\mathrm{\σ}}_0^{2}I{\left[{\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\right]}^T$

$={\left(G_n^T{G}_n\right)}^{-1}{\mathrm{\σ}}_0^2$

Order

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{\mathrm{\σ}}_0^2$

$=H$

$=\left[\begin{array}{cccc}{h}_{11}& h_{12}& h_{13}& h_{14}\\ h_{21}& h_{22}& h_{23}& h_{24}\\ h_{31}& h_{32}& h_{33}& h_{34}\\ h_{41}& h_{42}& h_{43}& h_{44}\end{array}\right]$

Then the horizontal dilution of precision of GPS location can be used range error σ ₀Be expressed as:

HDOP＝(h ₁₁+h ₂₂) ^1/2σ ₀

Final train positioning is resolved the result and is defined as:

$S_{k+1}=w\·S_{k+1,\mathrm{pos}}+(1-w)\·(S_k+\frac{v_k+v_{k+1}}{2}{\mathrm{\Δt}}_{k+1/k})$

In the formula, S _{K, pos}Be the location positioning result of the k moment by the least-squares estimation acquisition; v _kBe k speed positioning result constantly; Δ t _K+1/kBe sampling time interval; S _K+1Be the constantly final positioning result of k+1.

Above-mentioned only is preferred embodiment of the present invention, is not used for limiting practical range of the present invention.That is to say that any equal variation and modification of being made according to claim scope of the present invention is all claim scope of the present invention and contains.

Claims (10)

1. a digital track map assistant GPS is realized the method for accurate train location, it is characterized in that, comprises the steps:

Step 1, input system data are carried out data extract and processing, and it further comprises: extract from observed datas such as the satellite ephemeris of GPS receiver and pseudoranges; And, comprise the digital track map data base of station track accurate geographic information;

Step 2, the auxiliary GPS integrity monitoring algorithm of the digital track map of realization, described GPS integrity monitoring algorithm is meant: one, set up a kind of local system of axes based on track circuit, be the WGS-84 coordinate transformation based on the new system of axes in this locality of track; Its two, monitoring GPS pseudo range observed quantity;

Step 3, the auxiliary train positioning of the digital track map of realization are resolved algorithm, and it comprises: utilize the auxiliary location compute that carries out train of digital track cartographic information; Utilize the auxiliary velocity calculated that carries out train of digital track cartographic information; And according to the comprehensive positioning result of weights;

The data of step 4, output train location estimation.

2. a kind of digital track map assistant GPS according to claim 1 is realized the method for accurate train location, its feature also is, in step 2, the new system of axes in described this locality is defined as follows: will be along station track, the train place section direction of having discerned of current of traffic as X-axis; To make progress perpendicular to horizontal surface, with the rectangular direction of X-axis as the Z axle; Will be perpendicular to the horizontal direction of track and Z axle to the right as Y-axis; The origin of coordinates is the starting point of this station track section on current of traffic; Because the running orbit one of train is positioned on the track, so under this new system of axes: Y coordinate and Z coordinate can be considered zero; The X coordinate depends on the actual distance that train moves on this station track section.

3. a kind of digital track map assistant GPS according to claim 1 is realized the method for accurate train location, and its feature also is, in the step 2, if the station track section at train place head node coordinate along current of traffic under the WGS-84 system of axes is (x _Heady _Head, z _Head), the tail node coordinate is (x _End, y _End, z _End); The three-dimensional location coordinates of train head under the WGS-84 system of axes is that (z), the three-dimensional location coordinates under new local system of axes is (x ', y ', z ') for x, y; Work as x _Head＜x＜x _End, y _Head＜y＜y _End, z _Head＜z＜z _EndThe time, the coordinate of train head under the new system of axes in this locality can be expressed as:

${\left(\begin{array}{c}{x}^{\′}\\ y^{\′}\\ z^{\′}\end{array}\right)}_{\mathrm{new}}={\left(\begin{array}{c}\sqrt{{(x-x_{\mathrm{head}})}^2+{(y-y_{\mathrm{head}})}^2+{(z-z_{\mathrm{head}})}^2}\\ 0\\ 0\end{array}\right)}_{\mathrm{WGS}-84};$

Described monitoring GPS pseudo range observed quantity further comprises: under the WGS-84 system of axes, carry out the first of train position with all visual GPS pseudo range measurement amounts and resolve, and will resolve the result and be converted to new local coordinate; And, the estimated distance residual error.

4. realize the method for accurate train location according to claim 1 or 3 described a kind of digital track map assistant GPSs, its feature also is: utilize constraint condition y=z=0 under new local system of axes, with method of least square backstepping pseudorange; Described new local coordinate can be calculated by following matrix equation formula:

${\left(\begin{array}{c}{x}^{\′}\\ y^{\′}\\ z^{\′}\\ \mathrm{\Δt}\end{array}\right)}_{\mathrm{new}}=R\×{\left(G^{T}G\right)}^{-1}{G}^T\×\left(\begin{array}{c}{l}_{\mathrm{observed}}^1\\ l_{\mathrm{observed}}^2\\ \·\·\·\\ l_{\mathrm{observed}}^n\end{array}\right);$

Wherein, G is the observing matrix of GPS least-squares estimation; l _Observed ⁱBe the pseudorange observed value of i satellite, i=1,2 ..., n; And,

Make y '=z '=0, then can instead push away obtaining l ⁱ, it is and l _Observed ⁱCorresponding pseudorange amount:

$\left(\begin{array}{c}{l}^1\\ l^2\\ \·\·\·\\ l^n\end{array}\right)=G^{\′}{\left[{\left(G^{\′}\right)}^T{G}^{\′}\right]}^{-1}{\left(G^{\′}\right)}^T\×{\left(\begin{array}{c}{x}^{\′}\\ 0\\ 0\\ \mathrm{\Δt}\end{array}\right)}_{\mathrm{new}};$

Wherein, G '=R * G.

5. realize the method for accurate train location according to claim 1 or 3 described a kind of digital track map assistant GPSs, its feature also is: in described estimated distance residual error process, compare the pseudo range observed quantity of each satellite in the previous step and the deviation of the pseudorange correlative that inverse obtains, obtain residual distance error vector ω, its evaluation method is as follows:

$\mathrm{\ω}={[l_{\mathrm{observed}}^1-l^1,l_{\mathrm{observed}}^2-l^2,...,l_{\mathrm{observed}}^i-l^i,...,l_{\mathrm{observed}}^n-l^n]}^T,i=\mathrm{1,2},...,n$

Definition $\mathrm{\Δ}{l}_i=|l_{\mathrm{observed}}^i-l^i|,$ Use Δ l _iWeigh and defend the quality of observed quantity that asterisk is the satellite of i, and as the foundation of differentiating the fault star, with Δ l _iWith detection threshold ξ _ThresholdCompare, have only error when GPS pseudorange observed value at detection threshold ξ _ThresholdIn, carry out GPS with such observed value and resolve to be only and satisfy accuracy requirement; If the error of GPS pseudorange observed value has exceeded detection threshold ξ _Threshold, such observed quantity is with disallowable.

6. a kind of digital track map assistant GPS according to claim 1 is realized the method for accurate train location, and its feature is that also in step 3, the auxiliary location compute that carries out train of the digital track cartographic information of described utilization further comprises:

One, set up conventional GPS positioning calculation equation:

Wherein,

Be the transition matrix between the coordinate under the coordinate under the WGS-84 system of axes and the earth's core geodetic coordinate system;

Its two, in GPS positioning calculation equation, add the cartographic information of track section and order:

Then the GPS observational equation just can be write as:

In the formula, Δ ρ _iBe the pseudorange deviation of i satellite with respect to the original hypothesis position; λ _kFor k moment longitude resolves the result;

For k moment latitude resolves the result; Δ λ _K+1Be the k+1 longitude correction of resolving constantly;

Be the k+1 latitude correction of resolving constantly; Δ h _K+1The height correction that obtains constantly for k+1; Δ t _K+1Be k+1 receiver clock correction constantly;

Its three, least-squares estimation, its least-squares estimation is separated and can be expressed as:

Wherein,

Like this, under hypothesis train elevation unmodified situation, just can obtain the latitude and longitude coordinates correction and the receiver clock correction of train position.

7. a kind of digital track map assistant GPS according to claim 1 is realized the method for accurate train location, and its feature is that also the auxiliary velocity calculated that carries out train of the digital track cartographic information of described utilization further comprises:

Resolve at GPS and to add the limiting condition of track section geography information to the train running speed direction in the equation, its velocity calculated model can be described as:

In the formula,

It is the pseudorange rate of change of i satellite; v _{East, k}, v _{North, k}Be respectively k and resolve the train east orientation and the north orientation running velocity of acquisition constantly; v _{East, k+1}, v _{North, k+1}Be respectively k+1 and resolve the train east orientation and the north orientation running velocity of acquisition constantly;

Be receiver clock correction rate of change; V is a measurement noise; k _Lat/longBe the ratio of unit latitude with the pairing actual length of unit longitude; G _DopplerBe and the observing matrix of GPS pseudorange rate of change, receiver speed and the corresponding observational equation of clock correction rate of change, its estimation equation is as follows:

In the formula,

For the earth at longitude and latitude

The East West radius of curvature of position, For the earth at longitude and latitude

The curvature of meridian radius of position.

8. a kind of digital track map assistant GPS according to claim 1 is realized the method for accurate train location, its feature also is, describedly specifically be meant: resolve model and distribute certain weight to come comprehensive train positioning result by giving above-mentioned two according to the comprehensive positioning result of weights, obtain final train position coordinate, the certain weight concrete grammar of described distribution is as follows:

The result who supposes to utilize the location compute model to position acquisition with mean square distance error (DRMS) size be σ _PosUtilize the velocity calculated model position acquisition the result with the mean square distance error size be σ _VelThe weight size that model distributed depends on the size of each self-align sum of errors HDOP value of two kinds of models, and weight w is expressed as:

$w=\frac{1}{{(1+\mathrm{HDOP})}^2}{\left(\frac{{\mathrm{\σ}}_{\mathrm{vel}}}{{\mathrm{\σ}}_{\mathrm{pos}}}\right)}^2;$

Wherein, the computation process of HDOP value is as follows:

Suppose δ L _nBe GPS observed quantity L _nError vector,

Be the user coordinates correction Error vector, G _n ^TBe GPS positioning calculation observing matrix, then

Covariance matrix be:

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\mathrm{cov}\left(\mathrm{\δ}{L}_n\right){\left[{\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\right]}^T;$

Every satellite all has range error separately, if they all are independent of each other, and variance is determined value σ ₀ ²Then

$\mathrm{cov}\left(\mathrm{\δ}{L}_n\right)={\mathrm{\σ}}_0^{2}I;$

Wherein, I is n * n rank identity matrix; Therefore:

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{G}_n^T{\mathrm{\σ}}_0^{2}I{\left[{\left(G_n^T{G}_n\right)}^{-1}{G}_n^T\right]}^T;$

$={\left(G_n^T{G}_n\right)}^{-1}{\mathrm{\σ}}_0^2$

Order

$\mathrm{cov}\left(\mathrm{\δ}{\hat{S}}_n\right)={\left(G_n^T{G}_n\right)}^{-1}{\mathrm{\σ}}_0^2$

$=H;$

$=\left[\begin{array}{cccc}{h}_{11}& h_{12}& h_{13}& h_{14}\\ h_{21}& h_{22}& h_{23}& h_{24}\\ h_{31}& h_{32}& h_{33}& h_{34}\\ h_{41}& h_{42}& h_{43}& h_{44}\end{array}\right]$

Then the horizontal dilution of precision of GPS location can be used range error σ ₀Be expressed as:

HDOP＝(h ₁₁+h ₂₂) ^1/2σ ₀；

Final train positioning is resolved the result and is defined as:

$S_{k+1}=w\·S_{k+1,\mathrm{pos}}+(1-w)\·(S_k+\frac{v_k+v_{k+1}}{2}{\mathrm{\Δt}}_{k+1/k});$

In the formula, S _{K, pos}Be the location positioning result of the k moment by the least-squares estimation acquisition; v _kBe k speed positioning result constantly; Δ t _K+1/kBe sampling time interval; S _K+1Be the constantly final positioning result of k+1.

9. a digital track map assistant GPS is realized the accurate train position fixing system, and it comprises:

Data extract and data processing module are used to extract and handle from observed quantity data such as the satellite ephemeris of GPS receiver and pseudoranges, and the data that comprise the digital track map data base input of station track accurate geographic information;

GPS integrity monitoring module is set up local system of axes and is carried out coordinate transformation, and passes through the comparison of GPS measured error and decision threshold, rejects the bigger observed quantity of error, thus monitoring GPS pseudo range observed quantity;

And train positioning is resolved module, and it is responsible for adding track section in described monitoring GPS pseudo range observed quantity cartographic information utilizes method of least square to carry out the location compute of train, and is responsible for finishing the velocity calculated of train;

Annexation between each module is as follows:

Quantitative data input after data extract and data processing module processing is responsible for monitoring mould monitoring GPS pseudo range observed quantity to GPS integrity monitoring module by this GPS integrity module; In this monitoring GPS pseudo range observed quantity, add the cartographic information of track section, resolve location compute and the velocity calculated that module is carried out train by train positioning; Export estimated result at last to train position.

10. a kind of digital track map assistant GPS according to claim 9 is realized the accurate train position fixing system, its feature also is, described GPS integrity monitoring module further comprises: data extract and processing module, it is responsible for gathering the almanac data and the pseudo range observed quantity of satellites in view, the cartographic information that in described monitoring GPS pseudo range observed quantity, adds track section, calculate the position coordinate of satellites in view, and the extraction of responsible track cartographic information and processing; Local system of axes is created and coordinate transferring, and it is responsible for setting up a kind of local system of axes based on track circuit, and is the WGS-84 coordinate transformation of train local coordinate based on track; Error judgement and detection threshold determination module, it is responsible for according to the least-squares estimation algorithm, calculation error criterion and decision threshold, monitoring GPS pseudo range observed quantity; And, pseudo range observed quantity screening and rejecting module, it filters out the GPS pseudo range observed quantity that satisfies error condition by the comparison of GPS measured error and decision threshold, rejects the bigger observed quantity of error;

Described train positioning is resolved module and further comprised: locating data is extracted and processing module, is responsible for adding in described monitoring GPS pseudo range observed quantity the cartographic information of track section; Train position resolves module, and it utilizes method of least square to carry out the location compute of train; Train speed resolves module, finishes the velocity calculated of train; And comprehensive module, it resolves model by non-two and distributes certain weight, thus comprehensive train positioning result obtains final train position coordinate.

Images