JP6414279B2 - GNSS positioning device - Google Patents

Description

  The present invention relates to a GNSS positioning apparatus that receives a ranging signal transmitted from a GNSS (Global Navigation Satellite System) satellite such as a GPS (Global Positioning System) satellite and performs position measurement. And a GNSS positioning method.

  Conventionally, a GNSS positioning method in which a receiver on the earth or in space receives a ranging signal from a plurality of GNSS satellites, and a GNSS positioning device connected to the receiver measures the position of the receiver is widely known. Yes. Observed values obtained from the ranging signal of the GNSS satellite include a pseudorange PR (Pseudo Range) and a carrier phase CP (Carrier Phase).

  The pseudo-range PR is an observation value of a positioning code string (C / A code, P2 code, etc.) carried on a carrier wave of a signal, a propagation time until a signal transmitted from a satellite reaches a receiver, This is the distance between the satellite and the receiver calculated by multiplying the speed of light by calculating from the phase difference between the values of the same positioning code string generated by synchronizing the clock of the receiver and the clock of the satellite. The distance obtained by the pseudo distance PR uses a value in which the clock of the receiver and the clock of the satellite are not completely synchronized, and the obtained distance includes an error due to a deviation of the clock. Called the “pseudo” distance.

  The carrier phase CP is a phase difference between the phase of the carrier tracked when the positioning code string is observed and the same carrier signal generated by synchronizing the clock of the receiver and the clock of the satellite. The phase difference has an ambiguity of 2π × N (N is an integer value of wave number), and the distance between the satellite and the receiver cannot be determined uniquely from the carrier phase alone, but the carrier phase Since the observation resolution is higher than the pseudorange and the noise is low, if the wave number N can be determined correctly, it can be used as a highly accurate distance observation amount.

  However, in GNSS positioning, it is proportional to the error caused by the satellite itself, such as the satellite clock error and the satellite position error obtained from the navigation calendar, and the electron density in the atmosphere when radio waves pass through the ionosphere. Therefore, errors due to atmospheric conditions such as errors due to slow radio wave speed (hereinafter referred to as “ionospheric delay”) and errors due to receivers such as multipaths occur. For this reason, the error in the distance observation value between the satellite and the receiver becomes large, and as a result, the accuracy of the position measurement (positioning) of the receiver is about 10 m.

  Therefore, in order to make GNSS positioning available to applications that require more precise location information, such as surveying and mapping, currently, accurate coordinates on the earth such as known electronic reference points Using observation data, the error due to the satellite itself and the error due to atmospheric conditions are estimated for each GNSS satellite, and the error correction amount (hereinafter referred to as “correction amount”) is provided to the GNSS positioning device as reinforcement information. A method for improving the positioning accuracy by correcting the error of the observation value obtained from the ranging signal of the GNSS satellite has been proposed. A system that realizes this is called a positioning reinforcement system, and many forms have been proposed and implemented.

  Such a GNSS positioning reinforcement system is roughly classified into a VRS (Virtual Reference Station) method and an FKP (Flachen Korrektur Parameter) method. In these methods, the service center generates reinforcement information effective for the entire area from observation data of a plurality of electronic reference points in the service area on the earth. The GNSS positioning device connected to the receiver in the area sends the approximate position information of the receiver to the service center using a mobile phone line or the like, and the reinforcement information at the position from the service center using the same line. Acquired and corrects errors using such reinforcement information, and performs highly accurate positioning. Since the amount of error to be corrected changes every moment, it is necessary to provide reinforcement information at a period corresponding to the required accuracy. In services using mobile phone lines, etc., the capacity of data transmission is large, and it is possible to provide reinforcement information at a sufficiently fast cycle (for example, every second) so that accuracy degradation due to time fluctuations can be ignored. .

  On the other hand, in the positioning reinforcement system, for example, when reinforcing information that can be used by a GNSS positioning device for a wide range such as the whole country in Japan is distributed uniformly by a quasi-zenith satellite, SBAS (Satellite-Based Augmentation System), As described above, the line capacity for data transmission is limited as compared with the case where the reinforcement information is distributed to each individual user using a line such as a mobile phone on the ground. For this reason, a technique has been proposed in which the reinforcement information is compressed (data amount is reduced) and distributed. As an example of compression, the amount of correction is temporally thinned, and the amount of correction that varies depending on the location of the receiver, such as a method that distributes at regular intervals and delay due to radio waves in the ionosphere, is spatially thinned, and is a representative point at equal intervals A method of distributing the value in is included.

  Therefore, when the compressed reinforcement information is used, the correction amount thinned out temporally and spatially is used, so that the correction amount itself has an error compared to the case where the reinforcement information before compression is used. That is, if the period of the correction amount obtained is long with respect to the temporal variation of the error amount of the GNSS observation value, an error in the correction amount (error due to temporal compression) that cannot be ignored occurs. Further, with respect to the spatial variation in the error amount of the GNSS observation value, if the interval between the representative points from which the correction amount is obtained is large, an error in the correction amount that cannot be ignored (error due to spatial compression) occurs. In a time zone or place where the local fluctuation of the error amount is large, even if the correction amount is interpolated and extrapolated temporally and spatially by a function such as a polynomial, the function cannot express the fine fluctuation and cannot be ignored. Error occurs.

  Therefore, by providing information on the correction amount error to the GNSS positioning device, or by devising an interpolation / extrapolation method using the thinned correction amount on the GNSS positioning device side, the influence of the correction amount error is reduced. There is a need for technology to mitigate.

  As a related art in this regard, there is a positioning method that performs processing by treating the interpolated ionospheric delay correction amount as an indeterminate value when interpolating and using the ionospheric delay correction amount distributed spatially. It has been proposed (for example, Non-Patent Document 1). In each GNSS satellite, the ionospheric delay contained in the pseudorange and carrier phase is the same in magnitude and opposite in sign (pseudorange is delayed and carrier phase is advanced), and the sensitivity varies depending on the carrier frequency of the signal. By using the property that the influence of the uncertainty is different for each signal, the error of the ionospheric delay correction amount is separated as a noise component, and the influence on the positioning result is reduced.

  In addition, the error distribution of the correction amount thinned out at the service center and the expansion coefficient of the error variance over time are provided to the GNSS positioning device, and the GNSS positioning device side corrects the correction amount for each GNSS satellite at the time of positioning. In the positioning calculation, a positioning method has been proposed that performs processing such as determining the use or non-use of these satellites and lowering the weight of corrected observation values for satellites with large error variance values. (For example, Patent Document 1).

  In addition, when accurate coordinates on the earth are obtained by decimating the observed values of GNSS satellites such as known electronic reference points, extrapolation using higher-order polynomials is not possible, and data cannot be obtained. A positioning method has been proposed in which an observation value at a time is generated and a difference from the observation value of the receiver is taken to eliminate an error included in the observation value of the receiver (for example, Patent Document 2). This is because when a correction amount thinned out at certain time intervals is obtained, the value at the time when the correction amount is not obtained is extrapolated from the value at the time when the correction amount is obtained by extrapolation using a high-order polynomial. This corresponds to obtaining and using.

U.S. Patent Publication No. 20110014906A1 (page 2, right column, lines 8 to 34) JP 2000-304843 (page 12, right column 43 rows to page 13, right column 27 rows)

D. Odijk: Weighting Ionological Corrections to Improve Fast GPS Positioning Over Medium Distance, ION GPS 2000 (1114 page 22 to 1116 page 19 line)

  However, in the positioning method as described in Non-Patent Document 1, it is possible to reduce an error due to compression of the ionospheric delay correction amount in which the sign of the value in the pseudorange and the carrier wave phase is different and the sensitivity is different for each signal frequency. However, if the signs of the values included in the pseudorange and the carrier phase are the same, such as satellite clock error, satellite orbit error, tropospheric delay, etc., an error due to compression with the same correction amount regardless of the signal frequency. There is a problem that cannot be reduced.

  Further, the positioning method as in Patent Document 1 has the following problems. In other words, the error variance value of the correction amount increases as time elapses from the time when the correction amount is obtained, and the weight of the corrected GNSS observation value in the positioning calculation of the receiver decreases, and thus the correction gain for the observation residual is corrected. Therefore, there is a problem that the convergence time is delayed in the static survey. Also, in mobile surveying such as automobiles, the error variance of the position of the receiver that expands with movement cannot be quickly reduced, and for example, the success rate of the process for determining the wave number of the carrier phase of the signal of the GNSS satellite There is a problem of lowering.

  Furthermore, the positioning method as in Patent Document 2 has the following problems. In other words, in positioning using the correction amount thinned out in time, when the correction amount at the time of positioning calculation is obtained and used by interpolation / extrapolation calculation, when the correction amount varies greatly over time, If the inflection point is included, the accuracy of the interpolation / extrapolation calculation is lowered, which causes a decrease in positioning accuracy. In addition, when the reinforcement information distribution interval is long due to restrictions on the channel capacity of the satellite that distributes the reinforcement information, and the receiver passes through the tunnel or under the overhead, the GNSS positioning device fails to receive the reinforcement information to be acquired. However, there is a problem that the time interval of the obtained correction amount is increased, and the calculation accuracy of interpolation / extrapolation is greatly reduced.

  The present invention relates to a GNSS positioning apparatus and a GNSS positioning method using correction amounts that are thinned out in time, even when the correction amount has a large time variation / time interval or includes an inflection point. An object of the present invention is to provide a GNSS positioning apparatus and a GNSS positioning method that suppress a decrease in calculation accuracy.

  The GNSS positioning apparatus according to the present invention is a GNSS positioning apparatus that calculates the position of a receiver that receives a ranging signal transmitted from a GNSS satellite, and corrects the observation values of the pseudorange and the carrier phase obtained from the ranging signal. A reinforcement information receiving unit that receives reinforcement information relating to the amount, and a first correction amount for correcting the position of the receiver at a position calculation time for calculating the position of the receiver based on the received reinforcement information A correction amount interpolation extrapolation unit, a receiver position prediction unit that predicts the position of the receiver at the position calculation time based on a state quantity estimation value including the position and speed of the receiver, and the receiver position prediction Calculating the distance from the GNSS satellite at the position calculation time based on the predicted position of the receiver and the satellite position at the position calculation time. A correction amount for calculating a second correction amount for correcting the position of the receiver at the position calculation time based on an observation value including a pseudo distance and a carrier wave phase at the position calculation time calculated from a distance signal An estimation unit; and a position calculation unit that corrects the position of the receiver based on a third correction amount obtained based on the first correction amount and the second correction amount. .

  According to the present invention, the first correction amount of the receiver position error estimated based on the plurality of time data of the reinforcement information, the receiver position based on the satellite position obtained from the GNSS signal receiver and the observation value of the GNSS satellite In order to perform the process of calculating the third correction amount of the receiver position error based on the second correction amount of the receiver position error estimated from the above, positioning in the positioning using the correction amount thinned out in time The correction amount at the time when the calculation is performed can be obtained with high accuracy.

It is a block diagram which shows the structure of the positioning apparatus by Embodiment 1 of this invention. It is a graph which shows the relationship between the 1st correction amount calculated in the correction amount interpolation and extrapolation part of Embodiment 1 of this invention, and time. It is a figure for demonstrating operation | movement of the correction amount estimation part of Embodiment 1 of this invention. It is a graph which shows the relationship between the value of the correction amount calculated in the correction amount averaging part of Embodiment 1 of this invention, and the probability density of a correction amount. It is a figure which shows the flowchart of Embodiment 1 of this invention. It is a block diagram which shows the structure of the positioning apparatus by Embodiment 2 of this invention. It is a figure for demonstrating operation | movement of the ambiguity adjustment part of Embodiment 2 of this invention. It is a figure which shows the flowchart of Embodiment 2 of this invention. It is a figure for demonstrating the operation effect of the positioning apparatus by Embodiment 2 of this invention.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a configuration of a positioning apparatus according to Embodiment 1 for carrying out the present invention. As shown in FIG. 1, the positioning apparatus according to the present embodiment includes a reinforcement information receiving unit 1, a reinforcement information storage unit 2, a correction amount interpolation / extrapolation unit 3, a receiver position prediction unit 4, a GNSS signal reception unit 5, It comprises a satellite position calculation unit 6, a correction amount estimation unit 7, a correction amount averaging unit 8, and a position calculation unit 9. Further, the operation of each part of the present embodiment will be described. Hereinafter, the following terms are used for the correction amount.
PRC: Pseudo Range Correction (pseudo distance correction amount)
CPC: Carrier Phase Correction (carrier phase correction amount)

  The pseudo distance correction amount indicates a pseudo distance error correction amount, and the carrier phase correction amount indicates a carrier phase error correction amount. When there are a plurality of pseudo distances such as C / A code and P2 code, the pseudo distance correction amounts are collectively expressed as PRC. Similarly, when there are a plurality of carrier phase such as L1 wave and L2 wave, their carrier phase correction amounts are collectively expressed as CPC.

  The reinforcement information receiving unit 1 uses information or a correction amount for calculating the correction amount or the correction amount of the signal of the GNSS satellite received from the quasi-zenith satellite or the SBAS by a method recommended or specified by the positioning reinforcement service provider. Receive reinforcement information about.

  The reinforcement information storage unit 2 stores the reinforcement information received by the reinforcement information reception unit 1. The reinforcement information received at a plurality of past times is stored so that information necessary for the calculation method in the correction amount interpolation / extrapolation unit 3 can be obtained.

  The correction amount interpolation / extrapolation unit 3 uses the reinforcement information stored in the reinforcement information storage unit 2 to calculate the first correction amount (pseudo distance correction) for the observed value of the GNSS satellite at time t when the position of the receiver is calculated. The amount prc1, the carrier phase correction amount cpc1) and the standard deviation values σ_prc1 and σ_cpc1 are calculated. FIG. 2 is a graph showing the relationship between the first correction amount (y) calculated by the correction amount interpolation / extrapolation unit 3 and time (t). As shown in FIG. 2, the first correction amount is extrapolated by polynomial expression using, for example, values of correction amounts at the past three times. If the past three times are t1, t2, and t3, and the correction amounts corresponding to the times are y1, y2, and y3, the correction amount y at time t + 1 is expressed by the following equation.

The correction amount at each time of each GNSS satellite includes, for example, satellite clock error Clk [m] (m indicates a unit meter), orbit error Orb [m], tropospheric delay Trop [m], ionospheric delay Ion [m] , And C / A code, P2 code, L1 wave carrier phase, L2 wave carrier phase signal-to-signal bias Bias (C / A, P2, L1, L2) [m] The pseudo distance correction amount and the carrier phase correction amount can be obtained. However, in the following equation, the satellite clock error Clk [m] included in the reinforcement information is used to further correct the correction error after correcting the observed value with the satellite clock error included in the navigation calendar of the GNSS satellite. Assume that.
C1PRC (C / A code pseudorange correction)
= −Clk + Orb + Trop + 120 / 154Ion + C / A_Bias
P2PRC (P2 code pseudo distance correction amount)
= -Clk + Orb + Trop + 154 / 120Ion + P2_Bias
L1CPC (L1 wave carrier phase correction amount)
= -Clk + Orb + Trop-120/154 Ion + L1_Bias
L2CPC (L2 wave carrier phase correction amount)
= -Clk + Orb + Trop-154/120 ION + L2_Bias
In addition, if there is a method that the positioning reinforcement service provider recommends or specifies in the calculation method of the correction amount at each time and the calculation method of interpolation / extrapolation, the calculation is performed according to them.

As the standard deviation values σ_prc1 and σ_cpc1, values set in advance or values obtained by obtaining reinforcement information or information on the time change rate of the correction amount from the outside are used. The calculation of the standard deviation of the correction amount using the information on the time change rate of the correction amount is, for example, for each pseudorange correction amount and carrier phase correction amount of each GNSS satellite,
UDRE: standard deviation of correction amount acquired from reinforcement information or externally cur_time: time t for calculating standard deviation of correction amount,
Ref_time: reinforcement information and the standard deviation of the correction amount acquired from outside UDRE_growh_rate: time change rate of the correction amount acquired from the reinforcement information and outside time_of_validity: scale adjustment acquired from the reinforcement information and outside As a value, Corrected_UDRE obtained by the following equation is used as σ_prc1 and σ_cpc1 at time t.
Corrected_UDRE = {(cur_time-ref_time) / time_of_validity * (UDRE_growth_rate-1) +1} * UDRE
In addition, if there is a method recommended by the service provider or specified in the specification, use it.

Next, the operation of the receiver position prediction unit 4 will be described. When the receiver has a position and a speed as state quantities, the receiver position prediction unit 4 out of the estimated state quantity of the receiver at time t-1 immediately before time t at which the position of the receiver is calculated , Position / velocity x _{t−1 | t−1} , v _{t−1 | t−1} and covariance matrix position / velocity term P ^xy _{t−1 | t−1} Predict the receiver position / velocity x _{t | t−1} and its covariance matrix P ^xy _{t | t−1} at time _t . Q is a process noise matrix.

When the receiver has position, speed, and acceleration as state quantities, among the estimated state quantities of the receiver at time t-1 immediately before time t at which the position of the receiver is calculated, the position, speed, Acceleration estimates x _{t-1 | t−1} , v _{t−1 | t−1} , a _{t−1 | t−1,} and the term P ^xya _{t−1 | t−} regarding the covariance matrix position / velocity / acceleration ₁ , the receiver position / velocity / acceleration x _{t | t−1} , v _{t | t−1} , a _{t | t−1} and its covariance matrix P ^xya _{t | t−} Predict ₁ Q is a process noise matrix.

  The GNSS signal receiving unit 5 receives a ranging signal of the GNSS satellite and outputs a navigation calendar and an observation value of the GNSS satellite. In the present embodiment, the term GNSS signal receiving unit is used as the ranging signal receiving unit.

The satellite position calculation unit 6 uses the orbit information of the GNSS satellites included in the navigation calendar output from the GNSS signal reception unit 5 and uses the orbit information of the GNSS satellites in accordance with the procedure described in the specifications of each GNSS satellite, so that the satellite position X ^sat (t ). In addition to the navigation calendar, a precise calendar provided by IGS (International GNSS Services) or the like can be used if it can calculate the satellite position.

  Subsequently, processing of the correction amount estimation unit 7 will be described. When a carrier phase tracking interruption (hereinafter referred to as “cycle slip”) occurs between time t−1 and time t due to an obstacle blocking the signal from the GNSS satellite, that is, at time t−1. If the difference between the carrier phase value and the carrier phase value at time t is greater than or equal to the threshold value, the tracking of the carrier phase is reset, and the real number estimation value and integer determination value of the wave number up to time t−1 become invalid. In the present embodiment, the correction amount estimation unit 7 performs the following process only when it is determined that no cycle slip has occurred between time t-1 and time t. The correction amount estimation unit 7 determines whether or not a cycle slip has occurred. (When it is determined that a cycle slip has occurred, the standard deviation of the second correction amount is set to infinity so that the weight of the second correction amount in the correction amount averaging unit 8 described later becomes zero.) (A method for making the weight of the second correction amount not to be zero even when the second correction amount is used and making it usable will be described in the second embodiment.)

The correction amount estimation unit 7 outputs the receiver position prediction value x _{t | t−1} at the time t output from the receiver position prediction unit 4 and the term P ^x _{t | t−1} relating to the position of the covariance matrix, the GNSS signal. The GNSS observation values PR and CP (pseudorange and carrier phase) output from the receiver 5, the position X ^sat (t) of each GNSS satellite at the time t output from the satellite position calculator 6 and the wave number (an integer value is determined) Using the integer determined value when the integer value is not determined, and the real number estimated value when the integer value is not determined), the second correction amount (pseudo distance correction amount prc2, carrier phase correction amount cpc2) at time t and its standard deviation value σ_prc2 and σ_cpc2 are calculated. FIG. 3 is a diagram for explaining the operation of the correction amount estimation unit 7 of the present embodiment. As shown in FIG. 3, the geometric distance ρ at time t is obtained by the following equation.

そして、時刻ｔにおける擬似距離、搬送波位相、波数から、次式により各衛星の第２の補正量ｐｒｃ２， ｃｐｃ２を計算する。
・ｐｒｃ２ ＝ ρ−（擬似距離−衛星時計誤差−相対論効果）
・ｃｐｃ２＝ ρ−（搬送波位相−衛星時計誤差−相対論効果）−波長×波数

Then, the second correction amounts prc2 and cpc2 of each satellite are calculated from the pseudorange at time t, the carrier phase, and the wave number by the following equation.
・ Prc2 = ρ− (pseudo distance−satellite clock error−relativistic effect)
Cpc2 = ρ− (carrier phase−satellite clock error−relativistic effect) −wavelength × wave number

As the integer determination value of the wave number, a value output from the position calculation unit 9 described later is used. As the wave number of the satellite for which the integer value of the wave number is not determined, the wave number estimated value Nt−1 | t−1 output from the position calculation unit 9 described later may be used. The satellite clock error of the above equation is calculated from the satellite clock error information included in the navigation calendar of the GNSS satellite according to the procedure described in the specifications of the GNSS satellite, and is different from the satellite clock error included in the reinforcement information. . Similarly, the relativistic effect is calculated according to the procedure described in the specification of the GNSS satellite.
The standard deviation is calculated by the following formula, for example.
A satellite whose wave number has an integer value is
Σ_prc2 = √ (geometric distance error variance value + pseudo distance observation error variance value)
Σ_cpc2 = √ (geometric distance error variance value + carrier phase observation error variance value)

For satellites whose wavenumbers have not been determined,
Σ_prc2 = √ (geometric distance error variance value + pseudo distance observation error variance value)
σ_cpc2 = √ (geometric distance error dispersion value + carrier phase observation error dispersion value + wavelength ² × wave number estimation error dispersion value)
In calculating the error variance value σ ² _p of the geometric distance, the value of the component relating to the position of the covariance matrix is used, and is calculated by the following equation.

The pseudo-range observation error variance is, for example, 0.5 ² [m ² ] and the carrier phase observation error variance is, for example, 0.002 ² [m ² ]. Set a reasonable value for it. The wave number estimation error variance value in the processing of the satellite in which the wave number integer value has not been determined will be described. The estimated error variance value of the wave number corresponds to the wave number of the corresponding satellite in the covariance matrix P _{t−1 | t−1} of the state quantity at time t−1 obtained by the processing of the position calculation unit 9 described later. Use the value of the diagonal component. Alternatively, for example, a covariance matrix Pt | t−1 after time update is obtained by a process equivalent to time-time update performed by the position calculation unit 9 using a Kalman filter, and the diagonal component corresponding to the wave number of the corresponding satellite is obtained. A value may be used.

FIG. 4 is a graph showing the relationship between the correction amount value calculated by the correction amount averaging unit 8 of the present embodiment and the probability density of the correction amount. The correction amount averaging unit 8 uses the first correction amount prc1, cpc1 and its standard deviation σ_prc1, σ_cpc1, the second correction amount prc2, cpc2, and the standard deviation σ_prc2, σ_cpc2 to actually use the third correction amount. prc3, cpc3 and its standard deviation value σ_prc3, σ_cpc3 are calculated. In FIG. 8, the first correction amount is C ₁ , its standard deviation is σ _C1 , the second correction amount is C ₂ , its standard deviation is σ _C2 , its third correction amount is C ₃ , and its standard deviation Are collectively expressed as σ _C3 . In other words, the third correction amount is calculated by weighting the first correction amount and the second correction amount with the respective standard deviation values and averaging them using the following calculation formula.

標準偏差σ＿ｐｒｃ３， σ＿ｃｐｃ３は次式にて計算する。

Standard deviations σ_prc3 and σ_cpc3 are calculated by the following equation.

For each of the pseudo distance correction amount and the carrier phase correction amount, a threshold is set according to the elapsed time from time t-1 to time t. If the difference between the first correction amount and the second correction amount is greater than or equal to the threshold value, the accuracy of either the first correction amount or the second correction amount is suspected. If the difference is greater than or equal to the threshold value, at least one of the following two processes can be selected in accordance with the system reliability requirement.
The weight of the second correction amount is set to zero, and the third correction amount = the first correction amount.
-Stop using the correction amount of the satellite until the next reinforcement information is received.

In general, since the correction amount calculated from the reinforcement information does not include the error correction amount due to the receiver, such as the clock error or multipath of the receiver, the first correction amount calculated by the correction amount interpolation / extrapolation unit 3 is used. The correction amount includes only an error due to the satellite itself and an error due to atmospheric conditions. On the other hand, the second correction amount calculated with the correction amount estimated value includes a correction amount for an error caused by the receiver. In order to eliminate this difference, one of the following processes is performed.
・ If you have an estimate of the error caused by the receiver separately, subtract that value from the second correction amount. ・ The position calculation unit 9 converts the observed value of the GNSS satellite into a single difference between satellites. In this case, the reference satellite is determined, the first correction amount and the second correction amount are converted into the correction amount of the single difference between satellites, and the correction amount is averaged at the level of the single difference between satellites. And

  The position calculation unit 9 performs the position calculation of the receiver using the third correction amount, its standard deviations σ_prc3, σ_cpc3, the GNSS observation value, the receiver state quantity estimation value, and its covariance matrix. Estimated value Xt | t−1 of receiver position and velocity, estimated value Nt | t−1 of the wave number of an undecided satellite, and covariance matrix Pt | t−1 of all state variables, third correction Using the quantity and the GNSS observation value, for example, by the time update and the observation update by the Kalman filter, the receiver position / velocity estimate value Xt | t, the wave number estimate value Nt | t, and the covariance matrix Pt | t at time t calculate. The standard deviation of the third correction amount may be used in addition to the observation noise in the observation update. For satellites for which the integer value of the wave number of the carrier phase has not been determined, an adjustment is made from the estimated value of the wave number and the value of the term corresponding to the wave number of the covariance matrix using LAMBDA (The Last Squares Adjustment Adjustment) or the like. Find numerical candidates. If an integer solution candidate passes a test such as a ratio test or a residual test, the wave number is determined, and the real number estimated value is replaced with an integer value.

  Next, the positioning process of this Embodiment is demonstrated using the flowchart of FIG. First, in step 1 (S1), it is confirmed whether or not new reinforcement information is obtained by the reinforcement information receiving unit 1. When the reinforcement information is obtained, the reinforcement information obtained in step 2 (S2) is stored in the reinforcement information storage unit 2. When reinforcement information is not obtained, it transfers to step 3 (S3). In step 3 (S3), it is determined whether or not to perform positioning at time t based on the presence or absence of the GNSS observation amount in the GNSS signal receiving unit 5 and a predetermined positioning cycle. When positioning is not performed at time t, the process proceeds to time t + 1. When positioning is performed at time t, the correction amount of each GNSS satellite is generated by the processing of steps 4 to 7 (S4 to S7). First, in step 4 (S4), using the reinforcement information stored in the reinforcement information storage unit 2, the correction amount interpolation / extrapolation unit 3 calculates a first correction amount at time t. Subsequently, in step 5 (S5), the correction amount estimation unit 7 determines whether or not a cycle slip has occurred. When the difference between the value of the carrier phase at t-1 and the value of the carrier phase at t is equal to or greater than the threshold, it is determined that a cycle slip has occurred. Even if there is no previous time data, such as immediately after initialization, it is determined that a cycle slip has occurred. If it is determined that no cycle slip has occurred, the receiver position prediction unit 4 predicts the receiver position in step 6 (S6), the satellite position calculation unit 6 calculates the satellite position, and the GNSS signal A correction amount estimation unit 7 estimates a second correction amount at time t from the observed value of the GNSS satellite output from the reception unit 5. If it is determined that a cycle slip has occurred, the process proceeds to step 7 (S7). In step 7 (S7), using the first correction amount and the second correction amount, the correction amount averaging unit 8 calculates a third correction amount at time t by weighted averaging. If it is determined in step 5 (S5) that a cycle slip has occurred, the weight of the second correction amount is set to zero, and the first correction amount is set to the third correction amount. After completing the processing in steps 4 to 7 (S4 to S7) for each GNSS satellite, the position calculation unit 9 calculates the position of the receiver using the third correction amount in step 8 (S8). Repeat this.

Next, the effects in the present embodiment will be described below as A) to E).
A) By using the third correction amount obtained by the weighted average of the first correction amount and the second correction amount for position calculation, the positioning accuracy can be improved by reducing the interpolation / extrapolation accuracy of the correction amount. It is possible to avoid a decrease, and to obtain a position calculation solution with a temporally uniform accuracy.
B) The total correction amount such as the pseudo-range correction amount and the carrier phase correction amount is not the individual correction amount such as the ionospheric delay amount, but the third correction amount is calculated as the first correction amount and the second correction amount by the weighted average. By adopting a configuration for obtaining the correction amount, an error due to compression of the correction amount other than the ionospheric delay can be reduced.
C) Since the third correction amount obtained by the weighted average of the first correction amount and the second correction amount is used for position calculation, the error of the first correction amount increases with time and the first correction amount increases. Even if the weight of the correction amount decreases, the weight of the third correction amount used for position calculation in order to perform the weighted average with the second correction amount can be increased. For this reason, it is possible to avoid a decrease in weight in calculating the position of the corrected GNSS observation value due to the error expansion of the correction amount over time.
D) Since the second correction amount is calculated in addition to calculating the first correction amount, the first correction amount is compared with the second correction amount, and the correction amount is calculated before the position calculation calculation is performed. It becomes possible to evaluate the reliability. If the accuracy of the correction amount obtained by interpolation / extrapolation is suspected to be low, such as when the difference between the two correction amounts is significant, the use of the correction amount can be stopped to avoid the risk of large positioning errors. .
E) Since the reinforcement information is not directly used for the calculation of the second correction amount, the latest correction information cannot be used for the calculation of the first correction amount when the reinforcement information is not acquired. When the reliability of the amount decreases and position calculation using only the first correction amount as the correction amount becomes impossible until the next reinforcement information is received, the weighted average of the first correction amount and the second correction amount Thus, the third correction amount having higher reliability than the first correction amount can be obtained, and the position can be calculated. Especially in situations where it is likely to fail to receive reinforcement information, such as in urban areas where there are many tunnels and underpasses, the ratio of time (availability) for obtaining a highly accurate position calculation solution as a whole increases. For example, mobile mapping The efficiency of the work of creating a map can be improved.

Embodiment 2. FIG.
FIG. 6 is a block diagram showing a configuration of the positioning apparatus in the second embodiment for carrying out the present invention. The positioning device of the present embodiment shown in FIG. 6 is obtained by adding an autonomous navigation device 10 and an ambiguity adjusting unit 11 to the configuration of FIG. 1 shown in the first embodiment. Further, the position calculation unit 9 is configured to add an estimated value of a sensor state quantity such as a scale factor and bias of the sensor of the autonomous navigation apparatus 10 and a covariance matrix thereof to the output. Further, the operation of each part of the present embodiment will be described.

  The autonomous navigation device 10 includes, for example, an odometer that outputs a speed obtained from the cumulative travel distance of a receiver such as a vehicle, a gyro that outputs angular velocity, an acceleration sensor that outputs acceleration, and the like. Output to the prediction unit 4.

The receiver position prediction unit 4 includes the position X _{t−1 | t−1} and its covariance matrix among the estimated state quantities of the receiver at the time t−1 immediately before the time t at which the position of the receiver is calculated. A term P ^x _{t−1 | t−1} relating to the position of the current position is used as an initial value, and the scale factor and the bias are corrected using an estimated value of the sensor state quantity obtained by the position calculation unit 9 described later. Then, for example, a predicted value X _{t | t−1} and a covariance matrix P ^x _{t | t−1} of the receiver position at time t when the position of the receiver is calculated from the sensor output value by a strapdown navigation operation are predicted. . When sensor outputs s _i (i = 1... n) are obtained at n times between t−1 and t, for example, a predicted value X _{t | t−1} of the receiver position is calculated by numerical integration of the following equation. To do.
<Repeat from i = 1 to n>

一般にｆは、センサの仕様や取り付け位置によって決まる非線形の関数である。共分散行列Ｐ^ｘ _{ｔ｜ｔ−１}は、例えば次式のように線形化したｆを用いて数値積分によって求める。ｄＱはプロセスノイズ行列である。実施の形態１の場合と比較して１ステップの時間幅が短いため、ＱではなくｄＱと記載している。
＜ｉ＝１〜ｎまで繰り返し＞

In general, f is a non-linear function determined by the specification of the sensor and the mounting position. The covariance matrix P ^x _{t | t−1} is obtained by numerical integration using f linearized as in the following equation, for example. dQ is a process noise matrix. Since the time width of one step is shorter than in the case of the first embodiment, dQ is described instead of Q.
<Repeat from i = 1 to n>

  The operations of the reinforcement information reception unit 1, the reinforcement information storage unit 2, the correction amount interpolation / extrapolation unit 3, the GNSS signal reception unit 5, and the satellite position calculation unit 6 are the same as those in the first embodiment.

The operation of the correction amount estimation unit 7 is the same as that of the first embodiment except for the following points. In the first embodiment, only when it is determined that a cycle slip does not occur between time t-1 and time t, the correction amount estimation unit 7 is processed, and when it is determined that a cycle slip has occurred. The standard deviation of the second correction amount is set to infinity, and the weight of the second correction amount in the calculation of the correction amount averaging unit 8 is set to zero. In this embodiment, when it is determined that the cycle slip has not occurred, the same processing as in the first embodiment, and when it is determined that the cycle slip has occurred, the second correction amount (prc2, cpc2) and its standard deviation (σ_prc2, σ_cpc2) are calculated.
・ Prc2 = ρ− (pseudo distance−satellite clock error−relativistic effect)
Cpc2 = ρ− (carrier phase−satellite clock error−relativistic effect)
Σ_prc2 = √ (geometric distance error variance value + pseudo distance observation error variance value)
Σ_cpc2 = √ (geometric distance error variance value + carrier phase observation error variance value)
Refer to the description of Embodiment 1 for details of the description of each section.
Next, the operation of the ambiguity adjusting unit 11 will be described with reference to FIG. The ambiguity adjustment unit 11 adjusts the wave number ambiguity of the carrier phase correction amount in the second correction amount output from the correction amount estimation unit 7, but the cycle slip occurs between time t−1 and time t. The process is performed only when it is determined that the error has occurred. That is, when a cycle slip occurs, the wave number up to time t-1 (the integer determined value when the integer value is determined, the real number estimated value when not determined) is not used, and the second at time t Since the carrier phase correction amount of the correction amount is calculated with the wave number as zero, the carrier phase correction amount of the second correction amount has an ambiguity that is an integral multiple of the wavelength. In FIG. 7, the carrier phase correction amount of the _first correction amount is expressed as C ₁ , and the carrier phase correction amount having an ambiguity that is an integral multiple of the wavelength of the second correction amount is expressed as C ₂ . For this reason, the difference between the first correction amount and the second correction amount becomes large. If the correction amount averaging unit 8 averages the difference as it is, a large error occurs in the third correction amount. Therefore, in order to avoid the occurrence of a large error in the third correction amount, the value of the first correction amount is used to calculate the wave number that should be applied when calculating the carrier phase correction amount of the second correction amount. An assumed value is obtained, and the second correction amount to which the assumed value is applied and the first correction amount are averaged.
In particular,
Difference Dif =
The first correction amount of carrier phase correction amount− (second correction amount of carrier phase correction amount−wavelength × N1) | is determined as N1 that minimizes and N2 that minimizes second. ,
When the absolute value | Dif1−Dif2 | of the difference Dif1 when N1 is applied and the difference Dif2 when N2 is applied is equal to or larger than a specified value, N1 is adopted as the assumed value.
If | Dif1-Dif2 | is equal to or greater than the specified value, it is determined that N1 is correct.
If | Dif1-Dif2 | is equal to or less than the specified value, the accuracy of either the first correction amount or the second correction amount is suspected. In this case, at least one of the following two processes can be selected in accordance with the system reliability requirement.
In the correction amount averaging unit 8, the weight of the second correction amount is set to zero so that the third correction amount = the first correction amount.
-Stop using the correction amount of the satellite until the next reinforcement information is received.
The operation of the correction amount averaging unit 8 is the same as that in the first embodiment.
The position calculation unit 9 performs position calculation calculation using the third correction amount, its standard deviation σ_prc3, σ_cpc3, the GNSS observation value, the state amounts of the receiver and the autonomous navigation sensor, and its covariance matrix. Estimated value Xt | t−1 of the position of the receiver, estimated value Nt | t−1 of the wave number of an undecided satellite, estimated value St | t−1 of the sensor state quantity, and covariance of all the state variables Using the matrix Pt | t−1, the third correction amount, and the GNSS observation value, for example, by the time update and the observation update by the Kalman filter, the estimated value Xt | t of the receiver position at the time t and the estimated value Nt of the wave number | T, sensor state quantity estimate St | t and covariance matrix Pt | t are calculated. The standard deviation of the third correction amount may be used in addition to the observation noise in the observation update. For satellites for which the integer value of the wave number of the carrier phase has not been determined, an adjustment is made from the estimated value of the wave number and the value of the term corresponding to the wave number of the covariance matrix using LAMBDA (The Last Squares Adjustment Adjustment) or the like. Find numerical candidates. If an integer solution candidate passes a test such as a ratio test or a residual test, the wave number is determined, and the real number estimated value is replaced with an integer value.

  Next, the positioning process of this Embodiment is demonstrated using the flowchart of FIG. The processing from step 1 (S1) to step 5 (S5) is the same as in the first embodiment. In Embodiment 2, when it is determined in step 5 (S5) that no cycle slip has occurred, sensor data of the autonomous navigation device is acquired from the autonomous navigation device 10 in step 6 (S6), and receiver position prediction is performed. The position of the receiver is calculated by the unit 4, the satellite position is calculated by the satellite position calculation unit 6, and the observation value of the GNSS satellite output from the GNSS signal reception unit 5 is used to calculate the time t at the correction amount estimation unit 7. A second correction amount at is estimated. If it is determined that a cycle slip has occurred, the sensor data of the autonomous navigation device is acquired from the autonomous navigation device 10 in step 7 (S7), the receiver position prediction unit 4 calculates the position of the receiver, and the satellite position The satellite position is calculated by the calculation unit 6, and the second correction amount at time t is estimated by the correction amount estimation unit 7 from these and the observed value of the GNSS satellite output from the GNSS signal reception unit 5. However, in the calculation of the carrier wave phase correction amount of the second correction amount, the wave number value is calculated as zero. Subsequently, in step 8 (S8), the ambiguity adjusting unit 11 adjusts the ambiguity of the second correction amount due to the wave number being zero, using the value of the first correction amount. Subsequently, in step 9 (S9), the correction amount averaging unit 8 uses the first correction amount and the second correction amount to calculate a third correction amount at time t by weighted averaging. After completing the processing in steps 4 to 9 (S4 to S9) for each GNSS satellite, the position calculation unit 9 calculates the position of the receiver using the third correction amount in step 10 (S10). Repeat this.

Next, specific effects in the present embodiment will be described below as A) to C). In the present embodiment, the following effects are obtained in addition to the effects of the first embodiment of the present invention.
A) By adding the ambiguity adjusting unit 11 as a component, the second correction amount can be used even when a cycle slip occurs between time t-1 and time t.
B) By adding sensor data of the autonomous navigation device 10 to the input of the receiver position prediction unit 4, even if the GNSS satellite is invisible for a long time, such as under a tunnel or overpass, it is compared with the case of the first embodiment. Thus, the receiver position at time t can be obtained with higher accuracy, and therefore the second correction amount with higher accuracy can be estimated.
C) FIG. 9 is a diagram for explaining the operation effect of the positioning device according to the present embodiment. FIG. 9 shows one receiver ((a) to (c)), and the movement of the position of the receiver over time is shown from left to right. When the receiver moves to the position (a), both the satellite that distributes the reinforcing information such as the quasi-zenith satellite and the GNSS satellite are visible from the receiver. When the receiver moves to the position (b), it passes through a tunnel or the like, thereby passing through a section where the satellite becomes invisible when viewed from the receiver. Then, when the receiver moves to the position (c) after passing through the tunnel or the like, it is assumed that the satellite becomes visible again as viewed from the receiver. The reinforcement information is continuously distributed from the satellite to the GNSS positioning device even when the receiver moves to any position of (a) to (c). When the receiver moves to the position (a), the GNSS positioning device connected to the receiver receives the reinforcement information from the GNSS satellite, and then extrapolates the correction amount until receiving the reinforcement information. To do. The “Fix state” before entering the invisible section in FIG. 9 indicates a state in which the wave number of the GNSS satellite signal equal to or more than the above-mentioned prescribed number is determined and Xt | t is obtained with high accuracy. Even when the receiver moves to the position (b), the GNSS satellite distributes the reinforcement information to the GNSS positioning device, but the receiver is connected to the receiver because it travels in the invisible section. The GNSS positioning device cannot obtain the latest reinforcement information. Therefore, immediately after becoming visible again, the GNSS positioning device predicted from the autonomous navigation device during the first correction amount obtained from the reinforcement information obtained up to the time of moving to the position (a) and the invisible section. A third correction amount is obtained by weighting with the second correction amount calculated from the position of the receiver, and positioning calculation is performed. Until the subsequent receiver receives the reinforcement information at the position (c), the GNSS positioning apparatus obtained from the reinforcement information obtained when the receiver moved to the position (a). The third correction amount is weighted by the weight of the first correction amount and the second correction amount calculated from the position of the receiver predicted from the navigation device using the position estimation value based on the position calculation result immediately after becoming visible as the initial value. The correction amount is obtained and the position calculation of the receiver is performed. As shown in FIG. 9, in a GNSS positioning device in an urban area, when a receiver passes through a tunnel or under an overhead, the GNSS positioning device fails to receive reinforcement information, and a correction amount interpolation / extrapolation unit Since the first correction amount is calculated in step 3, the latest reinforcement information cannot be used. In such a case, for example, when the correction amount at a plurality of past times is calculated from the reinforcement information, the correction amount of the observation value of the GNSS satellite immediately after the satellite leaves the invisible section can be calculated from the latest reinforcement information. The correction amount must be extrapolated from the correction amount at the previous time with respect to the expected correction amount time, and the accuracy of the correction amount is reduced. Therefore, when the wave number of the GNSS satellite signal exceeding the specified number is determined immediately before entering the invisible section and Xt | t is obtained with high accuracy, the position is predicted by the autonomous navigation device using it as an initial value. If it is about several tens of seconds, the position prediction accuracy capable of estimating the second correction amount can be maintained, and the decrease in accuracy of the first correction amount due to the extrapolation described above can be compensated. The accuracy of position prediction using the autonomous navigation device is higher than the prediction accuracy of the receiver position prediction unit in the first embodiment, and is effective for long-time prediction such as when passing under a tunnel or an overhead. The time that can be predicted depends on the specifications of the autonomous navigation device 10.

  1 reinforcement information receiving unit, 2 reinforcement information storage unit, 3 correction amount interpolation / extrapolation unit, 4 receiver position prediction unit, 5 GNSS signal reception unit, 6 satellite position calculation unit, 7 correction amount estimation unit, 8 correction amount averaging Unit, 9 position calculation unit, 10 autonomous navigation device, 11 ambiguity adjustment unit

Claims (3)

In a GNSS positioning device that calculates the position of a receiver that receives a ranging signal transmitted from a GNSS satellite,
Reinforcement information receiving unit that receives reinforcement information related to the correction amount of the observation value of the pseudorange and the carrier phase obtained from the ranging signal;
A correction amount interpolation extrapolation unit for predicting a first correction amount for correcting the position of the receiver at a position calculation time for calculating the position of the receiver based on the received reinforcement information;
A receiver position prediction unit that predicts the position of the receiver at the position calculation time based on a state quantity estimation value including the position and speed of the receiver calculated by the position calculation unit at the previous position calculation time; ,
A distance from the GNSS satellite at the position calculation time is calculated based on the position of the receiver predicted by the receiver position prediction unit and the satellite position at the position calculation time, and is calculated from the distance and the distance measurement signal. A correction amount estimator that calculates a second correction amount for correcting the position of the receiver at the position calculation time, based on the observed value including the pseudorange and the carrier wave phase at the position calculation time;
A position calculator that corrects the position of the receiver based on a third correction amount obtained based on the first correction amount and the second correction amount;
GNSS positioning device equipped with.   The position calculation unit includes an ambiguity adjustment unit that adjusts the ambiguity of the wave number of the carrier phase correction amount in the second correction amount output from the correction amount estimation unit. GNSS positioning device.   The position calculation unit includes an autonomous navigation device that outputs sensor data for autonomous navigation, and specifies position information of the receiver based on the data of the sensor for autonomous navigation. GNSS positioning apparatus of description.

Images