JP5164546B2 - Positioning method - Google Patents

Description

  The present invention relates to a positioning method, and more particularly to a positioning method for measuring a position based on a signal transmitted from a positioning satellite.

For example, in Patent Literature 1, a receiver position is predicted based on the previous positioning result, and a pseudo-range prediction amount is calculated based on the predicted receiver position. Next, the difference between the predicted amount of the pseudorange and the observed amount is calculated, and when this difference exceeds the allowable error range, it is determined that the received wave has multipath and is not used for the positioning calculation. ing.
JP 2003-57327 A

  According to the prior art, when a moving body such as an automobile equipped with a positioning device (receiver) is accelerated or decelerated, an error occurs in the predicted position, and an error also occurs in the predicted amount of the pseudo distance. In such a case, if multipath detection is performed for each pseudo distance based on the difference between the predicted amount and the observed amount, the error due to acceleration / deceleration and the error due to multipath cannot be determined, and multipath False detection and non-detection increase.

  The present invention has been made in view of such problems, and reduces false detection and non-detection of multipath.

  In order to solve the above problems, the present invention employs the following means.

Pseudorange between the treatment process according to the basic technique of the positioning method of the present invention includes the steps of receiving orbit information 及 beauty time information positioning satellite is transmitting, the receiver and the positioning satellite based on the received time information calculating a, and calculating the position of the positioning satellite based on the received orbit information 及 beauty said time information, said calculated pseudoranges and the current of the receiver based on the position of the positioning satellite calculating a position, a step of calculating a predicted position of the receiver, calculating the predicted amount of the pseudo-range based on the position of the predicted position 及 beauty positioning satellite of said receiver, the received time when pseudorange but based on a difference between the pseudo distance and the predicted amount of the pseudorange in the difference, and reception time of the predicted amount of pseudoranges and the pseudorange 1 epoch before the previous one Calculating a prediction measurement errors of the difference, and having the steps of: determining the occurrence of multipath the prediction observation error based on the result of dividing by the standard deviation of the prediction observation error.

According to the present invention, since the error due to acceleration / deceleration continuously changes with time, the error in the predicted amount of the pseudo distance due to acceleration / deceleration is reduced by taking the time difference for the difference between the pseudo distance and the predicted pseudo distance. it is possible to accurately detect multipath Te, as a result, the ability to reduce erroneous detection and non-detection of multipath, it is possible to display the highly accurate position that is not affected by multipath.

  Hereinafter, the best embodiment will be described with reference to the accompanying drawings. FIG. 1 is a diagram illustrating a positioning device (receiver) according to the first embodiment. Note that multipath false detection and non-detection reduction can be realized by dividing a pseudo-range predicted observation error by its standard deviation and performing a statistical test, as will be described later.

  As shown in FIG. 1, the receiver includes a receiving unit 101 and a calculating unit 102. The receiving means 101 has processing functions of an antenna, a down converter, an analog / digital converter, a quadrature detector, a C / A (coarse / acquisition) code generator, a correlation detector, and a decoder. Signals sent from positioning satellites (positioning signals) are received by antennas, navigation messages including orbit information, positioning satellite status information, ionospheric delay calculation parameters, etc. are detected, pseudorange (distance from satellite), Doppler Measure observation data such as frequency and signal strength. A positioning satellite is a device that transmits a signal for positioning from space or the ground, such as a GPS satellite or a pseudo satellite.

  A computing means 102 such as a CPU (central processing unit) detects a multipath based on a navigation message, a pseudorange, a Doppler frequency and a signal intensity from the receiving means 101, and detects a position, velocity and acceleration. Is calculated.

  FIG. 2 is a diagram for explaining an operation procedure of the positioning apparatus (receiver) shown in FIG.

  Step 201: The receiving means 101 receives a signal from a positioning satellite with an antenna, and performs a down-conversion, analog / digital conversion, quadrature detection, C / A code generation, correlation detection, and decoding processing, thereby obtaining a navigation message. Is detected, and the reception time, pseudorange, Doppler frequency and radio wave intensity are measured. These observation data are sent to the calculation means 102.

  Step 202: When the number of received positioning satellite signals is 4 or more, the calculation means 102 proceeds to step 203. If it is 3 or less, the process proceeds to step 201.

  Step 203: If the computing means 102 detects a multipath before one epoch, it proceeds to step 205. If not detected, the process proceeds to step 204. Here, epoch means the time of observation data acquisition (see Tatsunori Sada: GPS Survey Technology, Ohmsha (October 2003)).

  Step 204: If the number of positioning satellite signals received before one epoch is three or less, the calculation means 102 proceeds to step 205. If it is four or more, the process proceeds to step 206.

  Step 205: The computing means 102 calculates the positioning satellite position based on the navigation message and the reception time. Based on the positioning satellite position and the pseudorange, a single positioning calculation process is performed to calculate an initial position. The elements of the estimator are position, velocity, acceleration and clock bias (clock error), assuming acceleration to be a first order Markov process. The initial position is set to the position of the estimated amount, and the speed, acceleration and clock bias are set to 0. The reciprocal of the time constant of acceleration and the variance of acceleration are set to predetermined values. The state transition matrix, system noise covariance matrix and observation noise matrix are set to predetermined values. Also, the estimation error covariance matrix is set to an initial value.

  Step 206: The computing means 102 calculates the positioning satellite position based on the orbit information and the reception time of the navigation message. Based on the ionospheric delay calculation parameters, positioning satellite position, positioning position, and reception time included in the navigation message, the ionospheric delay in each positioning satellite is calculated, subtracted from the pseudorange, and corrected.

  Step 207: The computing means 102 calculates the tropospheric delay in each positioning satellite based on the positioning satellite position and the positioning position using the troposphere model. Predetermined values are used for relative humidity, temperature, and atmospheric pressure. The Saastamoinen model is used as the troposphere model. Subtract the tropospheric delay from the pseudorange and correct.

Step 208: The computing means 102 calculates the distance between the predicted positioning position and the positioning satellite position, which is an element of the one-step prediction amount, and sets this as the predicted pseudo distance. The predicted pseudorange is subtracted from the pseudorange to calculate the predicted observation error of the pseudorange. The pseudo-range predicted observation error ν (k) is expressed as Equation 1 as an approximate expression. Here, y (k) is an observation amount (here, pseudorange), and η (k | k−1) is a one-step prediction amount (here, position, velocity, acceleration, clock bias).

Step 209: The calculation means 102 calculates the matrix C (k) based on the predicted positioning position and the positioning satellite position using Expressions 2 to 6. Here, (xu, yu, zu) is the predicted positioning position, (xq, yq, zq) is the positioning satellite position, q is the positioning satellite number, i is a vector whose elements are all 1, k is an epoch, and m is a positioning satellite. Is a number.

The covariance matrix M (k) of the pseudo-range predicted observation error is calculated using Equation 7. Here, P (k | k−1) is an estimation error covariance matrix of the one-stage prediction amount η (k | k−1), and R is an observation noise matrix.

Using Cholesky factorization, a regular matrix L (k) that is Eq. 8 with respect to M (k) is calculated. Here, the matrix L (k) corresponds to the standard deviation of the predicted observation error.

Νs (k) obtained by dividing the predicted observation error by its standard deviation using Equation 9 is a normal distribution in which each element is independent of each other.

Step 210: The computing means 102 follows the χ square distribution with m−1 degrees of freedom for the test statistic T (k) expressed by Equation 10.

  Therefore, a risk factor of 0.05 is set, and a hypothesis that a multipath has occurred and a hypothesis that a multipath has not occurred are established. When the test statistic T (k) is equal to or less than the rejection limit value, it is determined that multipath has not occurred. If it is larger, it is determined that multipath has occurred. The risk factor may be a value other than 0.05.

  Step 211: If it is determined that multipath has occurred, the process proceeds to Step 201. If it is determined that multipath has not occurred, the process proceeds to step 212.

  Step 212: The calculation means 102 estimates the position, velocity, acceleration, clock bias, and their estimated error covariance using the Kalman filter based on the pseudorange single difference as follows.

The Kalman gain matrix K (k) is calculated using Equation 11.

Using Equation 12 and Equation 13, the estimated amount η (k | k) and single predicted amount η (k + 1 | k) of position, velocity, acceleration, and clock bias are calculated. Here, Φ (Δt, α) is a state transition matrix, Δt is a sampling interval of observation data, and α is a reciprocal of a time constant of acceleration. The position of the estimated amount becomes the positioning position, and the position of the one-step predicted amount becomes the predicted positioning position.

The estimated error covariance matrices P (k | k) and P (k + 1 | k) of η (k | k) and η (k + 1 | k) are calculated using Equations 14 and 15. Here, Q (k) is a covariance matrix of system noise.

Step 213: The calculation means 102 calculates prediction errors eP and eH of the three-dimensional position and the planar positioning position based on the pseudo-range predicted observation error ν (k) using Expressions 16 to 19. Here, Ajl is an element of j row and l column of matrix A, νq is an element of vector ν, eρ is an error in pseudorange due to satellite clock, satellite position, ionosphere delay, troposphere delay, and receiver clock. For example, eρ is set to 2.1 m.

  As described above, when a moving body such as an automobile equipped with a positioning device is accelerated or decelerated, an error occurs in the predicted position and an error occurs in the predicted amount of the pseudo distance in multipath detection. The error due to acceleration / deceleration and the error due to multipath cannot be determined, and multipath misdetection and non-detection increase. However, according to the operation procedure shown in FIG. 2, the position of one epoch is predicted based on the position, velocity, and acceleration, so that the position can be predicted with high accuracy. For this reason, errors due to acceleration / deceleration can be reduced, and multipath can be accurately detected.

  Further, according to the operation procedure shown in FIG. 2, when an error due to acceleration / deceleration occurs, the error gradually increases and the covariance value increases. On the other hand, an error due to multipath appears suddenly, and the covariance value is small when it occurs. Therefore, by dividing the prediction error of the pseudo distance of all received signals by the standard deviation, the error due to multipath appears larger than the error due to acceleration / deceleration, and it becomes easy to detect the multipath accurately.

  In addition, according to the operation procedure shown in FIG. 2, the accuracy (quality) of the positioning position can be managed because the positioning position prediction error is calculated based on the pseudo-range prediction observation error, the positioning position, and the satellite position. it can.

  Next, a modified example of the positioning method will be described.

(Positioning method with estimated jerk)
In the operation procedure shown in FIG. 2, assuming that the elements of the estimation amount are position, velocity, acceleration, jerk (jerk acceleration) and clock bias, and the jerk is a first-order Markov process, the position, velocity, acceleration and jerk are Since the one-step prediction amount is calculated based on the predicted value, the predicted positioning position is less affected by acceleration / deceleration, and the error due to acceleration / deceleration included in the pseudo-range prediction amount is smaller than in the method that predicts only acceleration. . For this reason, multipath can be detected accurately.

(Positioning method using pseudo-range single difference)
In the operation procedure shown in FIG. 2, the elements of the estimation amount are position, velocity, and acceleration, the positioning satellite with the highest elevation angle is the reference satellite, and the observation amount is a pseudo value obtained by subtracting the pseudo distance of the reference satellite from the pseudo distance of each positioning satellite. In step 208, the predicted pseudorange of the reference satellite is subtracted from the predicted pseudorange of each positioning satellite, and this is set as the single difference of the predicted pseudorange. The pseudorange of the reference satellite is subtracted from the pseudorange of each positioning satellite to calculate a single pseudorange difference. The single difference of the predicted pseudorange is subtracted from the single difference of the pseudorange, and this is used as the predicted observation error of the pseudorange. In step 209, the calculation means 102 calculates the matrix C (k) based on the predicted positioning position and the positioning satellite position using Expressions 20 to 24. 1 is the number of the reference satellite.

  In step 212, the estimated amount is calculated by using the pseudo-range single difference as an observed amount in Expression 12.

When the above-described processing is performed, the number of elements of the matrix C (k) is changed from m × 10 to (m−1) × 9, and the number of elements of M (k) and P (k | k) is changed from m × m to (m −1) × (m−1), and the amount of calculation can be reduced. As a result, it is possible to execute even calculation means with low calculation performance and low cost.

(Positioning method using difference in predicted observation error)
In the operation procedure shown in FIG. 2, in Step 208, using Formula 25 instead of Formula 1, subtracting the predicted pseudorange from the pseudorange and subtracting it one epoch before, this is calculated as the predicted observation error of the pseudorange. And

When the covariance matrix M (k) of the predicted observation error is calculated using Equation 26 instead of Equation 7 in Step 209, the error due to acceleration / deceleration changes continuously with time. By taking the time difference of the subtraction, the error due to acceleration / deceleration can be reduced, and multipath can be easily detected accurately.

(Positioning method using Doppler frequency)
In the operation procedure shown in FIG. 2, the estimated amount is a position, velocity, acceleration, clock bias, and positioning satellite velocity, and the observed amount is a pseudorange, Doppler frequency, and positioning device velocity. In step 208, using the equation 27 instead of the equation 1, the relative speed of the positioning satellite and the positioning device is obtained from the current epoch and the Doppler frequency one epoch before, the pseudo epoch one epoch before the epoch pseudo-range, and the positioning The relative speed between the satellite and the positioning device is subtracted to calculate the pseudorange prediction error. Here, yp (k) is the pseudorange, yD (k) is the Doppler frequency, f is the frequency of the carrier wave, and c is the speed of light.

In step 209, using the equation 28 instead of the equation 7, the covariance matrix M (k) of the predicted observation error of the pseudorange is calculated based on the observation noise matrix of the pseudorange and the Doppler frequency. Here, Rp is a pseudo-range observation noise matrix, and RD is a Doppler frequency observation noise matrix.

  In step 212, the position difference at the time obtained by adding 0.01 seconds to the reception time and the reception time, for example, is divided by 0.01 seconds to calculate the velocity of the positioning satellite and used in Equation 12. When the above-described processing is performed, since the Doppler frequency is not easily affected by multipath and diffraction, both the Doppler frequency can be accurately calculated for acceleration / deceleration, errors due to acceleration / deceleration can be reduced, and multipath can be detected accurately.

  FIG. 3 is a diagram illustrating a positioning device (receiver) according to the second embodiment.

  The positioning device (receiver) in this embodiment includes a receiving unit 101, a calculating unit 102, a speed sensor 103, an angular velocity sensor 104, and an acceleration sensor 105.

  The reception unit 101 and the calculation unit 102 are the same as those described in the first embodiment. However, the calculation means 102 also calculates the bias, position, speed, acceleration, and direction of the angular velocity sensor 104 and the acceleration sensor 105 based on the vehicle speed, the angular velocity, and the acceleration.

  A speed sensor 103 such as a vehicle speed sensor is installed on an axle or the like in an automobile and outputs a speed. An angular velocity sensor 104 such as a gyro outputs an angular velocity. The acceleration sensor 105 such as an accelerometer outputs acceleration.

  The operation procedure of the positioning device (receiver) shown in FIG. 3 will be described with reference to FIG.

  Step 301: Steps 201 to 207 shown in FIG. 2 are executed, a navigation message is detected, the reception time, pseudorange, Doppler frequency and radio wave intensity are measured, and the pseudorange is corrected.

  Step 302: Steps 208 to 210 are executed to determine whether or not a multipath has occurred.

  Step 303: If it is determined that multipath has occurred, the process proceeds to step 306. If it is determined that multipath has not occurred, the process proceeds to step 304.

  Step 304: Step 212 is performed to calculate estimated values and predicted values of position, velocity and acceleration.

  Step 305: Step 213 is executed to calculate a positioning position prediction error.

  Step 306: The speed sensor 103 outputs the speed and sends it to the computing means 102. The angular velocity sensor 104 outputs an angular velocity and sends it to the computing means 102. The acceleration sensor 105 outputs acceleration and sends it to the computing means 102. The computing means 102 receives the speed, angular velocity and acceleration.

  Step 307: If the computing means 102 determines that multipath has occurred, it proceeds to step 310. If it is determined that multipath has not occurred, the process proceeds to step 308.

  Step 308: If the positioning position prediction error is within the threshold value, the calculation means 102 proceeds to step 309. If so, go to step 310.

  Step 309: The computing means 102 calculates the bias and position error of the angular velocity sensor 104 and the acceleration sensor 105 based on the estimated position, velocity and acceleration values calculated in step 304, and the position, velocity, acceleration and direction before one epoch. Calculate velocity error and bearing error.

  Step 310: The computing means 102 corrects the angular velocity output from the angular velocity sensor 104 and the acceleration output from the acceleration sensor 105 based on the biases of the angular velocity sensor 104 and the acceleration sensor 105. The corrected angular velocity is added to the azimuth before one epoch, the azimuth error is subtracted, and the azimuth is calculated. The corrected acceleration is added to the speed one epoch before, the speed error is subtracted, and the speed is calculated. The calculated speed is added to the position one epoch before, the position error is subtracted, and the position is calculated.

  According to the operation procedure shown in FIG. 4 (second embodiment), when it is determined that multipath has not occurred, the estimated position, velocity, and acceleration are used so that the positioning position prediction error is within the threshold. The bias, position error, velocity error, and azimuth error of the angular velocity sensor 104 and the acceleration sensor 105 can be calculated and corrected. Therefore, the biases of the angular velocity sensor 104 and the acceleration sensor 105 can be calculated with high accuracy, and even with the low-precision and low-priced angular velocity sensor 104 and acceleration sensor 105, a highly accurate position can be output.

  As described above, according to the embodiment of the present invention, when an error due to acceleration / deceleration occurs, the error gradually increases and the covariance value increases. On the other hand, an error due to multipath appears suddenly, and the covariance value is small when it occurs. Using this fact, the prediction error of the pseudorange of all received signals is divided by the standard deviation. The error due to multipath appears larger than the error due to acceleration / deceleration, and the multipath can be detected with high accuracy. Thereby, the precision of a positioning position can be improved.

  In addition, by detecting multipath based on the difference between the single pseudorange and the predicted amount, the number of matrix elements used in multipath detection and estimation calculation is reduced, and multipath detection is performed at pseudorange. The amount of calculation can be reduced as compared with the case of doing so. As a result, it is possible to execute even calculation means with low calculation performance and low cost.

  In addition, a time difference is obtained by subtracting the predicted pseudo distance from the pseudo distance, and multipath occurrence is determined based on the time difference. Since the error due to acceleration / deceleration changes continuously with time, the error due to acceleration / deceleration can be reduced, and multipath can be easily detected accurately.

  In addition, multipath is detected based on the pseudorange and the Doppler frequency. Since the Doppler frequency is not easily affected by multipath and diffraction, acceleration / deceleration can be accurately calculated based on the Doppler frequency, errors due to acceleration / deceleration can be reduced, and multipath can be detected accurately.

  Further, since the position of one epoch is predicted based on the position, speed, and acceleration, the position can be predicted with high accuracy. For this reason, errors due to acceleration / deceleration can be reduced, and multipath can be accurately detected.

  In addition, since the position one epoch ahead is predicted based on the position, speed, acceleration, and jerk, the position can be predicted with high accuracy. For this reason, errors due to acceleration / deceleration can be reduced, and multipath can be accurately detected.

  In addition, since the prediction error of the positioning position is calculated based on the predicted observation error of the pseudo distance, the positioning position, and the satellite position, the accuracy (quality) of the positioning position can be managed.

  Further, when it is determined that the multipath has not occurred, the estimated position, velocity, and acceleration values within which the positioning position prediction error is within the threshold value are used to bias the angular velocity sensor 104 and the acceleration sensor 105, the position error, and the velocity. Error and azimuth error can be calculated and corrected. Therefore, the biases of the angular velocity sensor 104 and the acceleration sensor 105 can be calculated with high accuracy, and even with the low-precision and low-priced angular velocity sensor 104 and acceleration sensor 105, a highly accurate position can be output.

  Further, by connecting a device using the positioning method of the present embodiment to a car navigation device, it is possible to display a highly accurate position that is not adversely affected by multipath. In addition, bias correction values for gyroscopes, accelerometers, and the like can be calculated with high accuracy based on the high-accuracy position obtained by the positioning method. For this reason, it is possible to calculate a highly accurate position by the self-contained navigation even in a state where a signal from a positioning satellite such as a GPS satellite cannot be received.

It is a figure explaining the positioning apparatus (receiver) concerning 1st Embodiment. It is a figure explaining the operation | movement procedure of the positioning apparatus (receiver) shown in FIG. It is a figure explaining the positioning apparatus (receiver) concerning 2nd Embodiment. It is a figure explaining the operation | movement procedure of the positioning apparatus (receiver) shown in FIG.

Explanation of symbols

101 receiving means 102 computing means 103 speed sensor 104 angular velocity sensor 105 acceleration sensor

Claims (6)

Receiving orbit information 及 beauty time information positioning satellites originates,
Calculating a pseudo-range between the positioning satellite and the receiver based on the received time information,
Calculating a position of the positioning satellite based on the received orbit information 及 beauty the time information,
Calculating the current position of the receiver based on the calculated pseudorange and the position of the positioning satellite;
Calculating a predicted position of the receiver,
Calculating a predicted amount of the pseudo-range based on the position of the predicted position 及 beauty positioning satellite of said receiver,
Prediction of the time difference of the pseudo distance based on the difference between the pseudo distance one epoch before the reception time point and the predicted amount of the pseudo distance, and the difference between the pseudo distance at the time of reception and the prediction amount of the pseudo distance Calculating an observation error;
Positioning method characterized in that it comprises the steps of: determining the occurrence of multipath on the basis of the result of dividing the prediction observation error with a standard deviation of the prediction observation error. The positioning method according to claim 1,
A positioning method, wherein a predicted position of one epoch is calculated based on a result of calculating the position, velocity, and acceleration of the receiver based on the pseudorange and the position of the positioning satellite . The positioning method according to claim 1,
A positioning method, wherein a predicted position of one epoch ahead is calculated based on the result of calculating the position, velocity, acceleration, and jerk of the receiver based on the pseudorange and the position of the positioning satellite . The positioning method according to claim 1,
A positioning method, comprising: calculating a prediction error of a predicted position of the receiver based on a predicted observation error of a time difference of the pseudorange, a predicted position of the receiver, and a position of the positioning satellite . The positioning method according to claim 4,
Wherein with multipath is determined to not occur, if the prediction error of the predicted position of the receiver is within the threshold, 1 epoch previous position, velocity, position of the receiver which issued calculated as 及 beauty orientation及 A positioning method characterized by calculating a bias, a position error, a speed error, and an azimuth error of an angular velocity sensor means for outputting an angular velocity based on the velocity and the velocity. The positioning method according to claim 4,
Wherein with multipath is determined to not occur, if the prediction error of the predicted position of the receiver is within the threshold, 1 epoch previous position, velocity, position of the receiver which issued calculated as 及 beauty orientation及 A positioning method comprising: calculating a bias of an angular velocity sensor means for outputting an angular velocity, a bias of an acceleration sensor means for outputting an acceleration, a position error, a velocity error, and an azimuth error based on the velocity and the velocity.

Images