JP2012052954A - Position finding apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a position finding apparatus capable of finding the current position with high accuracy and on a real time basis even when a relatively large receiver clock drift occurs.SOLUTION: A position finding apparatus 20 comprises a receive signal analyzer 21 that analyzes receive signals; a receiver clock 22 that provides reception time information to the receive signal analyzer 21; and a Kalman filter 40 that estimates the position of a user, the velocity of the user and the receiver time, wherein the Kalman filter 40 is provided with state predicting means 41 that predicts a state, state estimator 42 that estimates the state and clock abnormal variation detector 43 that detects any abnormal variation in the receiver clock 22, the state predicting means 41 being further provided with a state predictor 41a that predicts a state variable vector of the current epoch on the basis of a state variable vector of one epoch before obtained from the state estimator 42 and a predicted state value corrector 41b that corrects the predicted state value on the basis of the result of detection by the clock abnormal variation detector 43.

Description

  The present invention relates to a positioning device for a satellite navigation system using a Kalman filter.

  A positioning device in a satellite navigation system such as GPS is obtained by measuring radio wave transmission time (hereinafter referred to as “satellite transmission time”), satellite position, satellite speed, and received radio wave carrier obtained from information added to the radio wave transmission. User position change, satellite position change, Doppler shift due to frequency shift of receiver oscillator, etc., user position, user speed and time of satellite navigation system system in receiver (hereinafter referred to as “receiver time”) This is an estimation device.

  In this positioning device, the user position, the user speed and the receiver time are estimated by modeling the change from the previous state to the current state and the relationship between the current state and the observed value, and the model equation is expressed by a Kalman filter. The solving method is common.

  In the Kalman filter, the user position, the user speed, and the receiver time are taken as state variables, and the current state variable value and its covariance are calculated using a prediction model from the state variable value and its error covariance matrix one epoch before. A process of predicting a matrix and a process of estimating the current state variable value and the covariance matrix from the state variable value and the covariance matrix predicted using the relational expression between the current observation amount and the state estimated value are performed. Epoch means when the observation data is acquired.

  Normally, when estimating a user position, a user speed, and a receiver time using a Kalman filter, the state variable vector x is generally expressed by [Equation 1].

  Here, p is the user position (three-dimensional vector), Δt is the difference between the time of the receiver clock and the time in the satellite navigation system (hereinafter “receiver clock offset”), and v is the user speed (three-dimensional vector). , ΔT is a receiver clock offset deviation (hereinafter “receiver clock drift”) per unit time in the receiver clock, and a superscript “T” indicates a transposed matrix. The user position refers to the position of the user who uses the receiver, and the user speed refers to the moving speed of the user.

  In a positioning device of a satellite navigation system using a Kalman filter, in a receiver clock equipped with an oscillator, it can be assumed that the amount of change in the receiver clock drift per epoch follows a Gaussian distribution, and the amount of dispersion is determined by the oscillator clock. Determined by frequency stability. By incorporating the above assumption into the prediction model, it is possible to improve the estimation accuracy of the receiver clock drift, and the estimation accuracy improves as the oscillator having a higher frequency stability is used.

  In a positioning device of a normal inexpensive satellite navigation system, a temperature-compensated crystal oscillator (hereinafter referred to as “TCXO”) is used as an oscillator. In this oscillator, the frequency stability is relatively high in an environment where the temperature is stable, but on the other hand, when a sudden temperature change occurs or an impact is applied, the receiver clock drift changes greatly. End up. In order to perform stable positioning even in such a state, it is necessary to set a large amount (variance) of the amount of change per epoch of the receiver clock drift. However, when the dispersion is set to be large, there is a problem that the accuracy deteriorates under a situation where the receiver clock drift is stable.

  As a technique for solving such a problem, an adaptive Kalman filter is used (for example, see Non-Patent Document 1). The adaptive Kalman filter shown in Non-Patent Document 1 can determine whether a given system noise model parameter is valid based on the variance of the observed noise, and can adjust the parameter if it is not valid. The state estimation reliability can be improved by performing an averaging process.

Y. Bar-Shalom, X. R. Li, and T. Kirubarajan, Estimation with Applications to Tracking and Navigation: Theory, Algorithms, and Software, Wiley, New York, 2001, pp. 424-425.

  However, in the conventional adaptive Kalman filter, even if the change in the receiver clock drift is stable, a large change in the receiver clock drift occurs when the observation value includes a large observation error. There is a possibility of misjudgment, and there has been a problem of causing deterioration in positioning accuracy. In addition, the conventional adaptive Kalman filter is configured to perform averaging processing in order to increase the reliability of state estimation, so that a change in receiver clock drift cannot be detected in real time, and there is a problem that positioning accuracy deteriorates. It was.

  An object of the present invention is to solve the above-described problem, and to provide a positioning device that can accurately determine the current position in real time even when a relatively large receiver clock drift occurs. .

  The positioning device of the present invention is a positioning device that measures the position of a user of a receiver that receives a positioning signal from a satellite in a satellite navigation system, the user position indicating the position of the user, and the user A user speed indicating a moving speed of the receiver, a receiver clock offset indicating a difference between a time of a receiver clock of the receiver and a time of the satellite navigation system, and the receiver per unit time of the receiver clock Based on a state variable vector including a receiver clock drift indicating a clock offset deviation as a state variable, a Kalman filter that performs state prediction and state estimation related to the satellite navigation system, a positioning signal from the satellite, Observation data acquisition means for acquiring observation data indicating a state, wherein the Kalman filter predicts the state variable vector And a state prediction means for calculating a covariance matrix of the prediction error, and the state prediction by calculating an estimated value of the state variable vector and a covariance matrix of the estimation error based on the calculation result of the state prediction means and the observation data State estimation update means for updating the calculation result of the means, receiver clock drift prediction value included in the prediction value of the state variable vector calculated by the state prediction means, and estimation of the state variable vector calculated by the state estimation update means A difference calculating means for calculating a difference from a receiver clock drift estimated value included in the value, and a variance of the receiver clock drift included in the covariance matrix of the prediction error when the difference exceeds a predetermined threshold The state estimation update unit is configured to correct the variance of the receiver clock drift. The receiver clock drift variance included in the covariance matrix of the prediction error calculated by the state prediction means is replaced with a modified receiver clock drift variance, and the estimated value of the state variable vector and the covariance matrix of the estimation error Is recalculated.

  With this configuration, the positioning device of the present invention enables the receiver clock included in the covariance matrix of the prediction error when the difference between the receiver clock drift predicted value and the receiver clock drift estimated value exceeds a predetermined threshold. Since the dispersion of the drift is corrected to a larger value, the current position can be measured in real time with high accuracy even when a relatively large receiver clock drift occurs.

  The positioning device of the present invention is a positioning device that measures the position of a user of a receiver that receives a positioning signal from a satellite in a satellite navigation system, the user position indicating the position of the user, A user speed indicating a moving speed of the user, a receiver clock offset indicating a difference between a time of a receiver clock included in the receiver and a time in the satellite navigation system, and the unit time per unit time in the receiver clock. A Kalman filter that performs state prediction and state estimation for the satellite navigation system based on a first state variable vector including a receiver clock drift indicating a receiver clock offset deviation as a state variable, and receives a positioning signal from the satellite Observation data acquisition means for acquiring observation data indicating the state of the satellite, and the Kalman filter includes the first state A state prediction unit that calculates a covariance matrix of a prediction value of a number vector and a prediction error, and a second state that includes only the user speed and the receiver clock drift among the calculation results of the state prediction unit as state variables A state estimation unit that calculates an estimated value of the second state variable vector and a covariance matrix of the estimation error based on a prediction value of the variable vector and a covariance matrix of the prediction error and the observation data; and the state prediction unit A receiver clock drift predicted value included in the calculated predicted value of the first state variable vector, and a receiver clock drift estimated value included in the estimated value of the second state variable vector calculated by the state estimating means; Difference calculating means for calculating the difference between the receiver and the receiver clock included in the covariance matrix of the prediction error calculated by the state prediction means when the difference exceeds a predetermined threshold Dispersion correcting means for correcting the variance of the lift to a larger value, and the reception included in the covariance matrix of the prediction error calculated by the state predicting means when the dispersion correcting means corrects the variance of the receiver clock drift The variance of the machine clock drift is replaced with the modified variance of the receiver clock drift to calculate the estimated value of the first state variable vector and the covariance matrix of the estimation error and update the calculation result of the state prediction means When the variance correction means does not correct the variance of the receiver clock drift, the estimated value of the first state variable vector and the covariance of the estimation error based on the calculation result of the state prediction means and the observation data A state estimation updating unit that calculates a matrix and updates a calculation result of the state prediction unit.

  With this configuration, the positioning apparatus of the present invention calculates the estimated value of the second state variable vector including only the user speed and the receiver clock drift as the state variables and the covariance matrix of the estimated error, and the state estimating means A receiver clock drift predicted value included in the predicted value of the first state variable vector calculated by the predicting means, and a receiver clock drift estimated value included in the estimated value of the second state variable vector calculated by the state estimating means; When the difference between the values exceeds a predetermined threshold, the variance of the receiver clock drift included in the covariance matrix of the prediction error calculated by the state prediction means is corrected to a larger value, so that a relatively large receiver clock Even if drift occurs, the current position can be measured in real time with high accuracy.

  Further, in the positioning device of the present invention, the receiver clock includes an oscillator having a lower stability than an oscillator included in the clock mounted on the satellite, and the dispersion correction unit includes a frequency short-term of the oscillator of the receiver clock. It has a configuration in which a value estimated by the stability is used as the threshold value.

  With this configuration, the positioning device of the present invention can measure the current position with high accuracy in real time even when using a low-stability oscillator typified by a temperature-compensated crystal oscillator, thereby reducing the manufacturing cost. Can be planned.

  The present invention can provide a positioning device having an effect that the current position can be measured with high accuracy in real time even when a relatively large receiver clock drift occurs.

It is a block block diagram in 1st Embodiment of the positioning signal receiver which concerns on this invention. It is a block block diagram in 1st Embodiment of the positioning apparatus which concerns on this invention. It is a flowchart in 1st Embodiment of the positioning apparatus which concerns on this invention. It is a block block diagram in 2nd Embodiment of the positioning apparatus which concerns on this invention. It is a flowchart in 2nd Embodiment of the positioning apparatus which concerns on this invention.

  First, Kalman filter processing will be described before describing an embodiment of the present invention.

(Kalman filter processing)
In the above [Equation 1], the user position p and the user speed v are respectively the user position p = [p _x , p _y , p when the x-axis, y-axis, and z-axis of the ECEF coordinate system are used. _z] ^T, user velocity _{v = [v x, v y} , denoted v ^{z] T.} Therefore, the state variable vector x shown in [Equation 1] includes 8 variables. Therefore, the error covariance matrix of the estimated state value ^ x is expressed by [Equation 2]. It should be noted that “^ x” indicates that a hat (^) indicating an estimated value is added to the upper part of x.

Here, each component in the matrix of the above equation is expressed by [Equation 3]. However, in [Expression 3], e _p is error for user position p, e _{Delta] t} is the error relates to the receiver clock offset Delta] t, e _v is the error regarding user velocity v, e _{[Delta] T} is an error regarding the receiver clock drift [Delta] T Yes, E {} means an expected value.

(K-1) A model equation (state equation) for predicting the state of k epoch from the estimated value of epoch is expressed by [Equation 4]. Here, e _{Δp (k)} , e _{ΔΔt (k)} , e _{Δv (k)} , and e _{ΔΔT (k)} are (k−1) epoch state estimates, and the state at k epochs is expressed as matrix A. This is a system noise vector representing the uncertainty between the predicted value and the actual value. I is a unit matrix.

The prediction process by the Kalman filter is expressed by [Equation 5]. Here, Q _{Δp (k)} and Q _{Δv (k)} are covariance matrices for e _{Δp (k)} and e _{Δv (k)} , respectively, and σ _{ΔΔt (k)} ² and σ _{ΔΔT (k)} ² are e It is dispersion about _{ΔΔt (k)} , e _{ΔΔT (k)} . Superscript (-) means a value obtained by the prediction process.

The relational expression (observation equation) between the observed quantity and the state estimated value is expressed by [Equation 6]. However, the superscript i (i = 1, 2,... N) of each variable indicates the satellite number. T _sv is the satellite transmission time of the satellite, t _user is the receiver time, φ _doppler is the Doppler observation value, p _sat (t _sv ) is the satellite when the signal at the satellite transmission time t _sv is acquired at the user position p. The position, I _on is the ionospheric delay error, T _r is the tropospheric delay, e _ρ is the observation error of the pseudorange ρ (ρ = c (t _sv −t _user )), and e _doppler is the observation error of the Doppler observation value.

The estimation process by the Kalman filter is expressed by [Equation 7]. However, K is the Kalman gain, P is the covariance matrix of the estimation error, R is the covariance matrix of the observation noise e _k.

In [Expression 7], the design matrix H is a Jacobian matrix _obtained by differentiating h _k (^ x _k ⁻ , 0) by x _k and is expressed by [Expression 8].

Here, c is the speed of light. H ⁱ _AOA is a unit direction vector from the satellite i to the receiver, and is represented by [Equation 9].

In order to accurately estimate the state of the system, for each of the prediction process and the estimation process, system noise covariance matrices Q _{Δp (k)} and Q _{Δv (k)} and variance values σ _ΔΔt that match the system characteristics are _{used. By providing (k)} ² , σ _{ΔΔT (k)} ^2, and the covariance matrix R _k of observation noise, the accuracy of estimation can be improved.

(First embodiment)
Next, the configuration of the positioning device according to the first embodiment of the present invention will be described.

  As shown in FIG. 1, the positioning signal receiver 1 in this embodiment includes a receiving unit 10, a positioning device 20, and a control unit 30. This positioning signal receiver 1 is mounted on a vehicle, for example.

  The receiving unit 10 includes an antenna that receives radio frequency (RF) band radio waves from a predetermined number (for example, four) of GPS satellites, a conversion circuit that down-converts an RF band signal into an intermediate frequency band signal, and the like. The down-converted received signal is output to the positioning device 20.

  The positioning device 20 is composed of, for example, a microcomputer, and analyzes an input received signal to measure the current position (user position) of the vehicle.

  The control unit 30 is configured by a microcomputer, for example, and controls operations of the receiving unit 10 and the positioning device 20 based on a program stored in advance in a memory.

  FIG. 2 is a block configuration diagram of the positioning device 20 in the first embodiment. As shown in FIG. 2, the positioning device 20 includes a reception signal analysis unit 21 that analyzes a reception signal, and a receiver clock 22 that provides reception time information to a reception signal analysis unit 21 and a state estimation processing unit 42 (described later). A Kalman filter 40 for estimating the user position, the user speed, and the receiver time. The receiver clock 22 in the present embodiment includes the TCXO as an oscillator, but is not limited thereto, and may be configured to include another oscillator. The received signal analyzer 21 constitutes observation data acquisition means according to the present invention.

  The Kalman filter 40 includes a state prediction processing unit 41 that performs state prediction, a state estimation processing unit 42 that performs state estimation, and a clock abnormality change detection unit 43 that detects abnormal changes in the receiver clock 22 in real time. . The clock abnormality change detection unit 43 constitutes a difference calculation unit according to the present invention.

  The state prediction processing unit 41 uses the state prediction processing unit 41 a that predicts the state variable vector of the current epoch based on the state variable vector one epoch before obtained from the state estimation processing unit 42, and the detection result of the clock abnormality change detection unit 43. A state prediction value correction unit 41b that dynamically corrects the state prediction value based on this. The state prediction processing unit 41a and the state prediction value correction unit 41b constitute a state prediction unit and a dispersion correction unit according to the present invention, respectively.

  The state estimation processing unit 42 receives observation data relating to the satellite from the received signal analysis unit 21, and also inputs the state variable vector at the current epoch and the covariance matrix of the prediction error from the state prediction processing unit 41, and obtains the state variable vector. It is updated and output to the state prediction processing unit 41a. The state estimation processing unit 42 constitutes state estimation update means according to the present invention.

  Next, the operation of the Kalman filter 40 of the positioning device 20 in this embodiment will be described.

  The reception signal analysis unit 21 analyzes the input reception signal and obtains information on the satellite transmission time, satellite position, and satellite speed of each satellite. The received signal analyzer 21 calculates a Doppler observation value based on the received carrier wave of the radio wave. Then, the received signal analysis unit 21 passes the calculated result to the state estimation processing unit 42.

  Hereinafter, the operation of the Kalman filter 40 will be described with reference to FIG.

The state prediction processing unit 41a of the state prediction processing unit 41 calculates the user of the current epoch from the covariance matrix of the user position, user speed, and estimation error one epoch before (k-1) obtained from the state estimation processing unit 42. The position and the user speed are predicted, and the covariance matrix P _k ⁻ of the prediction error is calculated using [Equation 5] (step S11).

  Here, the amount of change in the receiver clock drift in the model is approximately the value estimated by the short-term stability of the TCXO frequency when the receiver clock drift is stable. Note that the short-term frequency stability means a fluctuation (irregular fluctuation) in the output frequency of the TCXO within a relatively short time period.

The state estimation processing unit 42 inputs data such as satellite transmission time, satellite position, satellite speed, Doppler observation value, and reception time from the reception signal analysis unit 21. Further, the state estimation processing unit 42 inputs data of the state variable vector x _k ⁻ and the covariance matrix P _k ⁻ of the current epoch predicted by the state prediction processing unit 41. And the state estimation process part 42 calculates | requires a Kalman gain by [Equation 7] using these input data, and estimates a state variable vector and a covariance matrix (step S12). However, when there is a correction by the state prediction value correction unit 41b, the state estimation processing unit 42 receives the data of the covariance matrix P _k ⁻ in which the variance value of the receiver clock drift is corrected from the state prediction value correction unit 41b. input.

The clock abnormality change detection unit 43 includes the receiver clock drift Δ ^ T _k included in the state variable vector ^ x _k estimated by the state estimation processing unit 42 and the state variable vector ^ x _k ⁻ predicted by the state prediction processing unit 41a. The receiver clock drift Δ ^ T _k ⁻ included in is input from the state estimation processing unit 42. Further, the clock abnormality change detection unit 43 calculates a difference (| Δ ^ T _k −Δ ^ T _k ⁻ |) between the two changes, and the reception estimated by the short-term frequency stability of the TCXO by [Equation 10]. By comparing with the standard deviation σ _ΔΔT of the machine clock drift, it is determined whether or not the receiver clock 22 has changed abnormally (step S13). Α is a threshold coefficient that determines the allowable range.

When the change in the receiver clock drift does not satisfy the number [10], the clock abnormality change detection unit 43 determines that an abnormal change in the receiver clock 22 has occurred, and a value σ greater than σ _ΔT ^{2 −} _newΔT ²⁻ is output to the state prediction value correction unit 41b as a correction value of the variance of the receiver clock drift. Note that α and σ _newΔT ²⁻ are determined in advance based on, for example, results of experiments and simulations.

  On the other hand, when the change in the receiver clock drift satisfies the number [10], the clock abnormality change detection unit 43 determines that an abnormal change in the receiver clock 22 has not occurred and ends the process.

The state predicted value correction unit 41b receives the receiver clock drift variance correction value σ _newΔT ^2- from the clock abnormality change detection unit 43. Then, the state prediction value correction unit 41 b receives the receiver clock drift variance σ _{ΔT (k)} ²⁻ in the prediction error covariance matrix P _k ⁻ received from the state prediction processing unit 41 a from the clock abnormality change detection unit 43. By replacing the received receiver clock drift variance correction value σ _newΔT ²⁻ , the covariance matrix P _k ⁻ of the prediction error is corrected (step S14).

In step S12, the state estimation processing unit 42 performs the estimation process again based on the corrected covariance matrix P _k ⁻ of the prediction error.

  As described above, according to the positioning device 20 in the present embodiment, when the difference between the receiver clock drift predicted value and the receiver clock drift estimated value exceeds a predetermined threshold, the prediction error covariance matrix is calculated. Since the dispersion of the included receiver clock drift is dynamically corrected to a larger value, the current position can be measured with high accuracy in real time even when a relatively large receiver clock drift occurs.

(Second Embodiment)
A second embodiment of the positioning device according to the present invention will be described.

  As shown in FIG. 4, the positioning device 50 in this embodiment is obtained by changing the Kalman filter 40 of the positioning device 20 (see FIG. 2) in the first embodiment to a Kalman filter 60. Therefore, the configuration related to the Kalman filter 60 will be described, and the description of the configuration overlapping with the description of the first embodiment will be omitted.

  The Kalman filter 60 includes a state prediction processing unit 61 that performs state prediction, a state estimation processing unit 62 that acquires reception time information from the receiver clock 22 and performs state estimation, and a clock abnormality that detects an abnormal change in the receiver clock 22. A change detection unit 63. The state prediction processing unit 61 includes a state prediction processing unit 61a that predicts a state, and a state prediction value correction unit 61b that corrects the state prediction value based on the detection result of the clock abnormality change detection unit 63.

  Next, operation | movement of the Kalman filter 60 of the positioning apparatus 50 in this embodiment is demonstrated based on FIG.

  The state prediction processing unit 61a predicts the user position and user speed of the current epoch from the covariance matrix of the user position and user speed one epoch before obtained from the state estimation processing unit 62, and the estimation error. Is calculated using [Equation 5] (step S21). Here, the fluctuation amount of the receiver clock drift in the model is about the value estimated by the short-term stability of the frequency of the TCXO when the receiver clock drift is stable.

The clock abnormality change detection unit 63 receives the user speed and the receiver clock drift (v _x , v _y , v _{z) from} among the state variables included in the predicted state variable vector x _k-1 from the state prediction processing unit 61a. , ΔT) and a covariance matrix relating to a user variable and a state variable vector of a receiver clock drift, which are partial matrices of the prediction error covariance matrix P _k−1 ⁻ . This covariance matrix is shown in [Equation 11].

  Further, the clock abnormality change detecting unit 63 sets the state variable vector as [Equation 12], sets the observation equation as [Equation 13], and performs [Equation 14] to perform the state estimation processing for detecting the clock abnormality as the state estimation processing unit 62. (Step S22). In this case, the state estimation processing unit 62 constitutes a state estimation unit according to the present invention.

Based on the receiver clock drift Δ ^ T _k estimated in step S22 and the receiver clock drift Δ ^ T _k ⁻ predicted by the state prediction processing unit 61a, the clock abnormality change detection unit 63 determines the difference between the two changes. (| Δ ^ T _k −Δ ^ T _k ⁻ |) is calculated and compared with the standard deviation σ _ΔΔT of the receiver clock drift estimated by the short-term frequency stability of the TCXO by [ _Equation 15]. It is determined whether or not the clock 22 has changed abnormally (step S23).

When the change in the receiver clock drift does not satisfy [Equation 15], the clock abnormality change detection unit 63 determines that an abnormal change in the receiver clock 22 has occurred, and has a value σ greater than σ _ΔT ^{2 −.} _newΔT ²⁻ is output to the state prediction value correction unit 61b as a correction value of the variance of the receiver clock drift.

The state prediction value correction unit 61b receives from the clock abnormality change detection unit 63 the receiver clock drift variance σ _{ΔT (k)} ²⁻ in the covariance matrix P _k ⁻ of the prediction error received from the state prediction processing unit 61a. The covariance matrix P _k ⁻ of the prediction error is corrected by replacing it with the correction value σ _newΔT ²⁻ of the receiver clock drift variance (step S24).

On the other hand, if the change in the receiver clock drift satisfies [Equation 15], the clock abnormality change detection unit 63 determines that no abnormal change in the receiver clock 22 has occurred, and the value predicted in step S21. Then, the state estimation processing unit 62 is caused to perform state estimation by [Equation 7] (step S25). When the prediction error covariance matrix P _k ⁻ is corrected in step S24, the state estimation processing unit 62 performs an estimation process using the corrected prediction error covariance matrix P _k ⁻ in step S25. In this case, the state estimation processing unit 62 constitutes state estimation update means according to the present invention.

Instead of the dispersion of the correction of the receiver clock drift described above, for example, the following equation 16] The covariance matrix of the prediction error by changing the sigma _{[Delta] T} ² scores sigma _NewderutaT in ² in Equation 11] A method for calculating a value satisfying [Equation 17] may be used. In this case, σ _newΔT ²⁻ is calculated by [ _Equation 18] using σ _newΔΔT ² .

Here, in the matrix on the right side of [Equation 16], the predicted value at k epoch is not used as it is for the covariance matrix related to the state variable vector of velocity v and receiver clock drift ΔT which is a submatrix of P _k ^−. You can change it if necessary.

  As described above, according to the positioning device 50 in this embodiment, the estimated value of the state variable vector including only the user speed and the receiver clock drift as the state variables and the covariance matrix of the estimated error are calculated, and this calculation is performed. A threshold value in which a difference between a receiver clock drift estimated value included in the estimated value of the state variable vector and a receiver clock drift predicted value included in the predicted value of the state variable vector calculated by the state prediction processing unit 61 is determined in advance. When exceeded, the receiver clock drift variance included in the covariance matrix of the prediction error calculated by the state prediction processing means 61 is corrected to a larger value, so that a relatively large receiver clock drift occurred. Even in this case, the current position can be measured in real time with high accuracy.

  As described above, the positioning device according to the present invention has the effect that the current position can be measured in real time with high accuracy even when a relatively large receiver clock drift occurs, and satellite navigation using the Kalman filter It is useful as a system positioning device.

10 receiving unit 20, 50 positioning device 21 received signal analyzing unit (observation data acquiring means)
22 receiver clock 30 control unit 40, 60 Kalman filter 41, 61 state prediction processing unit 41a, 61a state prediction processing unit (state prediction unit)
41b, 61b State predicted value correction unit (dispersion correction means)
42 State estimation processing unit (state estimation update means)
43, 63 Clock abnormality change detector (difference calculating means)
62 State estimation processing unit (state estimation means, state estimation update means)

Claims (3)

A positioning device for positioning a user of a receiver that receives a positioning signal from a satellite in a satellite navigation system,
A receiver clock indicating a difference between a user position indicating the position of the user, a user speed indicating the moving speed of the user, a time of a receiver clock included in the receiver and a time in the satellite navigation system. A Kalman filter for performing state prediction and state estimation for the satellite navigation system based on a state variable vector including an offset and a receiver clock drift indicating a shift of the receiver clock offset per unit time in the receiver clock as a state variable When,
Observation data acquisition means for receiving a positioning signal from the satellite and acquiring observation data indicating the state of the satellite; and
The Kalman filter is
State prediction means for calculating a covariance matrix of predicted values and prediction errors of the state variable vectors;
A state estimation updating unit that calculates a covariance matrix of the estimation value and estimation error of the state variable vector based on the calculation result of the state prediction unit and the observation data, and updates the calculation result of the state prediction unit;
The difference between the receiver clock drift prediction value included in the predicted value of the state variable vector calculated by the state prediction unit and the receiver clock drift estimation value included in the estimated value of the state variable vector calculated by the state estimation update unit Difference calculating means for calculating
Dispersion correction means for correcting the variance of the receiver clock drift included in the prediction error covariance matrix to a larger value when the difference exceeds a predetermined threshold;
The state estimation updating means corrects the receiver clock drift variance included in the prediction error covariance matrix calculated by the state prediction means when the variance correction means corrects the receiver clock drift variance. A positioning apparatus characterized by recalculating the covariance matrix of the estimated value and estimated error of the state variable vector in place of the receiver clock drift variance. A positioning device for positioning a user of a receiver that receives a positioning signal from a satellite in a satellite navigation system,
A receiver clock indicating a difference between a user position indicating the position of the user, a user speed indicating the moving speed of the user, a time of a receiver clock included in the receiver and a time in the satellite navigation system. State prediction and state estimation for the satellite navigation system based on a first state variable vector including an offset and a receiver clock drift indicating a shift of the receiver clock offset per unit time in the receiver clock as a state variable A Kalman filter that performs
Observation data acquisition means for receiving a positioning signal from the satellite and acquiring observation data indicating the state of the satellite; and
The Kalman filter is
State prediction means for calculating a covariance matrix of predicted values and prediction errors of the first state variable vector;
Based on the prediction value of the second state variable vector including only the user speed and the receiver clock drift among the calculation results of the state prediction means as the state variables, the covariance matrix of the prediction error, and the observation data State estimation means for calculating a covariance matrix of the estimated value and estimation error of the second state variable vector;
Receiver clock drift predicted value included in the predicted value of the first state variable vector calculated by the state predicting means, and receiver included in the estimated value of the second state variable vector calculated by the state estimating means A difference calculating means for calculating a difference from the clock drift estimated value;
Dispersion correction means for correcting the variance of the receiver clock drift included in the covariance matrix of the prediction error calculated by the state prediction means when the difference exceeds a predetermined threshold;
When the dispersion correction means corrects the dispersion of the receiver clock drift, the dispersion of the receiver clock drift included in the covariance matrix of the prediction error calculated by the state prediction means is corrected to the dispersion of the receiver clock drift. Replacing the calculation value of the first state variable vector and the covariance matrix of the estimation error to update the calculation result of the state prediction means, and the dispersion correction means corrects the variance of the receiver clock drift. When there is no state, a state in which the estimated value of the first state variable vector and the covariance matrix of the estimated error are calculated based on the calculation result of the state prediction unit and the observation data, and the calculation result of the state prediction unit is updated A positioning apparatus comprising: an estimation updating means; The receiver clock includes an oscillator having a lower stability than an oscillator included in the clock mounted on the satellite,
The positioning apparatus according to claim 1, wherein the dispersion correcting unit uses a value estimated based on a short-term frequency stability of an oscillator of the receiver clock as the threshold value.

Images