JP2009236517A - Gps receiver - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a GPS receiver that effectively avoids lowering of the accuracy of a subsequent measured location even when the accuracy of geolocation of the first time is low. <P>SOLUTION: The GPS receiver includes a current location estimating means for estimating a first range wherein the GPS receiver can be located currently, based on the results of GPS geolocation, a past location estimating means for estimating a second range wherein the GPS receiver could be located at a certain time point in the past, based on the first range, and a past measured location determining means which determines whether a past measured location calculated by the GPS geolocation at the time point in the past is included in the second range. The GPS receiver is so constituted that the past measured location is not used for calculation of an estimated location in the case when the past measured location is not included in the second range. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a GPS receiver that performs GPS positioning using a GPS (Global Positioning System) signal. Specifically, the present invention relates to a GPS receiver that outputs an estimated position calculated based on a result of the GPS positioning to the outside.

  GPS is a positioning system for measuring position information and the like of an object using GPS signals transmitted from GPS satellites orbiting the earth, and is widely used in daily life.

  In GPS, a plurality of GPS satellites share the same frequency band and transmit a GPS signal by a CDMA (Code Division Multiple Access) method. That is, the GPS signal is 1575.42 MHz by a navigation message as a communication signal including GPS satellite orbit information and clock information and a PRN code (Pseudo Random Noise code) as a spread code determined for each GPS satellite. The carrier is obtained by BPSK (Binary Phase Shift Keying) modulation. The GPS receiver selects the number of GPS satellites more than necessary in principle, generates a PRN code corresponding to the selected GPS satellite, and multiplies the phase of the generated PRN code in synchronization with the GPS signal to multiply the navigation data. Is demodulated. And it calculates using the navigation data obtained by demodulating, and obtains position information, moving speed, etc. of the object. Hereinafter, position measurement, movement speed measurement, and the like using a GPS signal by a GPS receiver will be referred to as “GPS positioning”.

  An end-user product including a GPS receiver includes, for example, a car navigation device mounted on a passenger car. In addition to GPS positioning, the car navigation system also performs well-known DR positioning using a DR (dead reckoning) sensor, and the final positioning result is obtained from the positioning results of both systems so as to complement the weaknesses of the two positioning methods. Calculate. And based on the positioning result obtained by calculation, the own vehicle position displayed on the screen is updated. When the destination is set by the user, the car navigation device performs navigation toward the destination together with the update of the vehicle position on the screen.

  In such a car navigation apparatus, the GPS satellite that can be used by the GPS receiver for GPS positioning changes every moment due to the shielding of the GPS signal by a building or the like. For this reason, the geometrical arrangement of the GPS satellites used for GPS positioning varies, and the error included in the GPS positioning solution fluctuates. As shown in FIG.

  FIG. 1A is a diagram for explaining variation in measurement positions due to GPS positioning. Line A in FIG. 1 (a) shows the actual travel trajectory of the passenger car, each plot marked with ● indicates a measurement position obtained by GPS positioning executed at a predetermined cycle, and line A ′ shows each measurement position in time series. Shows the connected trajectory. As shown in FIG. 1 (a), the positioning error fluctuates due to variations in the geometrical arrangement of GPS satellites used for positioning. Therefore, a trajectory A ′ connecting the measurement positions in time series is a passenger car. It can be seen that the actual travel locus A greatly varies.

  In a car navigation device or the like, it is not preferable to display such an irregularly trajectory on the screen. For this reason, a general GPS receiver has a configuration for calculating an estimated value obtained by statistically removing an error component from a GPS positioning result using a Kalman filter. The GPS receiver corrects and outputs the GPS positioning result based on the estimated value obtained by the Kalman filter integrating information such as past GPS positioning results and measurement conditions. FIG. 1B is a diagram showing the relationship between the same traveling locus A as in FIG. 1A and the locus A ″ that connects the measurement positions corrected based on the estimated values by the Kalman filter in time series. 1B, the error between the actual travel locus A and the corrected measurement position is reduced by the Kalman filter process, and the measurement positions adjacent in time are continuous. Since the position is corrected, the trajectory connecting the corrected measurement positions in time series draws a smooth trajectory.

  Here, for example, when a passenger car enters a tunnel, a GPS signal from a GPS satellite is shielded by the tunnel, so the GPS receiver cannot perform GPS positioning. If such a non-positioning state of the GPS receiver continues for a long time, for example, a change in the moving distance and direction from the previous measurement position becomes large, so that the accuracy of the estimated value of the Kalman filter decreases and the error of the GPS positioning result increases. For this reason, the GPS receiver is configured to reset the estimated value of the Kalman filter when the non-positioning state continues for a certain time or more.

  Further, when the passenger car exits from the tunnel, the GPS receiver resumes GPS positioning because there is no element that blocks the GPS signal. At this time, the measurement position by the GPS positioning may jump greatly.

  For example, in the first GPS positioning (hereinafter referred to as “initial positioning”) performed when returning from a long-time non-positioning state to a positioning ready state, the carrier frequency and the phase of the PRN code fluctuate during non-positioning. When the values used in GPS positioning before entering the non-positioning state (various setting values used when capturing the GPS satellites) are used, it is difficult to re-acquire the GPS satellites. For this reason, the GPS receiver sets a wider search range for the carrier frequency and the phase of the PRN code when reacquiring the GPS satellite. For this reason, the GPS receiver takes time to capture the GPS satellites, and it may not be possible to acquire a sufficient number of GPS satellites in the first positioning, and it may not be possible to calculate accurate GPS positioning results. At this time, the measurement position greatly jumps.

  In addition, if the Kalman filter estimated value is reset after a non-positioning state has continued for a certain period of time, the initial value is set for the Kalman filter based on the initial positioning result with low accuracy, and the subsequent estimated values are used for the initial positioning with low accuracy. It will be estimated based on the result. For this reason, the accuracy of the estimated value of the Kalman filter is significantly reduced immediately after the passenger car leaves the tunnel. Since the Kalman filter is designed to calculate a new estimated value using a past estimated value, once the estimated value includes a large error, there is a drawback that it takes time until the error is removed.

  FIG. 2 is a diagram for explaining a problem when the accuracy of the estimated value of the Kalman filter is lowered. Line A in FIG. 2 shows the same traveling trajectory as in FIG. 1A, etc., line B shows the boundary between the tunnel and the outside of the tunnel, plot C shows the measurement position obtained by the first positioning, and each plot marked with ● Indicates each measurement position after the second time when not processed by Kalman filter (measurement position obtained by GPS positioning after the next (second time) of the first positioning), and each plot marked with × is when processed by Kalman filter The measurement position after the second time is shown. As shown in FIG. 2, once the accuracy of the estimated value of the Kalman filter is once deteriorated due to the measurement position C greatly jumped, it takes time until the measurement position processed by the Kalman filter converges to a good value (near the straight line A). It turns out that it takes.

  Regarding techniques for promptly performing initial positioning, Patent Documents 1 to 3 disclose several methods. The GPS receiver described in Patent Document 1 monitors the synchronization level between a carrier and a signal generated inside the GPS receiver, and determines whether the GPS signal is shielded by a shield such as a tunnel. If it is determined that the GPS signal is blocked, searching for another GPS satellite is prohibited as an alternative. Further, the GPS receiver described in Patent Document 2 preferentially searches for GPS satellites captured in the past when detecting the interruption of the GPS signal. For this reason, in the GPS receiver described in Patent Documents 1 and 2, it is not necessary to search for a new GPS satellite by expanding the search range when passing through a tunnel, for example, and prompt initial positioning is achieved.

Further, the GPS receiver described in Patent Document 3 calculates a frequency shift amount of a reference frequency oscillator inside the GPS receiver based on the carrier frequency. Then, based on the calculated frequency deviation amount, the search range of the carrier frequency and the phase of the PRN code is set, and the next search operation is performed. The GPS receiver described in Patent Document 3 achieves quick initial positioning by setting an appropriate search range in advance.
JP-A-8-201503 JP-A-8-211142 Japanese Patent Laid-Open No. 10-268025

  However, none of Patent Documents 1 to 3 raises the problem described above, that is, the problem that the accuracy of the estimated value of the Kalman filter is significantly lowered when the measurement position by the initial positioning jumps. There has been no study on means for solving the problem.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a GPS receiver capable of effectively avoiding a subsequent decrease in accuracy of the measurement position even when the accuracy of the initial positioning is low. Is to provide.

  A GPS receiver according to an embodiment of the present invention that solves the above problem performs GPS positioning using a GPS signal, calculates an estimated position based on a measurement position obtained by the GPS positioning, and outputs the estimated position to the outside. A current position estimating means for estimating a first range in which the GPS receiver can be currently located based on a result of GPS positioning, and a GPS receiver at a past time based on the first range. Past position estimation means for estimating a second range where the position can be located, and past measurement position determination for determining whether or not a past measurement position calculated by GPS positioning at a certain past time is included in the second range And a past measurement position is not used for calculation of the estimated position when the past measurement position is not included in the second range.

  According to the GPS receiver configured in this way, when the accuracy of the initial positioning is low and the position jumps, the Kalman filter effectively avoids the accuracy of the estimated value from being affected by the initial positioning result. Can do. Thereby, the measurement position by the GPS positioning after the first positioning is not affected by the first positioning, and the accuracy does not deteriorate due to the influence.

  The GPS receiver includes a DOP comparison means for comparing a DOP (Dilution of Precision) in the current GPS positioning and a DOP in the GPS positioning at a past time when the past measurement position is not included in the second range. It is good also as composition provided further. The GPS receiver configured as described above does not use the past measurement position for calculating the estimated position when the DOP in the current GPS positioning is better.

  The GPS receiver may further include a DOP determination unit that determines whether or not the DOP in the current GPS positioning is equal to or less than a specified value when the past measurement position is not included in the second range. . The GPS receiver configured in this manner does not use the past measurement position for calculation of the estimated position when the DOP in the current GPS positioning is equal to or less than the specified value.

  The current position estimating means may be configured to estimate the first range based on a geometric arrangement of GPS satellites used at the time of GPS positioning and a pseudorange error obtained from the GPS positioning result.

  The past position estimating means may be configured such that the second range is obtained by shifting the first range based on the measurement direction and the measurement speed included in the positioning result by the current GPS positioning.

  Here, the above-mentioned certain point in the past is, for example, a point in time when the GPS positioning is resumed from a state where GPS positioning cannot be continued for a certain period of time or longer.

  In the GPS receiver, the estimated position may be calculated using a Kalman filter. When the past measurement position is not included in the second range, the GPS receiver configured as described above resets the Kalman filter so that the past measurement position is not used in the estimation calculation using the Kalman filter.

  According to the GPS receiver of the present invention, even if the accuracy of the first positioning is low and the measurement position jumps, it is possible to effectively avoid the subsequent decrease in accuracy of the estimated value of the Kalman filter.

  Hereinafter, the configuration and operation of a car navigation device according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 3 is a block diagram illustrating a configuration of the navigation device 200. As shown in FIG. 3, the navigation device 200 includes a GPS receiver 100, a gyro sensor 102, a vehicle speed sensor 104, A / D conversion units 102 a and 104 a, a CPU (Central Processing Unit) 108, a ROM (Read Only Memory) 110, A random access memory (RAM) 112, a hard disk drive (HDD) 114, a display unit 116, and an input unit 118 are provided. The CPU 108 controls the entire navigation device 200 and implements various functions through cooperative operations with each component of the navigation device 200.

  The GPS receiver 100 captures and tracks more than the number of GPS signals necessary for GPS positioning from a plurality of GPS satellites, performs GPS positioning, and outputs the obtained GPS positioning results to the CPU 108. GPS positioning by the GPS receiver 100 is performed once per second, for example.

  The gyro sensor 102 and the vehicle speed sensor 104 are well-known DR sensors. The gyro sensor 102 measures the angular velocity related to the direction in the horizontal plane of the vehicle on which the car navigation device 200 is mounted, and outputs the measurement result to the A / D conversion unit 102a. Moreover, the vehicle speed sensor 104 detects the rotational speed of the left and right drive wheels of the vehicle, and outputs the number of pulses corresponding to the detection result to the A / D converter 104a. The signals input to the A / D converters 102a and 104a are converted into digital signals and then input to the CPU 108.

  The ROM 110 stores various programs for executing navigation processing. These programs are loaded into the work area of the RAM 112 when the navigation device 200 is activated. The CPU 108 executes the program on the work area and appropriately reads the map data stored in the HDD 114 and causes the display unit 116 to display the map.

  Specifically, the CPU 108 performs DR positioning calculation based on the output of each DR sensor, and acquires the DR positioning result. Next, the obtained DR positioning result and the GPS positioning result of the GPS receiver 100 are compared with each other in consideration of the estimated error value for each positioning result. Then, a positioning result determined to be highly accurate is selected based on the comparison result, and the selected positioning result is determined as a final positioning result (hereinafter referred to as “final positioning result”). If there is no output of the GPS positioning result, the DR positioning result is set as the final positioning result without performing the above comparison process.

  Then, the CPU 108 searches a map database stored in the HDD 114 based on the determined positioning result, and extracts a corresponding map image. Finally, map matching is performed, and the vehicle position mark is superimposed on the extracted map image and displayed on the display unit 116. At this time, if the user operates the input unit 118 to set a destination, the CPU 108 also performs navigation processing for the destination.

  Next, the GPS receiver 100 will be described in detail. FIG. 4 is a block diagram showing the configuration of the GPS receiver 100. As shown in FIG. 4, the GPS receiver 100 includes a down converter unit 1, a received signal processing unit 2, and a positioning calculation control unit 3. The down-converter unit 1 receives the GPS signal and converts it to an intermediate frequency, and then outputs it to the received signal processing unit 2. The reception signal processing unit 2 and the positioning calculation control unit 3 operate in cooperation to execute GPS signal acquisition, tracking, and positioning processes using the signal output from the down converter unit 1. The positioning calculation control unit 3 outputs the GPS positioning result obtained by the cooperative operation with the received signal processing unit 2 to the CPU 108.

  The down converter unit 1 includes a GPS antenna 10, an RF (radio frequency) input unit 11, a BPF (Band Pass Filter) 12 and 14, an LNA (Low Noise Amplifier) 13, a down converter 15, an AGC (Auto Gain Control) 16, a TCXO. (Temperature Compensated Crystal Oscillator) 17, frequency synthesizer 18, and A / D converter 19.

  A GPS signal transmitted from a GPS satellite is received by the GPS antenna 10 and input to the BPF 12 via the RF input unit 11. Next, the input signal (hereinafter referred to as “RF signal”) is filtered into a predetermined band by the BPF 12, and after the noise outside the GPS band is attenuated through the LNA 13 and the BPF 14 which are low noise amplifiers, the down converter 15. To enter.

  The TCXO 17 is a local oscillator that oscillates at a predetermined frequency, and the oscillation frequency is lower than the frequency of the RF signal input to the down converter 15. Further, the frequency synthesizer 18 generates a local oscillator signal based on the input from the TCXO 17 and outputs it to the down converter 15. The down converter 15 converts the RF signal into an intermediate frequency advantageous for signal processing, that is, an IF (Intermediate Frequency) signal under the control of the AGC 16 using the input local oscillator signal and outputs it to the A / D converter 19. To do.

  The A / D converter 19 samples the IF signal, performs quadrature demodulation, and converts it into an I (In-phase) signal and a Q (Quadra-phase) signal. The I signal is an in-phase component at the time of quadrature demodulation, and the Q signal is a component that is orthogonal to the I signal. Hereinafter, for convenience of explanation, the I signal and the Q signal are collectively referred to as an “IQ signal”. The A / D converter 19 outputs the IQ signal obtained by the conversion process to the received signal processor 2.

  The reception signal processing unit 2 includes a channel 21 and an NCO (Number Controlled Oscillator) 22. The reception signal processing unit 2 is configured to include a plurality of channels 21 so that GPS signals from a plurality of GPS satellites can be simultaneously captured and tracked by the CDMA method.

  The positioning calculation control unit 3 includes a CPU 31, an RTC (Real-Time Clock) 32, a ROM 33, a RAM 34, and an interface 35. The CPU 31 operates based on a clock output from the frequency synthesizer 18 and performs overall control of the reception signal processing unit 2 and the positioning calculation control unit 3. The RTC 32 is a clock IC (Integrated Circuit) that is operated by a crystal oscillator (not shown), and functions as a time measuring means. The ROM 33 stores a program (for example, the above-described Kalman filter) and data for performing a positioning calculation. The program is loaded into the work area of the RAM 34 when the navigation device 200 is activated. The CPU 31 operates according to a program loaded in the work area of the RAM 34, controls the NCO 22, and performs a positioning calculation based on a signal from each channel 21. Then, the GPS positioning result obtained by the calculation is output to the CPU 108 via the interface 35.

  Specifically, the CPU 31 determines the Doppler shift amount of the IQ signal input to each channel 21 based on the navigation message acquired by the previous positioning, the previous GPS positioning result, the current time measured by the RTC 32, and The search range of the phase of the PRN code necessary for capturing the GPS signal is estimated, and the Doppler shift amount and the set value of the search range are generated based on the estimation result and output to the NCO 22. The CPU 31 designates a PRN code to be searched for in each channel 21 through the NCO 22 for each channel 21 so that GPS signals of different GPS satellites are captured and tracked in each channel 21.

  The NCO 22 operates based on the clock output from the frequency synthesizer 18, generates a reference code for the PRN code based on the set value from the CPU 31, and performs necessary data exchange with each channel 21 to set each channel 21. Control.

  Each channel 21 performs Doppler shift compensation on the IQ signal from the A / D converter 19 based on the Doppler shift amount set by the CPU 31 under the control of the NCO 22. Similarly, the PRN code included in the GPS signal and the reference code (that is, generated on the GPS receiver 100 side) are multiplied and multiplied by the PRN code reference code and the IQ signal within the search range set by the CPU 31. A correlation value with the PRN code is obtained. Each channel 21 captures the IQ signal, that is, the GPS signal, when the obtained correlation value exceeds a predetermined threshold value. As the integration processing time at this time is set longer, the influence of noise included in the GPS signal can be removed, and the capture sensitivity increases. This integration processing time is also set by the CPU 31.

  The CPU 31 detects the phase difference between the IQ signal captured by each channel 21 and the reference code based on the correlation value obtained by each channel 21, and synchronizes the phase of the reference code with the phase of the IQ signal. The IQ signal is demodulated to obtain navigation data. The CPU 31 performs feedback control corresponding to each channel 21, and performs feedback control of the NCO 22 so that the output value of each channel 21 is maximized, thereby tracking the GPS signal captured by each channel 21. Note that if the acquisition and tracking of the GPS signal fails, the CPU 31 retry the acquisition and tracking of the GPS signal after resetting the search range. When the total number of GPS signals captured and tracked on the channel 21 is greater than or equal to the number necessary for GPS positioning, the CPU 31 receives data necessary for positioning including navigation data from each channel 21 to determine the positioning and speed. Measurement is performed, each measurement result is subjected to a Kalman filter, a GPS positioning result is obtained and output to the CPU 108. As described above, the GPS receiver 100 performs a positioning calculation once every second and outputs the obtained GPS positioning result to the CPU 108.

  When a GPS positioning result obtained through such processing (hereinafter referred to as “GPS positioning calculation processing”) is output from the GPS receiver 100 to the CPU 108, the CPU 108 uses the GPS positioning result, the DR positioning result, and the like. To determine the final positioning result.

  The above is the basic operation of the GPS positioning performed by the GPS receiver 100. However, the GPS receiver 100 has the above-described problem, that is, the problem that the accuracy of the subsequent measurement position deteriorates when the measurement position by the first positioning jumps. In order to solve the problem, the GPS measurement position improvement process shown in the flowcharts of FIGS. Hereinafter, the GPS measurement position improvement processing according to the embodiment of the present invention will be described with reference to the flowcharts of FIGS. In the flowcharts of FIGS. 5 to 7 and the following description, the step is abbreviated as “S”.

  As shown in FIG. 5, every time the CPU 31 of the GPS receiver 100 performs the above-described GPS positioning (here, processing before applying the measurement result to the Kalman filter) (S1), the non-positioning state continues for a certain time or more. It is determined whether or not (S2). For example, when GPS positioning fails continuously and the non-positioning state continues for a certain time or longer (S2: YES), the Kalman filter has a large change in the moving distance and direction of the passenger car from the position measured in the past to the current position. The accuracy of the estimated value decreases. For this reason, the CPU 31 turns on the post filter reset flag in order to reset the estimated value of the Kalman filter in the post filter reset process of S33 described later before executing the next GPS positioning calculation process (S3). Further, in order to execute a processing routine for determining whether or not to reset the estimated value of the Kalman filter, the post-filter reset determination execution flag is turned on (S4), and the GPS measurement position improvement processing proceeds to S11 in FIG. If the duration of the non-positioning state is less than the predetermined time (S2: NO), the CPU 31 advances the process to S11 in FIG. 6 without executing the flag setting process of S3 and S4. In the initial setting immediately after the power is turned on, the post filter reset flag is on and the post filter reset determination execution flag is off. These flags are set to initial values by reset processing immediately after power-on.

  In the process of S11 of FIG. 6, if the post filter reset determination execution flag is on (S11: YES), the CPU 31 executes the processes after S12 of FIG. If the post filter reset determination execution flag is off (S11: NO), the process proceeds to S31 in FIG. 7 without executing the processes after S12 in FIG.

  In the process of S12, the CPU 31 returns to the positioning-possible state from the non-positioning state for a certain time or longer (hereinafter referred to as “after the positioning-possible state return”), and performs GPS positioning (hereinafter, “measurement position POS_cur”). It is determined whether or not it is required. (In the process of S12, it is not determined whether or not “measurement position POS_pre” described later is obtained). If the measurement position POS_cur is obtained (S12: YES), it is determined whether the position is obtained by GPS positioning after the return to the positioning enabled state and before the measurement position POS_cur is obtained (S13). ).

  If the position is not obtained by GPS positioning before the measurement position POS_cur is obtained (S13: NO), the measurement position POS_cur is the measurement position obtained by the initial positioning. In this case, the CPU 31 stores the measurement position POS_cur in a cache memory (not shown) as the measurement position POS_pre obtained by the initial positioning (S14). Note that the measurement positions that can be held in the cache memory are limited to the most recent measurement position due to the capacity of the cache memory.

  Next, the CPU 31 sets a counter for the time limit t_jdg to an initial value (S15), and proceeds to S31 in FIG. The time limit t_jdg is a time limit for determining whether or not the characteristic processing of the GPS measurement position improvement processing (processing such as S21 in FIG. 6) can be performed.

  Further, when the position has not been obtained by GPS positioning after returning to the positioning enabled state (that is, neither of the measurement positions POS_cur and POS_pre has been obtained) (S12: NO, S16: NO), the CPU 31 is in a non-positioning state. Since the process is still continued, the process proceeds to S31 in FIG. 7 without executing the process in FIG. When only the measurement position POS_pre is obtained by GPS positioning after returning to the positioning enabled state (S12: NO, S16: YES), the time limit t_jdg is counted up (S17). Then, it is determined whether the counted up time limit t_jdg exceeds the specified time (S18).

  When the time limit t_jdg exceeds the specified time (S18: YES), the CPU 31 executes a predetermined timeout process (that is, turns off the post filter reset determination execution flag so as not to execute the process of S21 in FIG. 6). Then, the measurement position POS_pre held in the cache memory is discarded (S20), and the process proceeds to S31 in FIG. If the time limit t_jdg does not exceed the specified time (S18: NO), the process proceeds to S31 of FIG. 7 without executing the processes of S19 and S20.

  When the position is obtained twice by GPS positioning after returning to the positioning enabled state (that is, the next measurement position POS_cur is obtained in addition to the measurement position POS_pre) (S12: YES, S13: YES), An error ellipse ELL_cur corresponding to the measurement position POS_cur is calculated (S21).

  The error ellipse represents a range on the two-dimensional plane in which the GPS receiver 100 can exist at the time of GPS positioning (that is, measurement position error) by an ellipse, and is obtained from the geometric arrangement of the GPS satellites used at the time of GPS positioning. It is calculated based on the error covariance matrix and pseudorange error (error included in the distance from the GPS satellite measured by the GPS receiver). For example, when the geometrical arrangement of GPS satellites is biased in one direction, an error ellipse having a long axis (long axis) in a direction orthogonal to the one direction is obtained. If the geometrical arrangement of the GPS satellites is uniform, an error ellipse having a short axis and a long axis equivalent in length is obtained. In the former case, the measurement position error in the major axis direction is larger than the measurement position error in other directions, and in the latter case, the measurement position error in any direction has the same value. The size of the entire error ellipse increases in proportion to the pseudorange error. The error ellipse may be an ellipse representing a range in the three-dimensional space where the GPS receiver 100 can exist during GPS positioning.

  Subsequently, the CPU 31 calculates an estimated error ellipse ELL_pre corresponding to the measurement position POS_pre based on the measurement direction DIR_cur, measurement speed SPE_cur, and error ellipse ELL_cur included in the same GPS positioning result as the measurement position POS_cur (S22). Specifically, the CPU 31 sets the error ellipse ELL_cur that is shifted in the opposite direction to the measurement direction DIR_cur by a distance corresponding to the measurement speed SPE_cur as the estimated error ellipse ELL_pre. The distance shifted at this time may be, for example, a value obtained by multiplying the measurement speed SPE_cur by the time difference between the measurement position POS_cur and the measurement position POS_pre.

  The GPS 31 determines whether or not the measurement position POS_pre falls within the estimation error ellipse ELL_pre (S23). If the measurement position POS_pre falls within the estimation error ellipse ELL_pre (S23: YES), the CPU 31 determines that the measurement position POS_pre is a value obtained under relatively stable measurement conditions and that the accuracy of the initial positioning is good. To do. Therefore, it is determined that if the estimation by the Kalman filter is continued (that is, the measurement position POS_pre is also used for the estimation of the Kalman filter without discarding the measurement position POS_pre as in the process of S20), a more accurate GPS positioning result is obtained. Then, the post filter reset flag is turned off (S24), the post filter reset determination execution flag is turned off (S25), and the process proceeds to S31 in FIG.

  When the measurement position POS_pre does not fall within the estimation error ellipse ELL_pre (S23: NO), the CPU 31 determines that there is a large variation between the measurement position POS_pre and the measurement position POS_cur, and a large position jump has occurred. . Then, it is determined whether or not HDOP (Horizontal Dilution of Precision) _cur corresponding to the measurement position POS_cur is better than HDOP_pre corresponding to the measurement position POS_pre, and HDOP_cur is equal to or less than a predetermined value (S26).

  Note that the determination process of S23 is repeated many times until the measurement position POS_pre falls within the estimation error ellipse ELL_pre unless time-out occurs (that is, within the time limit t_jdg).

  When HDOP_cur is better than HDOP_pre and HDOP_cur is equal to or less than the predetermined value (S26: YES), the CPU 31 has high accuracy of the measurement position POS_cur and low accuracy of the measurement position POS_pre, that is, the accuracy of the first positioning is low, Judge that the accuracy of GPS positioning is high. In this case, if the low-precision measurement position POS_pre is set to the initial value of the Kalman filter after resetting the Kalman filter, the problem described above, that is, until the accuracy of the estimated value of the Kalman filter is significantly reduced and the estimated value converges to a good value. Problems that take a long time. Therefore, the CPU 31 turns on the post filter reset flag to set the good measurement position POS_cur as the initial value of the Kalman filter (S27), executes the processes of S19 and S20, and advances the process to S31 of FIG.

  If the non-positioning state continues in S31 of FIG. 7 (S31: NO), the CPU 31 uses the estimated value newly estimated by inheriting the past estimated value including the estimated value at the time of non-positioning. The GPS positioning result is corrected and output to the CPU 108 (S35). If GPS positioning has been performed (S31: YES), it is determined whether or not the post filter reset flag is on (S32).

  If the post filter reset flag is on (S32: YES), the CPU 31 once resets the estimated value of the Kalman filter and turns off the post filter reset flag (S33, S34). Then, for example, an initial value of the Kalman filter is set based on the measurement position remaining in the cache memory, the GPS positioning result is corrected using the estimated value estimated based on the set value, and is output to the CPU 108 (S35). ).

  On the other hand, if the post filter reset flag is off (S32: NO), the CPU 31 does not reset the estimated value of the Kalman filter, and uses the estimated value newly estimated by inheriting the estimated value in the past. Is corrected and output to the CPU 108 (S35).

  Here, the GPS measurement position improvement processing will be described using the examples shown in FIGS. 8A and 8B and FIG. In FIG. 9, each plot marked with ● represents the second and subsequent measurement positions when not processed with the Kalman filter, and each plot marked with Δ represents the second and subsequent measurement positions when processed with the Kalman filter.

  When the measurement position POS_pre falls within the estimation error ellipse ELL_pre as shown in FIG. 8A, the initial positioning result is obtained under relatively stable measurement conditions as described above, and the Kalman filter is used using the measurement position POS_pre. Even if the estimated value is obtained, the accuracy of the estimated value does not decrease, or even if it decreases, it is considered that it does not cause a problem. For this reason, the CPU 31 corrects the GPS positioning result by setting a new estimated value using the measurement positions POS_pre and POS_cur remaining in the cache memory without resetting the Kalman filter. That is, in this case, in the Kalman filter, after the estimated value is once reset after returning to the positioning enabled state, a new estimated value is calculated using the measurement positions after the measurement position POS_pre.

  Also, as shown in FIG. 8B, when the measurement position POS_pre does not fall within the estimation error ellipse ELL_pre, the accuracy of the measurement position POS_pre is low, and the accuracy shown in FIG. There is a possibility that it takes time until the estimated value converges to a good value. In this case, after resetting the estimated value of the Kalman filter in the process of S33, the CPU 31 sets a new estimated value using the next measured position POS_cur as an initial value without using the measured position POS_pre, which is concerned about low accuracy, Correct the GPS positioning result. That is, in this case, the measurement position POS_pre with low accuracy is not set as the initial value of the Kalman filter.

  As a result, the second and subsequent measurement positions processed by the Kalman filter are affected by the first positioning result as in the example shown by the x mark plot in FIG. It is a position close to the position by GPS positioning (that is, the position before the Kalman filter processing measured under relatively stable measurement conditions, and the position indicated by the ● mark plot in FIG. 9) (the Δ mark plot in FIG. 9). Thus, the position is corrected to draw a smooth trajectory when the measurement positions are connected in time series.

  As described above, according to the GPS receiver 100 of this embodiment, even if the accuracy of the first positioning is low and the measurement position jumps, it is possible to effectively avoid the subsequent decrease in accuracy of the estimated value of the Kalman filter.

  The present invention is not limited to the above embodiment, and various modifications are possible within the scope of the technical idea of the present invention. For example, the present invention can be applied not only to the car navigation apparatus of the present embodiment but also to a PND (Personal Navigation Device) that is a simple navigation apparatus that does not hold a map database, a mobile terminal such as a mobile phone, and the like.

In the process of S26 of FIG.
(1) HDOP_cur is better than HDOP_pre or (2) HDOP_cur is equal to or less than a predetermined value as a determination condition. There may be.

It is a figure for demonstrating the effect of a Kalman filter. It is a figure for demonstrating the problem when the precision of the estimated value of a Kalman filter falls. It is a block diagram which shows the structure of the car navigation apparatus of embodiment of this invention. It is a block diagram which shows the structure of the GPS receiver of embodiment of this invention. It is a flowchart which shows the GPS measurement position improvement process performed in embodiment of this invention. It is a flowchart which shows the GPS measurement position improvement process performed in embodiment of this invention. It is a flowchart which shows the GPS measurement position improvement process performed in embodiment of this invention. It is a figure for demonstrating the GPS measurement position improvement process performed in embodiment of this invention. It is a figure for demonstrating the GPS measurement position improvement process performed in embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Down converter part 2 Received signal processing part 3 Positioning calculation control part 100 GPS receiver 102 Gyro sensor 104 Vehicle speed sensor 108 CPU
200 Navigation device

Claims (7)

In a GPS receiver that performs GPS positioning using a GPS (Global Positioning System) signal, calculates an estimated position based on the measurement position obtained by the GPS positioning, and outputs the estimated position to the outside.
Current position estimating means for estimating a first range in which the GPS receiver can currently be located based on the result of the GPS positioning;
Based on the first range, past position estimating means for estimating a second range in which the GPS receiver could be located at a certain time in the past;
Past measurement position determining means for determining whether or not a past measurement position calculated by the GPS positioning at a certain time in the past is included in the second range, and the past measurement position is the second range When not included in the GPS receiver, the past measurement position is not used for the calculation of the estimated position. DOP comparison means for comparing DOP (Dilution of Precision) in the current GPS positioning and DOP in the GPS positioning at a certain past time when the past measurement position is not included in the second range. In addition,
The GPS receiver according to claim 2, wherein when the current DOP in the GPS positioning is better, the past measurement position is not used for the calculation of the estimated position. DOP determination means for determining whether or not the current DOP in the GPS positioning is below a specified value when the past measurement position is not included in the second range;
3. The GPS receiver according to claim 1, wherein when the current DOP in the GPS positioning is equal to or less than a predetermined value, the past measurement position is not used in the calculation of the estimated position. 4. .   The current position estimation means estimates the first range based on a geometrical arrangement of GPS satellites used at the time of GPS positioning and a pseudorange error obtained from the GPS positioning result. The GPS receiver according to claim 3.   The past position estimating means sets the second range as a result of shifting the first range based on a measurement direction and a measurement speed included in a current positioning result by the GPS positioning. The GPS receiver according to any one of claims 1 to 4.   6. The past certain point in time is a point in time when the GPS positioning is continued from a state where GPS positioning cannot be performed continuously for a certain period of time or longer. GPS receiver.   The calculation of the estimated position is an estimation calculation using a Kalman filter, and when the past measurement position is not included in the second range, the past measurement position is not used for the estimation calculation using the Kalman filter. The GPS receiver according to any one of claims 1 to 6, wherein the Kalman filter is reset.

Images