JP5510112B2 - Trajectory information generation apparatus, method, and program - Google Patents

Description

  The present invention relates to a technique for generating trajectory information indicating a travel trajectory of a vehicle.

  2. Description of the Related Art Conventionally, a technique for acquiring an estimated current position of a vehicle by self-contained navigation based on detection results from a direction sensor or a distance sensor is known. For example, in the technique disclosed in Patent Document 1, an estimated current position of a vehicle is acquired based on a detection result by an azimuth sensor or a distance sensor, and the estimated position trajectory is compared with a candidate position on the road to determine the most promising road. A technique for specifying and correcting the estimated current position based on the candidate position on the most promising road is disclosed. In addition, a technique for specifying an error circle for narrowing down roads for which candidate positions are specified based on the estimated current position and the positioning position by GPS is disclosed.

JP 2000-298028 A

The azimuth sensor and distance sensor used for self-contained navigation as in the prior art are sensors that detect the azimuth difference from the reference azimuth and the distance from the reference position, so it is assumed that the reference azimuth and reference position are inaccurate. The current position is incorrect. In addition, since the accumulated amount of error of the sensor increases with time, the estimated current position becomes inaccurate as the distance from the reference azimuth or reference position increases. Therefore, in the conventional technique, a candidate position is set on the road based on the estimated current position, and correction is performed so that the candidate position is regarded as the estimated current position. However, since the estimated current position acquired by the sensor includes an error, the correction is not always correct. Once incorrect correction is made, it becomes difficult to correct the current position because each piece of information detected by the sensor is a relative displacement from a reference such as a reference azimuth. End up.
The present invention has been made in view of the above problems, and an object thereof is to provide a technique for improving the accuracy of a self-contained navigation locus.

  In order to achieve the above object, in the present invention, the reliability of a time-series matching trajectory and a time-series GPS trajectory is acquired, and a trajectory with higher reliability of the GPS trajectory and the matching trajectory is defined as a correction target trajectory. To do. Then, the self-contained navigation information is corrected so that the difference between the self-contained navigation trajectory and the corrected target trajectory is reduced.

  That is, in the vehicle, the GPS trajectory and the matching trajectory can be acquired as information that can specify the position and orientation in the prescribed coordinate system. Therefore, by selecting a higher-reliability trajectory of the GPS trajectory and the matching trajectory as a correction criterion for the autonomous navigation information, the autonomous navigation trajectory is based on the information with higher reliability among the information obtained in the vehicle. Can be corrected. As a result, it is possible to effectively improve the accuracy of the self-contained navigation trajectory.

  Also, a configuration is adopted in which the GPS trajectory is used as a corrected target trajectory when the reliability of the GPS trajectory exceeds a predetermined reference, and the matching trajectory is used as a corrected target trajectory when the reliability of the GPS trajectory is less than a predetermined reference. You may do it. That is, the GPS trajectory is generated based on the GPS information, and the GPS information does not depend on the self-contained navigation information. On the other hand, the matching trajectory is specified based on a comparison between the self-contained navigation trajectory and the road shape indicated by the map information. Therefore, if the reliability of the GPS information is high and accurate, the GPS information is suitable as a reference for correcting the self-contained navigation information. Therefore, if the reliability of the GPS trajectory exceeds a predetermined standard, the GPS trajectory is corrected based on highly reliable information that does not depend on the self-contained navigation information. It becomes possible to correct the navigation information.

  In addition, the GPS information is affected by multipath and the like in addition to the relationship between the vehicle and the GPS satellite, so that the regularity of the error is less than that of the self-contained navigation information and can change suddenly. Therefore, when the reliability of the GPS trajectory is less than a predetermined reference, the reference for correcting the self-contained navigation information is used as a matching trajectory, so that information with low reliability (high probability of error correction) It is possible to prevent the self-contained navigation information from being corrected. As a result, it is possible to effectively improve the accuracy of the self-contained navigation trajectory.

  Here, the self-contained navigation trajectory acquisition means only needs to be able to acquire the self-contained navigation trajectory indicated by the time series output signal of the sensor mounted on the vehicle. For example, the time series output signal of the sensor mounted on the vehicle is obtained. It is only necessary that the position and direction of the vehicle can be indirectly acquired in time series by specifying the relative position and the displacement of the direction from the reference by acquiring and detecting the movement of the vehicle. Therefore, various physical quantities in the vehicle can be assumed as the operation of the vehicle. For example, if a sensor that acquires vehicle speed, acceleration, angular velocity, or the like is mounted on a vehicle, it is possible to acquire a self-contained navigation locus based on the output of the sensor. In addition, the self-contained navigation is a navigation method in which the movement of the vehicle is detected by a sensor mounted on the vehicle itself, and the position and direction of the vehicle in a specified coordinate system are indirectly specified based on the position displacement and the direction displacement from the reference. Therefore, it is not possible to directly acquire a position in a specified coordinate system such as latitude and longitude.

  The matching trajectory acquisition means only needs to be able to perform a map matching process that extracts the road that best matches the self-contained navigation trajectory and the road shape indicated by the map information and regards the vehicle as traveling on the road. In other words, a matching trajectory is obtained by performing a process for identifying the position and direction of a vehicle on a road regarded as a road on which the vehicle is traveling by performing a map matching process over a plurality of times. I can do it. Here, in the map information, since the position on the road that actually exists can be specified within a specified coordinate system such as latitude and longitude, if the matching trajectory is used as a reference for correction, only the autonomous navigation information can be used. When the position and direction of the specified vehicle are not possible in reality, it can be corrected to a position and direction that are realistic. The reliability of the matching trajectory only needs to indicate the reliability of information such as the position and direction of the vehicle indicated by the matching trajectory. For example, the shape of the road (road width, etc.) identified by the map matching process and the map It can be specified by the state of the vehicle during the matching process.

  The GPS trajectory acquisition unit only needs to be able to acquire a GPS trajectory based on time-series GPS information. Therefore, it is only necessary that a signal from a GPS satellite is acquired and the current position and current direction of the vehicle can be specified in a specified coordinate system based on the signal. The prescribed coordinate system can be assumed to be a coordinate system composed of latitude and longitude, a coordinate system composed of longitude, latitude and altitude, or the like. The GPS information is information that directly indicates the position and orientation in a specified coordinate system such as latitude and longitude. Therefore, even if an error may be included, latitude and longitude can be trusted within the error range in GPS information. For this reason, in GPS information, a position and azimuth completely different from the true position and azimuth are not detected. On the other hand, in the sensor mounted on the vehicle, the position and orientation in the coordinate system are indirectly specified by the relative displacement from the reference. Therefore, if the reference is inaccurate, the true position and orientation are completely different. Different positions and orientations may be detected. Therefore, the accumulated error in the self-contained navigation trajectory increases with time, but if correction using GPS information is performed, the accumulated error is canceled and the accuracy of the self-contained navigation trajectory can be improved early. Become.

  Furthermore, since the error of GPS information is poor in regularity and can change suddenly, it is not reliable to make one sample of GPS information a target for correction, but if GPS information is captured in a time-series set, The reliability of position and orientation can be statistically increased by increasing the total frequency. On the other hand, the self-contained navigation information includes errors due to the movement of the vehicle, etc., and due to the nature of specifying relative displacement from the reference by accumulating the vehicle speed, acceleration, direction, etc., the accumulated error in the self-contained navigation trajectory It increases with time. Therefore, if the self-contained navigation information is corrected so that the difference between the self-contained navigation trajectory and the GPS trajectory is reduced by comparing the time-series self-contained navigation trajectory with the time-series GPS trajectory, the accumulated error of the self-contained navigation trajectory is corrected by the correction. And the cumulative error can be prevented from increasing with time. Therefore, the accuracy of the self-contained navigation trajectory can be improved.

  The self-contained navigation information correcting unit only needs to be able to correct the self-contained navigation information so that the difference between the self-contained navigation locus and the corrected target locus is reduced by comparing the self-contained navigation locus with the corrected target locus. That is, by comparing the self-contained navigation trajectory with the correction target trajectory and specifying the difference between the two, it is possible to specify the correction to be performed on the self-contained navigation information in order to reduce the difference between the two. The correction may be correction for eliminating the difference between the self-contained navigation trajectory and the correction target trajectory, but is correction for reducing the difference without eliminating all the differences between the self-contained navigation trajectory and the GPS trajectory. It is preferable. In the latter case, if the correction is repeated, the self-contained navigation trajectory can be gradually brought closer to the correction target trajectory. In addition, even if the correction target trajectory is incorrect due to inaccuracies, a correction that eliminates the difference between the self-contained navigation trajectory and the correction target trajectory by a single correction is performed. It is possible to suppress the influence of erroneous correction rather than performing it.

  The reliability of the GPS trajectory only needs to indicate the reliability of information such as the position and direction of the vehicle indicated by the GPS trajectory. For example, the reliability of the GPS trajectory is compared by comparing the self-contained navigation trajectory with the GPS trajectory. Can be specified. That is, the error of the self-contained navigation information is regularly generated as compared with the error of the GPS information, and the degree of sudden change is small. Therefore, although the error of the self-contained navigation information accumulates with time, the reliability of the output information does not differ significantly between the output information output at close times. For this reason, the shape of the self-contained navigation trajectory indicated by the time-series output information of the sensor is accurate.

  Therefore, the higher the degree of coincidence between the self-contained navigation locus and the GPS locus, the more accurate the GPS locus can be considered, and the higher the degree of coincidence, the higher the reliability of the GPS locus is defined. Is possible. As a result, the trajectory with high reliability between the GPS trajectory and the matching trajectory is used as the correction target trajectory, and when the reliability of the GPS trajectory exceeds a predetermined reference, the self-contained navigation information is obtained using the GPS trajectory as the correction target trajectory. By correcting, it becomes possible to improve the accuracy of the self-contained navigation trajectory. Here, the higher the degree of coincidence between the shape of the self-contained navigation trajectory and the GPS trajectory, the higher the reliability of the GPS trajectory may be set. The degree may be changed continuously or may be changed stepwise.

  Further, the vehicle trajectory indicated by the self-contained navigation trajectory, the matching trajectory, and the GPS trajectory may include one or both of the azimuth and position of the vehicle. Therefore, a separate trajectory for each of the azimuth and the position may be used as the correction target trajectory based on the reliability for each of the azimuth and the position. In other words, the correction target locus is specified for each of the azimuth and position of the vehicle. Then, the vehicle direction indicated by the self-contained navigation information is corrected so that the difference between the vehicle direction indicated by the self-contained navigation track and the vehicle direction indicated by the correction target track is reduced, and the vehicle position and the correction target indicated by the self-contained navigation track are corrected. The position of the vehicle indicated by the self-contained navigation information is corrected so that the difference from the position of the vehicle indicated by the trajectory is reduced. According to this configuration, it is possible to perform correction while setting a more accurate trajectory as the correction target trajectory for each of the azimuth and the position.

  Furthermore, as a specific example of the method for calculating the GPS reliability, the reliability of the GPS trajectory is statistically compared by comparing each of the position and orientation indicated by the GPS trajectory with each of the position and orientation indicated by the autonomous navigation trajectory. Alternatively, a configuration that calculates automatically may be employed. For example, if the azimuth difference between the vehicle direction indicated by the self-contained navigation information for a plurality of time points and the vehicle direction indicated by the GPS information for a plurality of time points is statistically concentrated on a specific azimuth difference, It is considered that is tilted by a specific orientation difference with respect to the GPS trajectory. In addition, when the position difference between the vehicle position indicated by the self-contained navigation information for a plurality of time points and the vehicle position indicated by the GPS information for the plurality of time points is statistically concentrated on a specific position difference, Is considered to be offset by a specific position difference with respect to the GPS trajectory. In this case, it can be said that the self-contained navigation track and the GPS track have a high degree of coincidence of shapes although the angles and positions are different.

  Therefore, it is determined whether or not the azimuth difference between the vehicle direction indicated by the self-contained navigation information for a plurality of time points and the vehicle direction indicated by the GPS information for the plurality of time points is statistically concentrated on a specific azimuth difference. In the frequency distribution created at this time, the greater the total frequency, the greater the frequency of the mode value, and the smaller the variance, the higher the reliability of the direction of the vehicle indicated by the GPS trajectory. Further, it is determined whether or not the position difference between the vehicle position indicated by the self-contained navigation information for a plurality of time points and the vehicle position indicated by the GPS information for the plurality of time points is statistically concentrated on a specific position difference. In the frequency distribution created at this time, the greater the total frequency, the greater the frequency of the mode, and the smaller the variance, the higher the reliability of the position of the vehicle indicated by the GPS trajectory.

  Furthermore, the reliability of the matching trajectory only needs to indicate the reliability of information such as the position and direction of the vehicle indicated by the matching trajectory, and is based on, for example, the road on which the vehicle is considered to have traveled by the map matching process. Reliability can be specified. For example, when the road width of a road on which the vehicle is traveling is narrow, the degree of freedom and error in the position of the vehicle are smaller than when the road width is wide. It can be considered high. In this case, it is sufficient that the reliability of the position indicated by the matching trajectory can be set higher as the road width of the road on which the vehicle is traveled is narrower, and the road on which the vehicle is traveled. Depending on the road width, the reliability may change continuously, or it may change in stages.

  Further, the reliability may be specified based on the direction change of the vehicle in the process of performing the map matching process. For example, when the azimuth variation of the vehicle in the process of performing the map matching process is small, the azimuth to be the map matching target is more stable than when the azimuth variation is large. Therefore, if a road is considered that the vehicle is traveling in a stable state, the probability that the vehicle is traveling on the road that is considered to be traveling is not stable. Higher than the state. Therefore, it is possible to consider that the reliability of the direction of the vehicle indicated by the matching trajectory is higher as the variation in the direction of the vehicle in the process of performing the map matching process is smaller. Here, it is only necessary to set the reliability of the azimuth indicated by the matching trajectory to be higher as the azimuth fluctuation of the vehicle is smaller, and the reliability may be continuously changed according to the magnitude of the azimuth fluctuation of the vehicle. However, a configuration that changes stepwise may be used.

  Furthermore, as in the present invention, the method of correcting the self-contained navigation information so as to reduce the difference between the time-series self-contained navigation trajectory and the time-series corrected target trajectory can be applied as a program or method. It is. In addition, the above-described device, program, and method may be realized as a single device or may be realized by using components shared with each part of the vehicle, and include various aspects. It is a waste. For example, it is possible to provide a navigation device, a method, and a program that include the above devices. Further, some changes may be made as appropriate, such as a part of software and a part of hardware. Furthermore, the invention is also established as a recording medium for a program for controlling the apparatus. Of course, the software recording medium may be a magnetic recording medium, a magneto-optical recording medium, or any recording medium to be developed in the future.

It is a block diagram of a locus information generating device. It is a flowchart which shows a self-supporting navigation information correction process. It is a flowchart which shows the comparison process of a GPS locus | trajectory and a self-supporting navigation locus | trajectory. (4A)-(4C) are explanatory views for explaining rotation and movement of a self-contained navigation locus. (5A) to (5D) are explanatory diagrams for explaining statistical processing for specifying a heading difference and a representative value. (6A) to (6C) are explanatory diagrams for explaining statistical processing for specifying a position difference and a representative value. It is a flowchart which shows a self-supporting navigation information correction process. (8A) is a diagram illustrating a state of correcting the self-contained navigation position, and (8B) is a diagram illustrating a state of correcting the self-contained navigation direction.

Here, embodiments of the present invention will be described in the following order.
(1) Configuration of the trajectory information generation device:
(2) Trajectory information generation processing:
(2-1) Map matching process:
(2-2) Self-contained navigation information correction processing:
(3) Other embodiments:

(1) Configuration of the trajectory information generation device:
FIG. 1 is a block diagram showing a configuration of a trajectory information generation device 10 according to the present invention. The trajectory information generation device 10 includes a control unit 20 including a CPU, a RAM, a ROM, and the like, and a recording medium 30, and the control unit 20 can execute a program stored in the recording medium 30 or the ROM. In the present embodiment, the navigation program 21 can be executed as this program. The navigation program 21 has a function of performing map matching processing based on self-contained navigation information to identify the position of the vehicle on the road and displaying the position of the vehicle on the map. In the present embodiment, the navigation program 21 can execute trajectory information generation processing for generating trajectory information indicating the trajectory of the vehicle by a plurality of methods. In particular, the navigation program 21 increases the self-contained navigation trajectory generated based on the self-contained navigation information. It has a function to generate accurately.

  Map information 30a is recorded in the recording medium 30 in advance. The map information 30a is information used for specifying the position of the vehicle, etc., node data indicating the position of the node set on the road on which the vehicle travels, and the shape for specifying the shape of the road between the nodes. Interpolation point data, link data indicating connection between nodes, feature data indicating the position and type of features existing on the road and its surroundings, and the like are included. In the present embodiment, the link data includes information indicating the road width of the road corresponding to the link.

  The vehicle on which the trajectory information generation device 10 according to this embodiment is mounted includes a user I / F unit 40, a vehicle speed sensor 41, a gyro sensor 42, and a GPS receiving unit 43. The navigation program 21 includes a self-contained navigation trajectory acquisition unit 21a, a matching trajectory acquisition unit 21b, a GPS trajectory acquisition unit 21c, and a self-contained navigation information correction unit 21d. The navigation program 21 cooperates with the vehicle speed sensor 41, the gyro sensor 42, and the GPS receiving unit 43 to execute a function of generating a self-contained navigation locus generated based on the self-contained navigation information with high accuracy.

  The vehicle speed sensor 41 outputs a signal corresponding to the rotational speed of the wheels included in the vehicle, and the gyro sensor 42 outputs a signal corresponding to the angular velocity acting on the vehicle. The self-contained navigation trajectory acquisition unit 21a is a module that causes the control unit 20 to realize a function of acquiring a self-contained navigation trajectory that is a vehicle trajectory indicated by time-series self-contained navigation information. That is, the control unit 20 acquires the output signals of the vehicle speed sensor 41 and the gyro sensor 42 as self-contained navigation information through an interface (not shown) by the processing of the self-contained navigation track acquisition unit 21a.

  Then, the control unit 20 specifies the relative position displacement of the vehicle from the reference position based on the output signal of the vehicle speed sensor 41, specifies the current vehicle position, and determines the reference based on the output signal of the gyro sensor 42. The current vehicle orientation is identified by identifying relative vehicle orientation displacement from the orientation. In this specification, the position of the vehicle specified based on the output signal of the vehicle speed sensor 41 is referred to as a self-contained navigation position, and the direction of the vehicle specified based on the output signal of the gyro sensor 42 is referred to as a self-contained navigation direction. . Note that the above-described reference position and reference direction may be any vehicle position or vehicle direction (traveling direction) specified at a predetermined time. For example, the vehicle position specified from GPS information described later at a predetermined time may be used. It can be configured by position, vehicle orientation, and the like. Further, the control unit 20 specifies information indicating the time-series self-contained navigation position and the self-contained navigation direction by specifying the self-contained navigation position and the self-contained navigation direction at a plurality of times, and acquires the information as a self-contained navigation locus.

  When the matching trajectory acquisition unit 21b performs a map matching process that regards the road on which the self-contained navigation trajectory and the shape of the road indicated by the map information 30a are the best match as the road on which the vehicle is traveling, and is identified by the map matching process This is a module that causes the control unit 20 to realize a function of acquiring a matching trajectory that is a trajectory of a series of vehicles. That is, the control unit 20 refers to the map information 30a and collates the road shape of the road existing around the vehicle with the autonomous navigation locus. Then, the road with the highest degree of coincidence between the two is identified, and the position and direction estimated as the position and direction of the vehicle on the road are regarded as the road on which the vehicle is traveling, and the position and direction of the vehicle are defined as the position and direction of the vehicle. Identify.

  In the present specification, the position of the vehicle specified by the map matching process as described above is called a matching position, and the direction of the vehicle specified by the map matching process is called a matching direction. Further, the control unit 20 specifies information indicating the time-series matching position and the matching direction by specifying the matching position and the matching direction at a plurality of time points, and acquires the information as a matching locus.

  Then, the control unit 20 displays the matching position, the matching direction, and the matching locus on the user I / F unit 40. That is, the user I / F unit 40 is an interface unit for a user to input an instruction or provide various kinds of information to the user, and includes a display unit, buttons, speakers, and the like (not shown). In the present embodiment, the control unit 20 displays a map on the display unit of the user I / F unit 40 and displays icons indicating the matching position, the matching direction, and the matching locus on the map. Therefore, the control unit 20 generates image data indicating an icon indicating a map, a matching position, a matching direction, and a matching locus, and outputs the image data to the user I / F unit 40. Based on the image data, the user I / F unit 40 displays a map, an icon indicating a matching position, a matching direction, and a matching locus on the display unit.

  Note that the control unit 20 can specify the reliability of the matching position and the matching direction specified by the map matching process by the process of the matching locus acquisition unit 21b. In the present embodiment, the control unit 20 identifies the reliability of the matching position based on the road width of the road that the vehicle identified by the map matching process is traveling, and performs the map matching process. The reliability of the matching direction is specified based on the direction change of the vehicle at.

  The GPS trajectory acquisition unit 21c is a module that causes the control unit 20 to realize a function of acquiring a GPS trajectory that is a vehicle trajectory indicated by time-series GPS information. That is, the control unit 20 causes the GPS receiving unit 43 to acquire GPS information for calculating the current position and current direction of the vehicle by the process of the GPS trajectory acquisition unit 21c, and the position and direction of the vehicle based on the GPS information. Is identified. In this specification, the position of the vehicle specified based on the GPS information is referred to as a GPS position, and the direction of the vehicle specified based on the GPS information is referred to as a GPS direction. The control unit 20 further specifies information indicating the time-series GPS position and GPS azimuth by specifying the GPS position and GPS azimuth at a plurality of time points, and acquires the information as a GPS trajectory. Note that the GPS information in the present embodiment includes GPS accuracy information indicating the accuracy of the GPS information.

  In the present embodiment, the GPS accuracy information includes an index acquired together with the GPS information and an index acquired based on the state of the vehicle. That is, the accuracy of GPS information decreases due to the influence of the relative relationship between the satellite and the vehicle and the communication environment (degree of multipath, etc.), and an index (HDOP (Horizontal Dilution Of Precision) indicating such a decrease in accuracy. The GPS information includes DOP (Dilution Of Precision) such as), a state where positioning with high accuracy is possible, and the number of satellites at the position. Therefore, in the present embodiment, the control unit 20 acquires the GPS accuracy information included in the GPS information together with the GPS information by the process of the GPS trajectory acquisition unit 21c.

  In addition, since the accuracy of GPS information depends on the state of the vehicle, in this embodiment, an index indicating the degree of decrease in the accuracy of GPS information due to the state of the vehicle is also acquired. Specifically, the control unit 20 acquires the output information of the vehicle speed sensor 41 and the output information of the gyro sensor 42 by the process of the GPS trajectory acquisition unit 21c, and the accuracy of the GPS information decreases as the vehicle speed of the vehicle decreases. Then, an index is obtained as GPS accuracy information such that the accuracy of the GPS information decreases as the maximum value of the change in the direction of the vehicle during the predetermined period increases. In the present embodiment, the GPS accuracy information is specified for each of the GPS position and the GPS azimuth, and the accuracy of the GPS position is specified based on an accuracy index included in the GPS information. Further, regarding the GPS azimuth (the GPS satellite azimuth described later), the accuracy is specified based on the accuracy index included in the GPS information and the index indicating the degree of decrease in the accuracy of the GPS information due to the state of the vehicle. In the present embodiment, the standardization is performed so that the state with the highest accuracy is 100 and the state with the lowest accuracy is 0 in each of the GPS position and the GPS azimuth.

  As described above, the present embodiment is configured to acquire the matching position and the matching direction while collating the self-contained navigation trajectory and the matching trajectory and display them on the user I / F unit 40. Furthermore, in the present embodiment, Adopts a configuration in which the self-contained navigation trajectory is corrected by the matching trajectory or the GPS trajectory in consideration of the properties of the self-contained navigation trajectory, the matching trajectory, and the GPS trajectory.

  That is, since the matching position and the matching direction displayed on the user I / F unit 40 are the position and direction specified by matching the self-contained navigation trajectory with the road shape, if the self-contained navigation trajectory is inaccurate, the matching is performed. The position and matching orientation will also be inaccurate. Specifically, in the self-contained navigation, the position of the self-contained navigation is specified by specifying the relative displacement of the vehicle from the reference position based on the output information of the vehicle speed sensor 41, and based on the output signal of the gyro sensor 42. A self-contained navigation azimuth is specified by specifying a relative displacement of the vehicle from the reference azimuth. Therefore, if the reference position of the self-contained navigation position and the reference direction of the self-contained navigation direction are inaccurate, the position and direction of the self-contained navigation trajectory will be inaccurate, and the position and direction completely different from the true position and direction will be the matching position and matching. It can be detected as an orientation. Further, since errors in the output information of the vehicle speed sensor 41 and the gyro sensor 42 accumulate with time, the accuracy of the self-contained navigation position and the self-contained navigation direction decreases with time, and even in this case, the position is completely different from the true position and direction. Or the orientation may be detected as the matching position and the matching orientation.

  On the other hand, the GPS information is information that directly indicates the position and orientation in a specified coordinate system such as latitude and longitude. Therefore, even if an error may be included, latitude and longitude can be trusted within the error range in GPS information. For this reason, in GPS information, a position and azimuth completely different from the true position and azimuth are not detected. Therefore, by correcting the self-contained navigation information so that the difference between the self-contained navigation trajectory and the GPS trajectory is reduced, it is possible to prevent the position and direction indicated by the self-contained navigation information from being completely different from the true position and direction. it can.

  Furthermore, since the node data, shape interpolation data, and the like included in the map information 30a indicate the position on the road that actually exists, such as latitude and longitude, the matching position and the matching direction are actual as the position and direction on the road. It becomes the position and direction which are possible. Therefore, if the self-contained navigation information is corrected so that the difference between the self-contained navigation trajectory and the matching trajectory is reduced, the position and direction indicated by the self-contained navigation information can be corrected to at least a realistic position and direction.

  In view of such trajectory characteristics, in the present embodiment, the control unit 20 uses more of the reliability of the trajectory set for each of the GPS trajectory and the matching trajectory by the processing of the self-contained navigation information correction unit 21d. A highly reliable trajectory is used as a correction target trajectory, and the self-contained navigation information is corrected so that the difference between the self-contained navigation trajectory and the correction target trajectory is reduced. In other words, the GPS trajectory and the matching trajectory are information that directly indicates the position and orientation in the prescribed coordinate system, and thus can serve as a reference for correcting the self-contained navigation trajectory that indirectly indicates the position and orientation. . Therefore, by selecting a higher-reliability trajectory of the GPS trajectory and the matching trajectory as a correction criterion for the autonomous navigation information, the autonomous navigation trajectory is based on the information with higher reliability among the information obtained in the vehicle. Can be corrected. As a result, it is possible to effectively improve the accuracy of the self-contained navigation trajectory. In the present embodiment, the reliability of the matching trajectory is acquired by the processing of the matching trajectory acquisition unit 21b, and the reliability of the GPS trajectory is acquired by the processing of the autonomous navigation information correction unit 21d. Details of these reliability levels will be described later.

  The correction of the self-contained navigation information only needs to be executed so that the difference between the self-contained navigation trajectory and the corrected target trajectory is reduced. In this embodiment, the control unit 20 has the highest degree of coincidence between the self-contained navigation trajectory and the corrected target trajectory. Rotate and move the self-contained navigation trajectory so that it rises, set a correction target for self-contained navigation information based on the trajectory after the rotation and movement, and self-contained navigation information so that the difference from the correction target is reduced Correct. When the correction target trajectory is a matching trajectory, the correction target, the matching position, and the matching direction set by performing the above-described rotation and movement are substantially equivalent. That is, when the correction target locus is a matching locus, the position of the correction target is the matching locus, and the direction of the correction target is the matching direction.

  When the corrected target locus is a GPS locus, the autonomous navigation locus is rotated and moved so that the degree of coincidence between the autonomous navigation locus and the GPS locus is the highest. That is, the control unit 20 rotates the self-contained navigation trajectory while maintaining the shape of the self-contained navigation trajectory, and compares the self-contained navigation position and the self-contained navigation azimuth obtained as a result of the parallel movement with the GPS position and the GPS azimuth. A state in which the difference between a plurality of positions and orientations is the smallest is regarded as the highest degree of coincidence. As a result, the rotation angle and the movement amount for the compared autonomous navigation trajectory are specified, and the control unit 20 adds the rotation angle to the current autonomous navigation direction and adds the movement amount to the autonomous navigation position. Result is regarded as the correction target.

  The control unit 20 outputs the output information of the vehicle speed sensor 41 and the gyro sensor 42 so that the self-contained navigation position and the self-contained navigation direction approach the correction target specified as described above by the processing of the self-contained navigation information correction unit 21d. Correct the output information. Here, the self-contained navigation position and the self-contained navigation direction specified on the basis of the output information of the vehicle speed sensor 41 and the gyro sensor 42 may be corrected. Of course, the reference position and the reference direction may be corrected. In the present embodiment, the self-contained navigation information correction unit 21d repeatedly performs correction so that the self-contained navigation position and the self-contained navigation direction approach the correction target. However, the self-contained navigation direction and the self-contained navigation direction are determined at each correction. It is configured not to match the correction target.

  That is, when the correction amount when correcting the self-contained navigation position and the self-contained navigation direction so as to coincide with the correction target is the first correction amount, the control unit 20 is independent by using the second correction amount smaller than the first correction amount. A configuration for correcting navigation information is adopted. As described above, once the self-contained navigation locus is erroneously corrected, it is difficult to correct the self-contained navigation position and the self-contained navigation direction correctly. Therefore, in the present invention, the self-contained navigation information is corrected by the second correction amount that is smaller than the first correction amount so that the influence is suppressed even if erroneous correction is performed. ing. Therefore, the accuracy of the self-contained navigation track can be easily improved. Note that the second correction amount is smaller than the first correction amount, and may be set so as to realize a part of the correction realized by the first correction amount.

(2) Trajectory information generation processing:
Next, trajectory information generation processing by the navigation program 21 will be described. The trajectory information generation process by the navigation program 21 is executed every predetermined period, and a self-contained navigation information correction process for improving the accuracy of the self-contained navigation trajectory is executed in the process. FIG. 2 is a flowchart showing a self-contained navigation information correction process. On the other hand, the control part 20 is performing the process of the matching locus | trajectory acquisition part 21b for every predetermined period in parallel with the said independent navigation information correction process. Here, first, the map matching process will be described.

(2-1) Map matching process:
In the map matching process, the control unit 20 acquires and collates the map information 30a and the self-contained navigation track by the processing of the matching track acquisition unit 21b. That is, the control unit 20 acquires the output information of the vehicle speed sensor 41 and the gyro sensor 42 by the process of the self-contained navigation information correction unit 21d, and specifies the displacement of the position and direction of the vehicle after acquiring the output information last time. . Then, the control unit 20 specifies the self-contained navigation position and the self-contained navigation direction based on the accumulated value of the displacement of the position from the reference position and the accumulated value of the displacement of the azimuth from the reference azimuth, and is specified within a predetermined period. The self-contained navigation locus and the self-contained navigation direction are used as the self-contained navigation locus. Note that the reference position and the reference orientation can be updated with the passage of time, but the initial values are determined based on GPS information or the like.

  In addition, the control unit 20 refers to the map information 30a and acquires link data, node data, and shape interpolation data indicating the shape of the road existing within a predetermined range around the autonomous navigation position. And the control part 20 collates the independent navigation locus | trajectory and each data acquired from the map information 30a, and specifies the road which most closely matches an autonomous navigation locus | trajectory.

  Next, the control part 20 acquires a matching locus | trajectory by the process of the matching locus | trajectory acquisition part 21b. That is, the control unit 20 regards the road on which the vehicle is most consistent with the self-contained navigation trajectory as the road on which the vehicle is traveling, and rotates and moves the self-contained navigation trajectory so that the self-contained navigation trajectory and the road most closely match. Then, the control unit 20 sets a position corresponding to the latest self-contained navigation position in the self-contained navigation trajectory rotated and moved as a matching position, and sets the traveling direction of the vehicle on the road at the matching position as a matching direction. Furthermore, the control unit 20 acquires a time-series matching position and a matching direction specified within a predetermined period as a matching locus.

Furthermore, the control unit 20 acquires the reliability of the matching locus. In the present embodiment, the reliability is acquired by a different method for each of the matching position and the matching direction. That is, the control unit 20 determines the reliability of the matching position based on the distance when the vehicle is regarded as traveling on a continuous road and the road width of the road regarded as the road on which the vehicle is traveling. Set the degree. In the present embodiment, the accuracy of the matching position is defined in five stages as shown in Table 1 below. That is, the control unit 20 sets the reliability of the matching position to 1 when the distance when the vehicle is considered to be traveling on a continuous road is equal to or less than the predetermined distance Ts (m), and the predetermined distance When it is larger than Ts, the reliability of the matching position is set to any one of 1 to 5. For this reason, the control unit 20 refers to the link data of the map information 30a and specifies the road width of the road regarded as the road on which the vehicle is traveling. Then, when the road width is equal to or less than the predetermined threshold T ₄ , the control unit 20 sets the reliability of the matching position to 5, and the road width is greater than the predetermined threshold T _{4 and} equal to or less than the predetermined threshold T ₃ . In some cases, the reliability of the matching position is set to 4. If the road width is greater than the predetermined threshold T _{3 and} less than or equal to the predetermined threshold T ₂ , the reliability of the matching position is set to 3, and the road width is greater than the predetermined threshold T _{2 and} less than or equal to the predetermined threshold T _1. If it is, the reliability of the matching position is set to 2. Then, the road width is in the case greater than the predetermined thresholds T ₁ sets the reliability of the matching position to 1. That is, the control unit 20 sets the reliability of the matching position higher as the road width of the road regarded as the road on which the vehicle is traveling is narrower.

Furthermore, the control unit 20 determines the reliability of the matching direction based on the distance when the vehicle is considered to be traveling on a continuous road and the direction change of the vehicle in the process of performing the map matching process. Set. In this embodiment, the accuracy of the matching direction is defined in five stages. That is, the control unit 20 sets the matching direction reliability to 1 when the distance when the vehicle is considered to be traveling on a continuous road is equal to or less than the predetermined distance Ts (m), and the predetermined distance When it is larger than Ts, the reliability of the matching direction is set to be 2 or more. Therefore, the control unit 20 further specifies the azimuth change within the predetermined period based on the output information of the gyro sensor 42 within the past predetermined period. Then, the control unit 20 sets the matching azimuth reliability to 3 when the azimuth change is smaller than the predetermined threshold Td, and sets the matching azimuth reliability when the azimuth change is equal to or larger than the predetermined threshold Td. Set to 2. That is, the reliability of the matching azimuth is set to be higher as the azimuth variation in the matching process is smaller.

  In the present embodiment, as described above, the reliability is set by different methods for the matching position and the matching direction. However, in a specific case, the reliability of the matching position and the matching direction is set by a method common to the matching position and the matching direction. That is, when the vehicle is traveling on a normal road, each of the matching position and the matching direction is specified based on the above-described Table 1 and Table 2, but GPS information cannot be acquired in the tunnel. Therefore, a method different from Tables 1 and 2 is also employed so that the matching trajectory is easily set as the correction target trajectory.

  Specifically, when it is considered that the vehicle is traveling on a road in the tunnel, the control unit 20 performs node processing and shape interpolation point data included in the map information 30a by processing of the matching trajectory acquisition unit 21b. , Referring to the link data, identify the shape of the road within a predetermined range around the vehicle, and determine the azimuth (forward in the direction of travel when driving on the road) and the time-series autonomous navigation position Collation is performed to identify the degree of coincidence of directions. Then, the reliability is set so that the higher the coincidence degree of the azimuth, the higher the reliability of the matching position and the matching direction. Note that the degree of azimuth coincidence can be specified by azimuth variance or the like. The reliability of the matching position and the matching direction as described above is set for each of the latest matching position and matching direction each time it is specified.

  In addition, this configuration is an example, and the degree of coincidence of the positions may be analyzed, or similar processing may be performed in a parking lot where GPS information cannot be acquired. Furthermore, since the reliability of the matching position and the reliability of the matching direction are the reliability of the information obtained as a result of the map matching process, the reliability is lowered when the accuracy of the map matching process can be lowered. It is good also as a structure. For example, other promising candidate roads (for example, acute-angle branch roads or parallel roads) within a predetermined period after the start of driving by the vehicle (start of power-on) or within a predetermined distance around the road on which the vehicle is traveling. If the difference between the direction indicated by the output information of the gyro sensor 42 and the direction of the road is more than a predetermined difference, the road regarded as the road on which the vehicle is traveling is the highway. When it is in an attached facility (service area, parking area, etc.), the reliability of the matching position and the matching direction is reduced, and a certain low value (such as reliability 1) is adopted. Is possible.

(2-2) Self-contained navigation information correction processing:
In the self-contained navigation information correction process shown in FIG. 2, the control unit 20 acquires GPS accuracy information by the process of the GPS trajectory acquisition unit 21c (step S100). Next, the control unit 20 performs a comparison process between the GPS locus and the self-contained navigation locus (step S105). The comparison process between the GPS trajectory and the self-contained navigation trajectory acquires the reliability of the GPS trajectory, and the first correction amount (rotation angle and movement amount) for obtaining the highest degree of coincidence between the self-contained navigation trajectory and the GPS trajectory. ), And is realized by the flowchart shown in FIG. 4A to 4C are diagrams for explaining the rotation and movement of the self-contained navigation trajectory in a coordinate system in which the longitude is the x-axis and the latitude is the y-axis.

  FIG. 4A shows an example of a self-contained navigation locus Tn by a solid curve arrow, and FIG. 4B shows an example of a GPS locus by a black circle and an arrow. That is, in FIG. 4B, the GPS position Gp is indicated by a black circle, the GPS satellite azimuth Gds at each GPS position Gp is indicated by a solid straight arrow, and the GPS coordinate azimuth Gdc between the GPS positions Gp is indicated by a broken straight arrow. ing. That is, the GPS information includes information indicating the direction of the vehicle, but it is also possible to regard the direction of the vehicle position vector as the vehicle direction separately from the direction of the vehicle. Therefore, in the present embodiment, the direction of the vehicle included in the GPS information is called the GPS satellite direction, the direction of the vehicle position vector is called the GPS position direction, and based on both the GPS satellite direction and the GPS position direction. It is configured to evaluate the GPS direction.

  In addition, although the self-contained navigation trajectory Tn is originally composed of a plurality of self-contained navigation positions and self-contained navigation directions, in the present embodiment, the self-contained navigation information has a shorter sampling period than the GPS information. A solid curved arrow indicates a self-contained navigation trajectory Tn, and the self-contained navigation position Np and the self-contained navigation azimuth Nd are illustrated only for one place. That is, in the self-contained navigation trajectory Tn in FIG. 4A, the solid curve arrow indicates that any position on the curve is the self-contained navigation position, and the direction indicated by the arrow at the tip of the curve arrow is the latest self-contained navigation direction. Yes, the tangent of the curved arrow at each autonomous navigation position indicates the autonomous navigation direction at each autonomous navigation position.

  As shown in FIGS. 4A and 4B, the self-contained navigation trajectory Tn and the GPS trajectory are generally similar but different. Since GPS information is affected by multipath and the like in addition to the relationship between the vehicle and the GPS satellite, the regularity of the error is less than that of the self-contained navigation information and can change suddenly. On the other hand, errors in output information from the vehicle speed sensor 41 and the gyro sensor 42 are regularly generated as compared with errors in GPS information, and the degree of sudden change is small. Therefore, although the error of the self-contained navigation information Tn accumulates with time, the reliability of the output information is not significantly different between the output information output at close times. For this reason, the shape of the self-contained navigation trajectory Tn is more accurate than the GPS trajectory. Therefore, when the self-contained navigation trajectory Tn is rotated and moved while maintaining the shape of the GPS trajectory, the degree of coincidence between the GPS trajectory and the self-contained navigation trajectory Tn is high. It can be evaluated that the reliability of the GPS trajectory is higher. In addition, when the reliability of the GPS trajectory is high and it can be a corrected target trajectory, the rotation angle when rotating and moving while maintaining the shape of the self-contained navigation trajectory Tn and having the highest degree of coincidence with the GPS trajectory The movement amount can be regarded as the first correction amount of the self-contained navigation direction and the first correction amount of the self-contained navigation position.

  Therefore, in the present embodiment, the first correction amount of the self-contained navigation direction is statistically specified based on the GPS satellite azimuth corresponding to the plurality of GPS positions and the self-contained navigation position, the inter-GPS coordinate azimuth, and the self-contained navigation azimuth. . That is, the azimuth difference between a plurality of GPS satellite azimuths and a plurality of autonomous navigation azimuths and the representative value of the azimuth difference between a plurality of GPS coordinate azimuths and a self-contained navigation azimuth has the highest degree of coincidence between the autonomous navigation locus and the GPS locus. The rotation angle of the self-contained navigation trajectory Tn in the state is considered.

  Since the self-contained navigation trajectory Tn before the rotation is actually composed of a plurality of self-contained navigation positions Np and self-contained navigation directions Nd, the self-contained navigation direction Nd and GPS directions (GPS satellite direction Gds and GPS coordinate direction Gdc at the same time) If the azimuth difference (rotational angle) is specified, the rotational angle at which the degree of coincidence between the GPS trajectory and the self-contained navigation trajectory Tn is the highest by rotating the self-contained navigation trajectory Tn at that time. Therefore, if the representative value of the azimuth difference between the self-contained navigation azimuth Nd and the GPS azimuth at a plurality of times is specified, the representative value is the rotation angle for maximizing the degree of coincidence between the GPS trajectory and the self-contained navigation trajectory Tn. Can be considered. In FIG. 4C, the self-contained navigation trajectory Tn before rotation is shown by a dashed curved arrow, and the self-contained navigation trajectory Tnr after rotation is shown by a dashed-dotted curved arrow. In FIG. 4C, the degree of coincidence between the GPS trajectory and the self-contained navigation trajectory Tn. The rotation angle for making the highest is shown by α.

  Further, in the present embodiment, the first correction amount of the self-contained navigation position is statistically specified based on the plurality of GPS positions and the self-contained navigation position. That is, the representative value of the position difference between the plurality of GPS positions and the plurality of autonomous navigation positions is regarded as the amount of movement of the autonomous navigation locus in the state where the degree of coincidence between the autonomous navigation locus and the GPS locus is the highest. For example, in FIG. 4C, the dashed curved arrow indicates the self-contained navigation trajectory Tnr before movement, and the solid curved arrow indicates the autonomous navigation trajectory Tnm after movement.

  Since the self-contained navigation trajectory Tnr before movement is a trajectory obtained by rotating the self-contained navigation trajectory at the rotation angle α described above, it is composed of a plurality of self-contained navigation positions and self-contained navigation directions after rotation. Therefore, if the position difference between the self-contained navigation position after rotation at the same time and the GPS position (the amount of movement along the x-axis and y-axis) is specified, the self-contained navigation trajectory Tnr after rotation at that time is moved. The amount of movement with the highest degree of coincidence between the GPS trajectory and the self-contained navigation trajectory Tn is specified. Therefore, if the representative value of the position difference between the self-contained navigation position and the GPS position at a plurality of times is specified, the amount of movement for the representative value to maximize the degree of coincidence between the GPS trajectory and the self-contained navigation trajectory Tn (FIG. 4C). X and Y) shown in FIG.

  The comparison process between the GPS trajectory and the self-contained navigation trajectory shown in FIG. 3 is a process for realizing the determination of the rotation angle and the movement amount as described above and the specification of the reliability based on the rotation angle and the movement amount. In order to execute this process, first, the control unit 20 obtains the travel distance from the time when the comparison process between the GPS locus and the autonomous navigation locus shown in FIG. Then, it is determined whether or not the travel distance is equal to or greater than a predetermined distance (step S200). If it is not determined in step S200 that the travel distance is greater than or equal to the predetermined distance, the processing after step S205 is skipped. That is, the comparison process is executed for each predetermined distance.

  When it is determined in step S200 that the travel distance from the time when the comparison process between the previous GPS trajectory and the self-contained navigation trajectory is executed is greater than or equal to the predetermined distance, the control unit 20 performs GPS by the process of the GPS trajectory acquisition unit 21c. A trajectory is acquired and a self-contained navigation trajectory is acquired by the processing of the self-contained navigation information correction unit 21d (step S205). Next, the control part 20 acquires the azimuth | direction difference of a GPS satellite azimuth | direction and an independent navigation azimuth | direction by the process of the autonomous navigation information correction | amendment part 21d (step S210). That is, the control unit 20 acquires GPS satellite azimuths and autonomous navigation azimuths regarding a plurality of positions based on GPS information and autonomous navigation information acquired within a predetermined period, and acquires azimuth differences between the azimuths at the same time.

  Further, the control unit 20 performs weighting according to the GPS accuracy by the processing of the self-contained navigation information correction unit 21d to create a frequency distribution of the azimuth difference (step S215). That is, since the accuracy of each GPS azimuth (GPS satellite azimuth) is specified by the GPS accuracy information, in statistical processing for specifying the azimuth difference between the GPS trajectory and the self-contained navigation trajectory, the accuracy of the GPS azimuth increases. The GPS information is configured to greatly contribute to the determination of the heading difference. Specifically, the control unit 20 acquires GPS accuracy, determines the frequency of each azimuth difference acquired in step S210 so that the higher the GPS azimuth accuracy, the higher the frequency, and adds it to the frequency distribution. In this way, a frequency distribution is created. The frequency only needs to be set so that the frequency becomes larger as the accuracy of the GPS azimuth becomes higher. For example, the accuracy information itself of the GPS azimuth standardized so as to become a larger value as the accuracy becomes higher can be used as the frequency. A value obtained by multiplying a predetermined coefficient may be used as the frequency.

  FIG. 5A to FIG. 5D are diagrams for explaining statistical processing for specifying a azimuth difference between a GPS azimuth and a self-contained navigation azimuth and a representative value thereof. In these 5A to FIG. 5D, the azimuth | direction in a coordinate system is shown by the straight arrow extended from a black circle on the left side, The frequency distribution produced based on the azimuth | direction shown on the left side is shown on the right side. 5A and 5B show statistical processing for the GPS satellite direction and the self-contained navigation direction.

  That is, FIG. 5A shows the GPS satellite azimuth Gds1 at a certain time and the autonomous navigation azimuth Nd1 at the same time. In this example, the azimuth difference between the self-contained navigation azimuth Nd1 and the GPS satellite azimuth Gds1 is 10 °, the GPS azimuth accuracy information is 40, and the GPS azimuth accuracy information is multiplied by 2. 80 is the total number of frequencies related to the orientation difference of 10 °. Further, in this example, a configuration is adopted in which the frequency is smoothed so that the frequency distribution has one peak even with a relatively small number of samples, and the difference in direction between the self-contained navigation azimuth Nd1 and the GPS satellite azimuth Gds1 is adopted. The frequency is set to ½ of the total number of frequencies described above. Further, the frequency of the azimuth difference obtained by adding 10 ° to the azimuth difference between the self-contained navigation azimuth Nd1 and the GPS satellite azimuth Gds1 is set to 1/4 of the total number of the above-mentioned degrees, and the self-contained navigation azimuth Nd1 and the GPS satellite azimuth Gds1 The frequency of the azimuth difference obtained by subtracting 10 ° from the azimuth difference is set to ¼ of the total number of the above-mentioned frequencies. That is, a distribution (Pa shown in FIG. 5A) having a significant frequency is formed around the direction difference between the self-contained navigation azimuth Nd1 and the GPS satellite azimuth Gds1, and statistical processing applied to the frequency distribution is performed.

  FIG. 5B shows the GPS satellite azimuth Gds2 at the time next to the GPS satellite azimuth Gds1 and the autonomous navigation azimuth Nd2 at the same time as the GPS satellite azimuth Gds2. In this example, the azimuth difference between the self-contained navigation azimuth Nd2 and the GPS satellite azimuth Gds2 is 20 °, and the GPS azimuth accuracy information is 20. Accordingly, the total number of frequencies related to the heading difference of 20 ° is 40, and the power of 20 °, which is the heading difference between the self-contained navigation heading Nd2 and the GPS satellite heading Gds2, is 20 (40 × 1/2). Further, the frequency of the heading difference obtained by adding 10 ° to the heading difference of 20 °, which is the heading difference between the autonomous navigation heading Nd2 and the GPS satellite heading Gds2, and the power of the heading difference obtained by subtracting 10 ° are 10 (40 × 1/4). It is. In FIG. 5B, the manner in which the distribution Pb created in this way is added to the frequency distribution is shown by hatching.

  When the frequency distribution is created based on the GPS satellite azimuth as described above, the control unit 20 performs processing for adding the frequency based on the azimuth difference between the GPS coordinate azimuth and the autonomous navigation azimuth to the frequency distribution. . For this purpose, first, the control unit 20 acquires the azimuth difference between the GPS coordinate azimuth and the self-contained navigation azimuth by the processing of the self-contained navigation information correction unit 21d (step S220). Specifically, the control unit 20 specifies a vector connecting two adjacent GPS positions based on GPS information acquired within a predetermined period, acquires a GPS coordinate direction based on the vector, and acquires a predetermined period. The self-contained navigation direction is acquired based on the self-contained navigation information acquired in the inside, and the difference between the directions at the same time is acquired.

  Furthermore, the control unit 20 performs weighting according to the GPS accuracy by the process of the self-contained navigation information correction unit 21d, and adds the azimuth difference to the frequency distribution (step S225). Since the azimuth between GPS coordinates is specified by two adjacent GPS positions, the accuracy of the azimuth between GPS coordinates depends on the accuracy of the GPS position. For this reason, the accuracy of the azimuth between GPS coordinates is specified by the geometric mean of the GPS accuracy related to the two adjacent GPS positions referred to when specifying the azimuth between the GPS coordinates. And the control part 20 determines the frequency of each azimuth | direction difference acquired in step S220 so that it may become so large that the precision of the said direction between GPS coordinates becomes high, and adds it to frequency distribution. In this case as well, the frequency only needs to be set so that the higher the GPS accuracy, the higher the frequency. In this example, the above-described geometric average itself is the frequency.

5C and 5D show statistical processing for the GPS coordinate direction and the self-contained navigation direction. That is, FIG. 5C shows the GPS coordinate direction Gdc1 at a certain time and the self-contained navigation direction Nd1 at the same time. In this example, the azimuth difference between the self-contained navigation azimuth Nd1 and the GPS coordinate azimuth Gdc1 is −10 °, and the GPS position accuracy information for specifying the GPS coordinate azimuth is 60 and 80. A state in which the azimuth accuracy information is 69 (= (60 × 80) ^1/2 ) is shown. In this case, the accuracy information 69 of the GPS coordinate azimuth is the total number of frequencies related to the azimuth difference −10 °, the frequency of the azimuth difference −10 ° is a value obtained by multiplying 69 by 1/2, and the azimuth difference −20 ° and A distribution Pc created so that the frequency with an azimuth difference of 0 ° is multiplied by 1/4 of 69 is added to the frequency distribution.

Similarly, in FIG. 5D, the GPS coordinate azimuth Gdc2 and the GPS coordinate azimuth Gdc2 at the next time of the GPS coordinate azimuth Gdc1 and the self-contained navigation azimuth Nd2 at the same time are shown. In this example, the azimuth difference between the self-contained navigation azimuth Nd2 and the GPS coordinate azimuth Gdc2 is 20 °, and the GPS position accuracy information for identifying the GPS coordinate azimuth is 80 and 100. The accuracy information is 89 (= (80 × 100) ^1/2 ). In this case, 89, which is the accuracy information of the azimuth difference between GPS coordinates, is the total number of frequencies related to the azimuth difference of 20 °, the frequency of the azimuth difference of 20 ° is a value obtained by multiplying 89 by 1/2, A distribution Pd created so that the frequency of ° is a value obtained by multiplying ¼ of 89 is added to the frequency distribution. In FIG. 5C, the manner in which the distribution Pc is added to the frequency distribution is indicated by hatching, and in FIG. 5D, the manner in which the distribution Pd is added to the frequency distribution is indicated by hatching.

  Next, the control unit 20 acquires the rotation angle of the self-contained navigation trajectory Tn through the processing of the self-contained navigation information correction unit 21d (step S230). That is, the control unit 20 specifies the representative value of the heading difference based on the frequency distribution created by the processes of steps S215 and S225, and acquires the representative value as the rotation angle of the autonomous navigation locus Tn. Specifically, the control unit 20 regards the azimuth difference having the largest frequency in the frequency distribution created by the processing of steps S215 and S225 as the representative value of the azimuth difference between the self-contained navigation trajectory Tn and the GPS trajectory, and the representative value is It is regarded as a rotation angle (α shown in FIG. 4C) when rotating the self-contained navigation trajectory so that the degree of coincidence between the self-contained navigation trajectory and the GPS trajectory becomes the highest. Therefore, in the present embodiment, the rotation angle is the first correction amount of the self-contained navigation direction for making the degree of coincidence between the self-contained navigation locus and the GPS locus the highest.

  Furthermore, the control part 20 acquires the reliability of GPS azimuth | direction by the process of the autonomous navigation information correction | amendment part 21d (step S235). In the present embodiment, the control unit 20 acquires the reliability of the GPS orientation based on the frequency distribution created by the processes in steps S215 and S225. Specifically, the control unit 20 specifies the azimuth difference having the largest frequency in the frequency distribution created by the processes in steps S215 and S225, the variance in the azimuth difference, and the total frequency in the frequency distribution. The reliability is higher so that the degree of misorientation with the largest frequency is higher, the reliability is higher as the variance in the misorientation is smaller, and the reliability is higher as the total frequency in the frequency distribution is larger. Set. In the present embodiment, the reliability of the GPS bearing is defined in five stages. That is, a map (not shown) that can specify the reliability according to the frequency of the azimuth difference having the largest frequency, the variance in the azimuth difference, and the total frequency in the frequency distribution is created, and the control unit 20 adds the map to the map. Based on this, the reliability of the GPS bearing is acquired.

  Next, the control unit 20 rotates the self-contained navigation trajectory by the processing of the self-contained navigation information correction unit 21d (step S240). That is, the control unit 20 rotates the self-contained navigation trajectory by the rotation angle acquired in step S230 while maintaining the shape of the self-contained navigation trajectory within a prescribed coordinate system. For example, as shown in FIG. 4C, the self-contained navigation trajectory Tnr is obtained by rotating the self-contained navigation trajectory Tn by the angle α around the most recent self-contained navigation position in the self-contained navigation trajectory Tn in the xy coordinate system.

  Next, the control unit 20 acquires a position difference between the GPS position and the self-contained navigation position by the process of the self-contained navigation information correction unit 21d (step S245). That is, the control unit 20 acquires a plurality of GPS positions and self-contained navigation positions based on the GPS information acquired within a predetermined period and the self-contained navigation information rotated in step S240, and calculates the position difference between the positions at the same time. get.

  Further, the control unit 20 creates a frequency distribution of the position difference by performing weighting according to the GPS accuracy by the process of the self-contained navigation information correction unit 21d (step S250). That is, since the accuracy of each GPS position is specified by the GPS accuracy information, the higher the accuracy of the GPS position, the greater the GPS information contributes to the determination of the position difference. Specifically, the control unit 20 acquires the GPS accuracy, determines the frequency of each positional difference acquired in step S245 so that the higher the accuracy of the GPS position, the higher the frequency, and adds it to the frequency distribution. In this way, a frequency distribution is created. The frequency only needs to be set so that the frequency increases as the accuracy of the GPS position increases. For example, the accuracy information itself of the GPS position that is standardized so as to increase as the accuracy increases can be used as the frequency. A value obtained by multiplying a predetermined coefficient may be used as the frequency.

  6A to 6C are diagrams for explaining statistical processing for specifying a positional difference between the GPS position and the autonomous navigation position and a representative value thereof. In these FIG. 6A to FIG. 6C, positions on the left side are indicated by black circles on the left side, and frequency distributions created based on the positions shown on the left side are shown on the right side. FIG. 6A shows a GPS position Gp1 at a certain time and a self-contained navigation position Np1 at the same time. 6B shows the GPS position Gp2 at the time next to the GPS position Gp1 and the autonomous navigation position Np2 at the same time as the GPS position Gp2. In FIG. 6C, the GPS position Gp3 at the time next to the GPS position Gp2 and The self-contained navigation position Np3 at the same time as the GPS position Gp3 is shown.

  As shown in these examples, the position difference between the self-contained navigation position Np1 and the GPS satellite position Gp1 can be defined by the position difference Δx in the x-axis direction and the position difference Δy in the y-axis direction. A frequency distribution is created for each difference. On the right side of FIGS. 6A to 6C, the positional difference Δx in the x-axis direction will be described. Of course, the process for creating the frequency distribution is the same for the position difference in the y-axis direction, except that the difference acquired as the position difference is different.

  Even in the case of position differences, when creating a frequency distribution, the total number of frequencies related to the position difference is specified from the accuracy information of the GPS position, and the frequency is smoothed so that the frequency distribution has one peak even with a relatively small number of samples. The structure to be adopted is adopted. In the example shown in FIG. 6A, the position difference Δx between the self-contained navigation position Np1 and the GPS satellite position Gp1 is 80, and the accuracy information of the GPS position is 80. In the example shown in FIG. 6B, the position difference Δx between the self-contained navigation position Np2 and the GPS satellite position Gp2 is 60, and the accuracy information of the GPS position is 40. Further, in the example shown in FIG. 6C, the position difference Δx between the self-contained navigation position Np3 and the GPS satellite position Gp3 is 40, and the accuracy information of the GPS position is 100.

  Regarding the GPS position, the accuracy information of the GPS position is configured to be the total number of frequencies, and in FIG. 6A, the accuracy information of the GPS position is 80. Therefore, the position difference between the autonomous navigation position Np1 and the GPS position Gp1 The total number of frequencies for Δx = 80 is 80. The frequency of the position difference Δx = 80 is 1/2 of the total number of frequencies, the frequency of the position difference obtained by adding 20 to the position difference Δx between the autonomous navigation position Np1 and the GPS satellite position Gp1, and the position difference obtained by subtracting 20 The frequency of is 1/4 of the total number of frequencies. As described above, a distribution (Pe shown in FIG. 6A) having a significant frequency is formed around the position difference Δx between the self-contained navigation position Np1 and the GPS satellite position Gp1, and added to the frequency distribution.

  In FIG. 6B, since the accuracy information of the GPS position is 40, the total number of frequencies related to the position difference Δx = 60 between the self-contained navigation position Np2 and the GPS position Gp2 is 40. Then, the distribution Pf is formed with the frequency of the positional difference Δx = 60 being 1/2 of the total number of frequencies and the frequency of the positional difference Δx = 40, 80 being 1/4 of the total number of frequencies. In FIG. 6C, since the accuracy information of the GPS position is 100, the total number of frequencies related to the position difference Δx = 40 between the self-contained navigation position Np3 and the GPS position Gp3 is 100. Then, the distribution Pg is formed with the frequency of the positional difference Δx = 40 being 1/2 of the total number of frequencies and the frequency of the positional difference Δx = 20, 60 being 1/4 of the total number of frequencies. In FIGS. 6A to 6C, the manner in which the distributions Pe to Pg are added to the frequency distribution is indicated by hatching.

  As described above, when the frequency distribution is created based on the GPS position, the control unit 20 acquires the amount of movement of the self-contained navigation trajectory by the processing of the self-contained navigation information correction unit 21d (step S255). That is, the control unit 20 specifies the representative value of the position difference based on the frequency distribution created by the process of step S250, and acquires the representative value as the movement amount of the autonomous navigation locus. Specifically, the control unit 20 specifies the position difference having the largest frequency in the frequency distribution created by the processing in step S250 for each of Δx and Δy, and represents each of them as a representative of the position difference between the autonomous navigation locus and the GPS locus. Consider it a value. The control unit 20 then moves the x-axis direction and the y-axis direction when moving the self-contained navigation trajectory so that the representative value has the highest degree of coincidence between the self-contained navigation trajectory and the GPS trajectory. (X and Y shown in FIG. 4C). Therefore, in the present embodiment, the amount of movement in the x-axis direction and the amount of movement in the y-axis direction are the first correction amount of the self-contained navigation position for making the degree of coincidence between the self-contained navigation locus and the GPS locus the highest. Become.

  Furthermore, the control part 20 acquires the reliability of a GPS position by the process of the autonomous navigation information correction | amendment part 21d (step S260). In this embodiment, the control part 20 acquires the reliability of GPS position based on the frequency distribution created by the process of step S250. Specifically, the control unit 20 specifies the position difference having the largest frequency in the frequency distribution created by the process of step S250, the variance in the position difference, and the total frequency in the frequency distribution. The reliability is higher so that the frequency of the position difference having the largest frequency is higher, the reliability is higher as the variance in the position difference is smaller, and the reliability is higher as the total frequency in the frequency distribution is larger. Set. In the present embodiment, the reliability of the GPS position is defined in five stages. That is, a map (not shown) that can specify the reliability according to the frequency of the position difference with the highest frequency, the variance in the position difference, and the total frequency in the frequency distribution is created, and the control unit 20 adds the map to the map. Based on this, the reliability of the GPS position is acquired.

  Since the reliability of the GPS azimuth and the GPS position as described above is specified based on statistics of a plurality of GPS azimuths and GPS positions, the reliability of the GPS trajectory is shown. On the other hand, the GPS accuracy information indicates the accuracy of each GPS azimuth and GPS position. Therefore, the reliability of the GPS azimuth and the GPS position is specified by performing statistical processing based on the GPS accuracy information, and the reliability as a locus specified from the GPS information can be defined. Note that the method of specifying the reliability is not limited to the above method, and for example, a configuration using the GPS accuracy information as the reliability may be employed.

  When the comparison process between the GPS trajectory and the self-contained navigation trajectory is performed as described above, the control unit 20 returns to the process shown in FIG. That is, the control unit 20 compares the reliability of the GPS azimuth with the reliability of the matching azimuth by the processing of the autonomous navigation information correction unit 21d, and determines whether the reliability of the GPS azimuth is equal to or higher than the reliability of the matching azimuth. Is determined (step S110). When it is determined in step S110 that the reliability of the GPS azimuth is equal to or higher than the reliability of the matching azimuth, the control unit 20 sets a correction target for the azimuth using the GPS trajectory as a correction target trajectory (step S115). . That is, the control unit 20 rotates the self-contained navigation trajectory by the rotation angle acquired in step S230, moves the self-contained navigation trajectory by the movement amount acquired in step S255, and the self-contained navigation direction at the latest time point is determined. The direction oriented as a result of rotation and movement is set as the correction target direction. For example, in the self-contained navigation trajectory Tnm after rotation and movement as shown in FIG. 4C, the direction in which the tip of the arrow is oriented becomes the direction in which the self-contained navigation direction at the latest time is oriented as a result of rotation and movement. The direction in which the tip of the arrow is oriented is the correction target direction.

  On the other hand, if it is not determined in step S110 that the reliability of the GPS azimuth is equal to or higher than the reliability of the matching azimuth, the control unit 20 sets an azimuth correction target using the matching trajectory as a correction target trajectory (step S120). That is, the control unit 20 sets the matching direction as the correction target direction.

  Next, the control unit 20 compares the reliability of the GPS position with the reliability of the matching position by the process of the autonomous navigation information correction unit 21d, and determines whether the reliability of the GPS position is equal to or higher than the reliability of the matching position. Is determined (step S125). When it is determined in step S125 that the reliability of the GPS position is equal to or higher than the reliability of the matching position, the control unit 20 sets a position correction target using the GPS path as a correction target path (step S130). . That is, the control unit 20 rotates the self-contained navigation trajectory by the rotation angle acquired in step S230, moves the self-contained navigation trajectory by the movement amount acquired in step S255, and the self-contained navigation position at the latest time point is determined. A position that exists as a result of rotation and movement is set as a correction target position. For example, in the self-contained navigation trajectory Tnm after rotation and movement as shown in FIG. 4C, the position of the tip of the arrow (the position of the white circle shown in FIG. 4C) is the result of rotation and movement at the latest time. Since the position exists, the position where the tip of the arrow exists becomes the position of the correction target.

  On the other hand, if it is not determined in step S125 that the reliability of the GPS position is greater than or equal to the reliability of the matching position, the control unit 20 sets a position correction target using the matching path as a correction target path (step S135). That is, the control unit 20 sets the matching position as the correction target position.

  Next, the control unit 20 acquires the reliability of the correction target by the process of the self-contained navigation information correction unit 21d (step S140). In this embodiment, since the correction target is set for each of the azimuth and the position, the reliability is acquired for each of the azimuth and the position. Here, the reliability of the correction target is the reliability of the correction target trajectory. Therefore, when step S115 is executed, the reliability of the GPS position is obtained as the reliability of the correction target of the position, and when step S120 is executed, the reliability of the matching position is acquired. Further, when step S130 is executed, the reliability of the GPS azimuth is acquired, and when step S135 is executed, the reliability of the matching azimuth is acquired as the reliability of the azimuth correction target.

  Further, the control unit 20 executes a self-contained navigation information correction process for correcting the self-contained navigation direction and the self-contained navigation position by the process of the self-contained navigation information correction unit 21d (step S145). FIG. 7 is a flowchart showing a self-contained navigation information correction process. In the self-contained navigation information correction process, the control unit 20 first acquires the second correction amount of the self-contained navigation position (step S300). In the present embodiment, the reliability of the corrected target locus (GPS locus or matching locus), the value obtained by subtracting the reliability of the autonomous navigation locus from the reliability of the corrected target locus, and the magnitude of the first correction amount of the autonomous navigation position. Based on this, the second correction amount for the self-contained navigation position is determined.

  Specifically, the higher the reliability of the corrected target locus, the larger the second correction amount of the self-contained navigation position, and the larger the value obtained by subtracting the reliability of the independent navigation locus from the reliability of the corrected target locus, It is defined in a map (not shown) so that the second correction amount increases and the second correction amount of the self-contained navigation position increases as the first correction amount of the self-contained navigation position increases. However, in the map, the second correction amount of the self-contained navigation position is smaller than the first correction amount. The control unit 20 acquires the second correction amount of the self-contained navigation position based on the map.

  That is, the higher the reliability of the corrected target locus, the lower the probability that the correction of the self-contained navigation locus that matches the corrected target locus is incorrect. Therefore, by setting the magnitude of the second correction amount of the self-contained navigation position so as to increase as the reliability of the GPS trajectory increases, the probability of erroneous correction is suppressed, and the accuracy of the self-contained navigation trajectory is improved early. It becomes possible to improve.

  The corrected target trajectory is a reference for specifying the first correction amount of the self-contained navigation position for making the degree of coincidence between the self-contained navigation trajectory and the corrected target trajectory the highest, and the self-contained navigation trajectory is a correction target. Therefore, the value obtained by subtracting the reliability of the self-contained navigation trajectory from the reliability of the correction target trajectory is the value obtained by subtracting the reliability of the correction target from the reliability of the correction reference, and the correction reference is more reliable than the correction target. The higher the degree, the larger the value. Therefore, by setting the magnitude of the second correction amount of the self-contained navigation position so that the larger the value obtained by subtracting the reliability of the self-contained navigation trajectory from the reliability of the correction target trajectory, the error correction probability is suppressed. In addition, the accuracy of the self-contained navigation trajectory can be improved early. Note that the reliability of the self-contained navigation path only needs to indicate the reliability of information such as the self-contained navigation position and the self-contained navigation direction. In this embodiment, the correction for the self-contained navigation position and the self-contained navigation direction is repeatedly performed, and the correction reference referred to in the correction performed in the past is the past correction target locus. Therefore, after the correction is performed in the present embodiment, the reliability of the past correction target trajectory serving as the reference for the correction is regarded as the reliability of the corrected independent navigation position and the independent navigation direction.

  Furthermore, the magnitude of the first correction amount of the self-contained navigation position indicates the degree of deviation between the self-contained navigation locus and the correction target locus. Therefore, the accuracy of the self-contained navigation trajectory can be improved early by setting so that the larger the difference between the two is, the larger the second correction amount of the self-contained navigation position is. In the present embodiment, since the reliability of the correction target locus and the self-contained navigation locus is defined in five stages, the higher the reliability of the correction target locus, the larger the second correction amount of the self-contained navigation position in steps. The second correction amount of the self-contained navigation position may be configured to increase stepwise as the value obtained by subtracting the reliability of the self-contained navigation track from the reliability of the corrected target track is large. Further, the second correction amount of the self-contained navigation position may be continuously changed according to the magnitude of the first correction amount of the self-contained navigation position, or may be configured to change stepwise. Good.

  Of course, the higher the reliability of the correction target trajectory, the larger the second correction amount tends to be. Here, the larger the value obtained by subtracting the reliability of the self-contained navigation trajectory from the reliability of the correction target trajectory, the greater the tendency of the second correction amount. The larger the first correction amount of the self-contained navigation position is, the larger the second correction amount tends to be. Therefore, the second correction amount may be configured to change m steps (m is a natural number) in accordance with changes in n steps (n is a natural number) of reliability. Furthermore, the second correction amount is one of the reliability of the correction target locus, a value obtained by subtracting the reliability of the autonomous navigation locus from the reliability of the correction target locus, and the magnitude of the first correction amount of the autonomous navigation position, or The configuration may be such that the second correction amount is determined according to the combination of the two.

  As described above, when the second correction amount of the self-contained navigation position is acquired in step S300, the control unit 20 determines whether the second correction amount of the self-contained navigation position is greater than 0 by the processing of the self-contained navigation information correction unit 21d. If it is not determined in step S305 that the second correction amount of the self-contained navigation position is greater than 0, step S310 is skipped.

When it is determined in step S305 that the second correction amount of the self-contained navigation position is greater than 0, the control unit 20 corrects the self-contained navigation position with the second correction amount of the self-contained navigation position (step S310). FIG. 8A is a diagram exemplifying how the self-contained navigation position is corrected by the second correction amount of the self-contained navigation position. In FIG. 8A, an example of the position correction target Gp when the correction target locus is a GPS locus is indicated by a one-dot chain line, and the self-contained navigation position Np before correction is indicated by a broken circle. In this example, the first correction amount of the self-contained navigation position Np is A ₁ and the second correction amount is A ₂ .

As shown in Figure 8A, the second correction amount A ₂ is configured to implement part of the correction achieved by the first correction amount A _1, and the first correction amount A ₁ size smaller. That is, the correction realized by the first correction amount A ₁ is, for example, correction for moving the self-contained navigation position Np along the vector V ₁ from the self-contained navigation position Np toward the correction target Gp. The vector V ₁ is, for example, In the example shown in FIG. 4C, the length along the x axis is X, and the length along the y axis is Y. On the other hand, the second correction amount A ₂ corresponds to correction for moving the self-contained navigation position Np along a vector V ₂ obtained by multiplying the vector V ₁ by a coefficient C (0 <C <1). The coefficient C is specified in step S310 described above. Therefore, for example, in the example shown in FIG. 4C, the correction by the second correction amount A ₂ can be specified by multiplying each of X and Y by the coefficient C indicating the magnitude of the second correction amount, and is indicated by a solid line. The corrected self-contained navigation position Npa is specified as shown in the circle. In the above correction, the control unit 20 has the second correction amount A _{2 that} is smaller than the first correction amount A ₁ that is the correction amount of the self-contained navigation position for making the degree of coincidence between the self-contained navigation locus and the correction target locus the highest. The self-contained navigation position is corrected by this. Therefore, even if an incorrect correction is made, the correction can be performed so that the influence is suppressed, and the accuracy of the self-contained navigation locus can be easily improved.

  Next, the control part 20 correct | amends the self-contained navigation direction in step S315,320. In this embodiment, the control part 20 employ | adopts the structure which correct | amends the self-contained navigation direction with a fixed correction amount. Therefore, the control unit 20 determines whether or not the reliability of the azimuth correction target satisfies a predetermined criterion by the process of the self-contained navigation information correction unit 21d (step S315). In the present embodiment, the control unit 20 determines that the reliability of the azimuth correction target is predetermined when the reliability of the azimuth correction target is equal to or higher than a predetermined minimum reliability required as a correction reference. It is determined that the standard is satisfied.

  In step S315, if it is not determined that the reliability of the azimuth correction target satisfies the predetermined criterion, step S320 is skipped. On the other hand, if it is determined in step S315 that the reliability of the azimuth correction target satisfies a predetermined standard, the control unit 20 uses the second correction amount of the self-contained navigation azimuth by the processing of the self-contained navigation information correction unit 21d. The self-contained navigation direction is corrected (step S320). Here, the second correction amount of the self-contained navigation direction is a constant value. For example, when the first correction amount is larger than 1 °, the second correction amount is set to 1 ° or the like, and the first correction amount is 1 °. In the following cases, the second correction amount is set to 0 ° or the like.

FIG. 8B is a diagram illustrating a state in which the self-contained navigation direction is corrected by the second correction amount of the self-contained navigation direction. In FIG. 8B, the correction target Gd of the azimuth when the correction target locus is a GPS locus is illustrated by a dashed-dotted straight arrow, and the self-contained navigation azimuth Nd before correction is illustrated by a broken straight arrow. In this example, the first correction amount of the self-contained navigation direction Nd is A ₁ , and the second correction amount is A ₂ .

That is, the correction realized by the first correction amount A ₁ is, for example, correction that rotates the self-contained navigation azimuth Nd to coincide with the correction target Gd, and the rotation angle is α in the example shown in FIG. 4C, for example. is there. On the other hand, the correction realized by the second correction amount A ₂ is, for example, a correction in which the self-contained navigation azimuth Nd is rotated by 1 ° to approach the correction target Gd. In the above correction, the control unit 20 has a constant value smaller than the first correction amount A ₁ that is the correction amount of the self-contained navigation direction in order to maximize the degree of coincidence between the self-contained navigation locus and the correction target locus. thereby correcting the self-contained navigation orientation by the second correction amount a _2. Therefore, the accuracy of the self-contained navigation trajectory can be improved in a state where the influence due to the erroneous correction is suppressed to be extremely small.

  When the self-contained navigation information is corrected by the above process, the process returns to the process shown in FIG. 2, and the control unit 20 updates the reliability of the self-contained navigation information by the process of the self-contained navigation information correction unit 21d (step S150). . That is, the control unit 20 sets the reliability of the correction target of the position acquired in step S140 as the reliability of the autonomous navigation position, and sets the reliability of the correction target of the direction acquired in step S140 as the reliability of the autonomous navigation azimuth. To do.

(3) Other embodiments:
The above embodiment is an example for carrying out the present invention, and various other embodiments can be adopted. For example, as long as the reliability of the GPS trajectory is high enough to be a reference for correction, a configuration in which the GPS trajectory is a correction target trajectory may be employed. That is, when the reliability of the GPS trajectory exceeds a predetermined reference, the GPS trajectory is used as the correction target trajectory, and when the reliability of the GPS trajectory is lower than the predetermined reference, the matching trajectory is used as the correction target trajectory. A configuration in which the self-contained navigation information is corrected so as to reduce the difference between the correction target trajectory and the correction target locus may be employed.

  Such a configuration can be realized by changing steps S110 and S125 of FIG. 2 in the above-described embodiment. Specifically, in step S110, the control unit 20 determines whether or not the reliability of the GPS azimuth satisfies a predetermined standard. If the predetermined standard is satisfied, step S115 is performed. If the predetermined standard is not satisfied, step S120 is performed. Configure to run. In step S125, the control unit 20 determines whether or not the reliability of the GPS position satisfies a predetermined criterion, and executes step S130 when the predetermined criterion is satisfied, and executes step S135 when the predetermined criterion is not satisfied. Configure as follows. The predetermined standard may be a standard for determining whether or not the accuracy of the self-contained navigation information can be improved when the self-contained navigation information is corrected using the GPS trajectory as the correction target trajectory. For example, a configuration in which a lower limit value is set and whether or not the lower limit value is exceeded can be adopted.

  According to this configuration, when the reliability of the GPS trajectory is high, the GPS trajectory is used as the correction target trajectory, and the self-contained navigation information can be corrected based on more objective information. That is, the GPS trajectory is generated based on the GPS information, and the GPS information does not depend on the self-contained navigation information, but the matching trajectory is identified based on the comparison between the self-contained navigation trajectory and the shape of the road indicated by the map information. Depends on navigation information. Therefore, if the reliability of the GPS information is high and accurate, the GPS information is suitable as a reference for correcting the self-contained navigation information. Therefore, if the reliability of the GPS trajectory exceeds a predetermined standard, the GPS trajectory is corrected based on highly reliable information that does not depend on the self-contained navigation information. It becomes possible to correct the navigation information.

  Furthermore, in the above-described embodiment, either the GPS trajectory or the matching trajectory is selected as the correction target trajectory based on the reliability, but neither the GPS trajectory nor the matching trajectory is suitable as a correction reference. If the reliability is as low as possible (the accuracy of the self-contained navigation information does not improve even if correction is performed), a configuration may be adopted in which correction is not performed.

  Furthermore, in the configuration in which the GPS trajectory is a correction target trajectory when the reliability of the GPS trajectory exceeds a predetermined reference, it is not essential to specify the reliability of the matching trajectory in order to determine the correction target trajectory. A configuration that specifies the reliability of the matching trajectory for purposes other than the purpose of determining the correction target trajectory may be employed. For example, when the matching trajectory becomes the correction target trajectory, the reliability of the matching trajectory is specified in order to specify the reliability of the corrected self-contained navigation trajectory, and the reliability is so low that it is not suitable as a correction criterion. In order to determine whether or not, the reliability of the matching trajectory may be specified.

  Furthermore, in the above-described embodiment, the position and orientation correction target is set based on the correction trajectory, and the second correction amount is set so as to gradually approach the correction target. The self-contained navigation information may be corrected without setting a typical correction target value. For example, a configuration that specifies a rotation angle and a movement amount of a self-contained navigation track so that at least the self-contained navigation track approaches a correction target track, and corrects the self-contained navigation information so as to perform rotation and movement corresponding to the rotation angle and movement amount. It is also good.

  Furthermore, in the above-described embodiment, the reliability is defined separately for the position and the azimuth, and the correction is performed in a state where the separate trajectory can be a reference for correction. However, the correction reference is a single trajectory. You may comprise as follows. For example, the reliability is defined for each of the GPS position, the GPS azimuth, the matching position, and the matching azimuth. If the highest reliability is the reliability of the GPS position, the GPS azimuth is also set as the correction target for the azimuth. A configuration in which only the GPS trajectory is used as a reference for correction can be employed.

  Furthermore, in the above-described embodiment, the GPS azimuth is specified based on the GPS satellite azimuth and the azimuth between GPS coordinates. However, either the GPS satellite azimuth or the azimuth between GPS coordinates may be used as the GPS azimuth. Furthermore, in the above-described embodiment, the second correction amount is determined by a method that differs depending on the position and orientation, but the same method may be used, or a method opposite to the above-described embodiment may be used. For example, the second correction amount of the position may be a constant movement amount, and the second correction amount of the azimuth increases as the reliability of the correction target trajectory increases, and the reliability of the autonomous navigation trajectory is determined from the reliability of the correction target trajectory. You may comprise so that it may become so large that the value which reduced the degree is large, and so that it may become so large that the 1st correction amount of a self-supporting navigation direction is large.

  Furthermore, other sensors such as an acceleration sensor may be added as a sensor for acquiring the self-contained navigation information. Further, the representative values of the GPS position and the GPS direction may be statistical average values or median values. Furthermore, the processing for comparing the self-contained navigation track and the GPS track may be changed according to the accuracy of the GPS information. For example, as the accuracy of the GPS information indicated by the GPS accuracy information is higher, a configuration in which a self-contained navigation locus with a shorter distance and a GPS locus are compared can be employed. In other words, since the accumulated error increases with time in the self-contained navigation trajectory, if the accuracy of the GPS information is high, the accuracy of the GPS information is higher than the acquisition of a large number of time-series GPS information to improve the statistical accuracy. It is better to compare with the self-contained navigation trajectory while it is high. Therefore, the accuracy of the self-contained navigation trajectory can be easily increased if the self-contained navigation trajectory with a shorter distance is compared with the GPS trajectory as the accuracy of the GPS information is higher.

  DESCRIPTION OF SYMBOLS 10 ... Trajectory information generation apparatus, 20 ... Control part, 21 ... Navigation program, 21a ... Independent navigation locus acquisition part, 21b ... Matching locus acquisition part, 21c ... GPS locus acquisition part, 21d ... Independent navigation information correction part, 30 ... Recording Medium 30a ... Map information 40 ... User I / F unit 41 ... Vehicle speed sensor 42 ... Gyro sensor 43 ... GPS receiver

Claims (9)

A self-contained navigation trajectory acquisition means for acquiring a self-contained navigation trajectory that is a vehicle trajectory indicated by time-series self-contained navigation information;
The matching trajectory, which is the trajectory of the vehicle in time series identified by the map matching process in which the road on which the self-contained navigation trajectory and the road shape indicated by the map information most closely match is regarded as the road on which the vehicle is traveling, is determined. Matching trajectory acquisition means for acquiring along with the reliability of the trajectory;
GPS trajectory acquisition means for acquiring a GPS trajectory that is the trajectory of the vehicle indicated by time-series GPS information;
Comparing the self-contained navigation trajectory with the GPS trajectory to obtain the reliability of the GPS trajectory, of the GPS trajectory and the matching trajectory as a correction target trajectory, and the self-contained navigation trajectory Self-contained navigation information correcting means for performing correction to rotate and move the self-contained navigation trajectory so as to reduce the difference from the correction target trajectory;
A trajectory information generation device comprising: A self-contained navigation trajectory acquisition means for acquiring a self-contained navigation trajectory that is a vehicle trajectory indicated by time-series self-contained navigation information;
A matching trajectory that is a trajectory of the vehicle in time series identified by a map matching process in which the road on which the self-contained navigation trajectory and the shape of the road indicated by the map information most closely match is regarded as a road on which the vehicle is traveling is acquired. Matching trajectory acquisition means;
GPS trajectory acquisition means for acquiring a GPS trajectory that is the trajectory of the vehicle indicated by time-series GPS information;
The self-contained navigation trajectory and the GPS trajectory are compared to obtain the reliability of the GPS trajectory, and when the reliability of the GPS trajectory exceeds a predetermined reference, the GPS trajectory is used as a correction target trajectory, and the GPS trajectory When the reliability of the trajectory is less than a predetermined reference, the matching trajectory is used as a correction target trajectory, and correction is performed to rotate and move the self-contained navigation trajectory so that the difference between the self-contained navigation trajectory and the correction target trajectory is reduced. and dead reckoning navigation information correction means for performing,
A trajectory information generation device comprising: The self-contained navigation information correction means includes
The higher the degree of coincidence of the shape of the self-contained navigation trajectory and the GPS trajectory, the higher the reliability of the GPS trajectory,
The locus | trajectory information generation apparatus in any one of Claim 1 or Claim 2. The vehicle trajectory indicated by the self-contained navigation trajectory, the matching trajectory, and the GPS trajectory includes the azimuth and position of the vehicle,
The self-contained navigation information correction means specifies the correction target locus for each of the azimuth and position of the vehicle,
Correcting the azimuth of the vehicle indicated by the self-contained navigation information so that a difference between the azimuth of the vehicle indicated by the self-contained navigation locus and the azimuth of the vehicle indicated by the correction target locus is reduced;
Correcting the position of the vehicle indicated by the self-contained navigation information so as to reduce a difference between the position of the vehicle indicated by the self-contained navigation locus and the position of the vehicle indicated by the correction target locus;
The locus | trajectory information generation apparatus in any one of Claims 1-3. The self-contained navigation information correction means includes
The frequency of the azimuth difference between the azimuth of the vehicle indicated by the self-contained navigation information for a plurality of time points and the azimuth of the vehicle indicated by the GPS information for a plurality of time points is large, and the frequency of the mode value is large. , The smaller the variance, the higher the reliability of the direction of the vehicle indicated by the GPS trajectory,
The total frequency of the frequency distribution of the position difference between the position of the vehicle indicated by the autonomous navigation information for a plurality of time points and the position of the vehicle indicated by the GPS information for a plurality of time points is large, and the frequency of the mode value is large. , The smaller the variance, the higher the reliability of the position of the vehicle indicated by the GPS trajectory,
The trajectory information generation device according to claim 4. The matching trajectory acquisition means includes
Increasing the reliability of the position indicated by the matching trajectory as the road width of the road regarded as the road on which the vehicle is traveling is narrower;
The trajectory information generation device according to claim 1 or claim 3 . The matching trajectory acquisition means includes
Increasing the reliability of the direction indicated by the matching trajectory as the azimuth fluctuation of the vehicle in the process of performing the map matching process is smaller,
The trajectory information generation device according to claim 1 or claim 3 . A step of acquiring a self-contained navigation trajectory that is a trajectory of the vehicle indicated by the time-series self-contained navigation information by the self-contained navigation trajectory acquisition means ;
The time-series trajectory of the vehicle identified by the map matching process in which the matching trajectory acquisition unit regards the road where the self-contained navigation trajectory and the shape of the road indicated by the map information best match the road on which the vehicle is traveling. obtaining a certain matching trajectory with reliability of the matching trajectory,
The GPS trace obtaining means obtaining a GPS trajectory is a trajectory of the vehicle indicated by the GPS information of the time series,
The self-contained navigation information correction means compares the self-contained navigation trajectory with the GPS trajectory to obtain the reliability of the GPS trajectory, and corrects a higher reliability trajectory of the GPS trajectory and the matching trajectory. And performing correction to rotate and move the self-contained navigation trajectory so that the difference between the self-contained navigation trajectory and the correction target trajectory is reduced.
Trajectory information generation method including A self-contained navigation trajectory acquisition function for acquiring a self-contained navigation trajectory that is a vehicle trajectory indicated by time-series self-contained navigation information;
The matching trajectory, which is the trajectory of the vehicle in time series identified by the map matching process in which the road on which the self-contained navigation trajectory and the road shape indicated by the map information most closely match is regarded as the road on which the vehicle is traveling, is determined. Matching trajectory acquisition function to acquire with the reliability of the trajectory,
A GPS trajectory acquisition function for acquiring a GPS trajectory that is the trajectory of the vehicle indicated by time-series GPS information;
Comparing the self-contained navigation trajectory and the GPS trajectory to obtain the reliability of the GPS trajectory, and using the GPS trajectory and the matching trajectory as a corrected target trajectory, A self-contained navigation information correction function for performing correction to rotate and move the self-contained navigation trajectory so as to reduce the difference from the correction target trajectory;
Track information generation program that makes a computer realize.

Images