JPH08334346A - Present location calculating device - Google Patents

Abstract

PURPOSE: To reduce the processing load of a map matching means without impairing the map matching accuracy of the means by allowing the means to perform the estimation of the next virtual present location and calculation of candidate points associated with the estimation on candidate points other than that having the highest reliability. CONSTITUTION: The map matching means of a microprocessor 24 estimates the present location of its own vehicle as a virtual present location, based on the relative displacement of the vehicle found from the traveling azimuth and traveling distance of the vehicle, calculates candidate points of the present location of the vehicle along a road together with the reliability of the points indicating the certainty of the points by collating the virtual present location with road data in a memory 25, and recognizes the candidate point having the highest reliability as the present location of the vehicle. The map matching means estimates the next virtual present position from candidate points other than the candidate point having the highest reliability and calculates candidate points of the next virtual present location. When the candidate points of the next virtual present location is calculated, the map matching means deletes the candidate points having a calculated reliability lower than a preset level.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is mounted on a moving body such as a vehicle and measures the traveling distance and traveling direction of the moving body.
Accordingly, the present invention relates to a current position calculation device that calculates the current position of the moving body.

[0002]

2. Description of the Related Art Conventionally, in a current position calculating device for calculating the current position of a vehicle traveling on a road, the current position of the vehicle is the traveling direction of the vehicle measured by a direction sensor such as a gyro and the vehicle speed sensor or the distance. It is calculated based on the travel distance of the vehicle measured by the sensor.

Further, the mileage of a vehicle is generally a distance coefficient which is a distance traveled by the vehicle per one rotation of the tire by measuring the rotational speed of the output shaft of the transmission or the tire. It is calculated by multiplying by.

Further, in order to correct the error of the current position obtained from the traveling direction of the vehicle and the traveling distance in this way, as described in Japanese Patent Laid-Open No. 63-148115, it is obtained so as to match the road. The current position of the vehicle
A so-called map matching technique is known, and this map matching technique can improve the accuracy of current position calculation.

[0005]

However, it is necessary to execute a process of searching for all roads located within a predetermined range with respect to the virtual current position of the vehicle, and the time required for this is the predetermined time. Depending on the density of the roads located within the range, especially when the roads exist in the city area, the processing time becomes extremely long. Further, the processing time required to calculate the reliability of the current position candidate point obtained on the road also becomes long.

Since the processing time for searching the road and the processing time for calculating the reliability are remarkably lengthened, other processing to be executed by the device, such as reading of data of a direction sensor, a vehicle speed sensor or There is a problem in that it may not be possible to appropriately read the data of the distance sensor or appropriately calculate the traveling distance of the vehicle based on the vehicle speed sensor or the data of the distance sensor.

It is an object of the present invention to provide a current position calculation device capable of reducing the processing load without impairing the accuracy of map matching.

[0008]

SUMMARY OF THE INVENTION An object of the present invention is a current position calculating device mounted on a vehicle for calculating the current position of the vehicle, the direction detecting means for detecting the traveling direction of the vehicle, and the traveling distance of the vehicle. Based on the relative displacement obtained based on the traveling azimuth and the traveled distance, the vehicle current position is estimated as a virtual current position, The virtual current position is collated with the road data of the road data storage means, the candidate point of the current position on the road is calculated together with the reliability indicating its credibility, and the candidate point with the highest reliability is recognized as the current position. Map matching means for performing the next virtual current position estimation and calculation of the candidate points associated with the candidate points other than the candidate point having the highest reliability. , Where, calculated reliability is achieved by the current position calculating device and deletes candidate point lower than degree of advance.

In this apparatus, the candidate points lower than the predetermined degree are, for example, candidate points having a reliability lower than the reliability of the candidate point having the highest reliability by a predetermined amount or more.

The map matching means preferably calculates the reliability of a candidate point found on the road based on the distance to the road around the virtual current position and the heading difference between the road heading and the vehicle heading. To do.

When the candidate point on the road corresponding to the virtual current position outside the road is not found, the map matching means sets the virtual current position outside the road as a candidate point,
The predetermined degree used when determining whether or not to delete the candidate points outside the road can be set to be easily deleted as compared with the candidate points on the road.

[0012]

In the map matching process, in principle, the virtual current position is calculated based on the relative traveling distance and the heading newly obtained with respect to the current position of the vehicle, and the virtual current position is calculated in the vicinity. In contrast to road data,
The position on the road that is judged to be the most probable (highest reliability) is obtained, and this is recognized as the current position. But,
In order to prevent misidentification due to error, data of candidate points other than the candidate point of the current position that is judged to be the most probable is not immediately discarded, and new candidate points are calculated based on that data as well. The most probable candidate point among the candidate points of is identified as the current position.

If data of candidate points other than the most probable candidate points are held in this way and new candidate points are obtained from these, there is a possibility that the data to be held will increase to a considerable number. This may impose an excessive processing load on the processor in executing the map matching process to be processed in real time. This problem is particularly noticeable in areas with dense roads such as urban areas. In the present invention,
The confidence level given to a candidate point, which indicates its credibility, is checked, and the candidate point having a reliability level lower than a predetermined level is deleted.

Thus, while maintaining the reliability of the map matching process, the number of objects to be processed in the process,
That is, it is possible to reduce the processing load and the time required to execute the processing.

[0015]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the configuration of the present position calculating apparatus according to the embodiment of the present invention. As shown in FIG. 1, the present position calculation device 10 includes an angular velocity sensor 11 that detects a yaw rate of a vehicle to detect a change in traveling direction, and a geomagnetic sensor 12 that detects a traveling direction of a vehicle by detecting geomagnetism. And a vehicle speed sensor 13 that outputs pulses at time intervals proportional to the rotation of the output shaft of the vehicle transmission.

Further, a display 17 for displaying a map around the current position, a mark indicating the current position, etc., a switch 14 for accepting a command for switching the scale of the map displayed on the display 17 from the user (driver), a digital map. CD-ROM 15 for storing data and its CD
A driver 16 for reading map data from the ROM 15. Further, the controller 18 for controlling the operation of each peripheral device described above is provided. In the present embodiment, the digital map data described above includes road data composed of coordinates indicating the ends of a plurality of line segments,
Alternatively, it includes road width data indicating the road width of the road, a highway flag indicating whether the road is a highway or a general road, and the like.

The controller 18 includes an A / D converter 19 for converting the signal (analog) of the angular velocity sensor 11 into a digital signal, and an A / D converter 20 for converting the signal (analog) of the geomagnetic sensor 12 into a digital signal. , Vehicle speed sensor 1
A counter 26 that counts the number of pulses output from the CPU 3 every 0.1 seconds, a parallel I / O 21 that inputs whether or not the switch 14 is pressed, and a DMA (that transfers the map data read from the CD-ROM 15 ( Direct Memory Acce
ss) It has a controller 22 and a display processor 23 for displaying a map image on the display 17.

The controller 18 also has a microprocessor 24 and a memory 25. The microprocessor 24 outputs the signal of the angular velocity sensor 11 obtained through the A / D converter 19, the signal of the geomagnetic sensor 12 obtained through the A / D converter 20, and the output pulse of the vehicle speed sensor 13 counted by the counter 26. Number, whether or not the switch 14 is pressed through the parallel I / O 21, and the map data from the CD-ROM 15 obtained through the DMA controller 22 are received and processed based on those signals, The current position is calculated and displayed on the display 17 via the display processor 23. The display of the vehicle position is performed by superimposing an arrow mark or the like on the map which is already displayed on the display 17, as shown in FIG. This allows the user to know the current position of the vehicle on the map. The memory 25 is
It includes a ROM that stores a program that defines the contents of processing (described later) for realizing such operations, and a RAM that is used as a work area when the microprocessor 24 performs the processing.

The operation of the present position calculating device 10 thus configured will be described below.

The operation of the apparatus 10 is generally performed by a process of calculating the traveling direction and the traveling distance of the vehicle, a process of determining the current position of the vehicle from the calculated traveling direction and distance, and the obtained vehicle position and Since it can be divided into three processes, that is, the process of displaying the azimuth, these processes will be described sequentially.

FIG. 3 illustrates the flow of processing for calculating the heading and traveling distance of the vehicle.

This processing is carried out at a fixed cycle, for example, 100 m.
This is a routine of the microprocessor 24 that is activated and executed for each S.

In this routine, first, the A / D converter 1
The output value of the angular velocity sensor 11 is read from 9 (step 301). Since the azimuth change is output as the output value of the angular velocity sensor 11, only a relative value in the traveling direction of the vehicle can be detected. Therefore, next, the output value of the geomagnetic sensor 12 is read from the A / D converter 20 (step 30
2) The estimated azimuth of the vehicle is determined using the absolute azimuth calculated from the output value of the geomagnetic sensor 12 and the azimuth change (angular velocity output) output from the angular velocity sensor 11 (step 303).

This azimuth determination can be performed, for example, for a long time,
When the vehicle speed is low, the error of the angular velocity sensor is large, so
When the vehicle speed is low for a certain period of time or more, the method is performed by using only the geomagnetic sensor direction.

Next, the number of pulses output from the vehicle speed sensor 13 is counted by the counter 26 every 0.1 seconds, and the counted value is read (step 304). By multiplying the read value by the distance coefficient, the distance advanced in 0.1 second is obtained (step 305).

Next, the traveling distance value per 0.1 second thus obtained is integrated with the value obtained last time,
It is checked whether or not the traveling distance of the vehicle has reached 20 m (step 306).
No, the current process is terminated and a new process is started.

As a result of the traveling distance calculation processing, when the accumulated traveling distance becomes a fixed distance, for example, 20 m (Yes in step 306), the traveling direction and traveling distance (20 m) at that time are output. (Step 307).
In step 307, the integrated distance is further initialized,
Start a new mileage integration.

Next, the process of calculating the virtual current position of the vehicle on the basis of the calculated heading and traveling distance and obtaining the candidate point of the vehicle based on the calculated virtual current position will be described.

FIG. 4 shows the flow of this processing.

This processing is a routine of the microprocessor 24 which is started and executed in response to the output of the traveling direction and the traveling distance from FIG. That is, this process is started every time the vehicle advances 20 m.

In this process, first, the sensor data output in the previous step 307, that is, the traveling direction and the traveling distance are read (step 41).

Next, a map matching process is performed (step 42). This includes a process of calculating the reliability of candidate points, as will be described in detail later.

Next, a reliability value P lower than the reliability of the display candidate point which is the highest reliability candidate point obtained by the map matching process by a constant value Q is obtained (step 43). The value of P is a predetermined value of about 40 to 50 when the maximum value of reliability is 100. The value of Q may be switched depending on whether the candidate point is a matching state or a free state described later. In that case, the Q value for the free state candidate points is set to the Q value for the matching state candidate points so that the free state candidate points are more easily deleted.
Lower the value.

Next, the reliability of candidate points other than the display candidate points is compared with the value of P obtained above (step 44). If the reliability is equal to or higher than the P value (step 44, No),
If it is smaller than the P value (Yes in step 44), the process proceeds to step 45. In step 45, the candidate (and various information described later) is deleted from the memory. By this deletion, the process of obtaining a new candidate point from this candidate point in the next map matching process is omitted, and the processing load can be reduced.

The processing from step 44 onward is repeated for all candidate points other than the display candidate points, and this processing ends (step 46).

FIG. 5 shows the details of the map matching process.

In the map matching process, first,
The amount of movement of the vehicle is obtained separately in the latitude and longitude directions. Further, the movement amount in each of these directions is added to the position of the vehicle candidate point obtained in the previous process for obtaining the vehicle candidate point, and the virtual current position (position where the current vehicle is estimated to exist) ( A) is calculated (step 421).
Details of this candidate point will be described later. If there is no candidate point obtained in the previous process of finding candidate points for the vehicle, such as immediately after the start of the device, the position set separately is used as the position of the previously obtained candidate point and the virtual current is used. Find position (A).

Next, the road data of the obtained map around the virtual current position (A) is read from the CD-ROM 15 via the driver 16 and the DMA controller 23 (step 422). At this time, if the candidate point used to obtain the virtual current position (A) is a candidate point on the road, a line segment within the distance search range D from the virtual current position (A) represents the road. Select a line segment or a line segment connected to it. If the candidate point is not a candidate point on the road (free candidate point), a line segment within the distance search range D from the virtual current position (A) is selected. At this time, the road search range D may be variable based on the reliability of the candidate points used to obtain the virtual current position (corrected last time). That is, for the virtual current position obtained from the highly reliable candidate point, the line segment included in the narrower range is selected, and conversely, the virtual current position obtained from the less reliable candidate point is wider range. Select the line segment included in.
The reason for changing the search range based on the reliability is that the reliability of the accuracy of the current position obtained last time is considered to be low when the reliability is low, so a wider range is searched to find the road. This is because it is more appropriate to obtain the correct current position.

As described above, in this embodiment, road data is approximated by a plurality of line segments 51 to 55 connecting two points as shown in FIG.
The one represented by the coordinates of the start point and the end point is used. For example, the line segment 53 has its starting point (x3, y3)
And the end point (x4, y4).

Next, from the line segments extracted in step 422, only the line segments whose azimuth is within the predetermined traveling direction and the required traveling direction are selected (step 42).
3) Further, for all of the extracted n line segments, a perpendicular is drawn from the virtual current position (A), and the perpendicular L
The length of (n) is obtained (step 424).

Next, based on the lengths of these perpendiculars, the error cost value ec (n) defined by the following equation is calculated for all the line segments selected in step 423.

Ec (n) = α × | θcar−θ (n) | +
β | L (n) | where θcar is the vehicle direction at the virtual current position (A), θ (n) is the direction of the line segment, and L (n) is the direction from the virtual current position (A) to the line segment. Distance, ie the length of the perpendicular, α
And β are weighting factors. The values of these weighting factors are
It may be changed depending on which of the deviation between the traveling direction and the direction of the road and the deviation between the current position and the road is given priority in selecting the road to which the current position belongs. For example, when a road whose direction is close to the traveling direction is important, α is increased.

Here, the candidate points will be further described. In an initial state such as immediately after the start of the device, the virtual current position (A) is uniquely determined by the user (driver) inputting predetermined information using the switch 14 and the like. Is located on the line segment corresponding to the road. However, after the vehicle has traveled, the virtual current position (A) may not exist on the line segment corresponding to the road due to an error in a sensor such as a gyro. as a result,
For example, as shown in FIG. 7, when a road is branched, that is, when two line segments 64 and 65 appear from a node 68 of the line segment 61 corresponding to the road, which line segment corresponds to which line segment. Often it is not possible to determine if a vehicle is present on the road.

Therefore, in such a case, in this embodiment, predetermined points existing on two conceivable line segments are set as candidate points, and their current position, error cost, and accumulation to be described later. Error cost, etc.
The memory 25 is configured to be stored in a predetermined area of the RAM. In addition, in order to facilitate the description, in the following description, one or more new candidate points will be generated from a single candidate point, unless it is clearly indicated that there are a plurality of candidate points.

Then, the calculated error cost ec
According to (n) and the cumulative error cost es related to the candidate points obtained in the previous processing, the cumulative error cost es in the current processing defined by the following formula
(N) is calculated (step 425).

Es (n) = (1−k) × es + k × ec (n) Here, k is a weighting coefficient larger than 0 and smaller than 1. This accumulated error cost es (n) represents how much the error cost calculated in the previous process is reflected in the error cost calculated in this process.

Furthermore, the calculated cumulative error cost es
The reliability trst defined by the following equation based on (n)
(N) is calculated (step 425).

Trst (n) = 100 / (1 + es (n)) As is clear from this equation, the accumulated error cost ec
The reliability trst increases as (n) increases.
(N) decreases and approaches 0 (zero). On the other hand, as it becomes smaller, the reliability trst (n) increases, and its value approaches 100.

By performing such processing, n existing within a predetermined range from the current position A for a certain candidate point
The reliability trst (n) associated with each line segment is obtained. The reliability is calculated for each of the selected line segments in this way, but in the present specification, for the sake of convenience, it is also treated as the reliability for the candidates obtained on the line segment.

Then, a point advanced from a certain candidate point along the corresponding line segment by a length corresponding to the distance R traveled by the vehicle is set as a new candidate point C (n) (step 4).
26). Therefore, when the number of line segments in which a certain candidate point is located or a line segment connected to this is such that the difference between the azimuth and the vehicle azimuth is a predetermined value or less is n, N new candidate points C (n) will be generated. If there is a candidate point in the free state described later,
Since there is no line segment where the candidate point is located, and when the next candidate point for this candidate point is calculated (steps 422 to 426), all line segments within the predetermined range D are extracted from the corresponding virtual current position. The number of candidate points can be large.

Finally, from these candidate points, the candidate point having the highest reliability is selected as a display candidate, and this display data candidate point data is output (step 427). More specifically, the candidate point C having the highest reliability value is used as a display candidate point CD, that is, a candidate point for displaying on the display 17, its position, accumulated error cost, reliability, which will be described later. A state flag indicating the matching state or the free state is stored in a predetermined area of the RAM of the memory 25, and the positions of candidate points other than the display candidate points, accumulated error cost, reliability, The status flag and the like are also stored in a predetermined area of the RAM.

Here, the action of map matching will be considered by taking a concrete road as an example. For example, as shown in FIG. 7, for a certain candidate point 62 existing on the line segment 61,
It is assumed that the current position A is represented by the position indicated by the point 63.
In such a case, a line segment 64 connected from the current position A to the line segment 61 in which the candidate point 62 is located, and the difference between the azimuth and the vehicle azimuth is a predetermined value or less, 65
And the distances L (1) and L (2) from the current position A to the line segments 64 and 65 are calculated, and the calculated distances and the angles θ (1) and θ (2 of the line segments 64 and 65 are calculated. ) And the vehicle heading θcar, etc., the associated error cost, accumulated error cost, and reliability are calculated. Further, based on the traveling distance R of the vehicle obtained in step 307 of FIG.
From a certain candidate point 62, a position advanced by a length corresponding to the traveling distance R along line segments 61 and 64 or line segments 61 and 65 is calculated, and points corresponding to this position are respectively candidate. Mark points 66 and 67.

Further, as shown in FIG. 8, with respect to the candidate point 66 on the line segment 64, the new current position A is represented at the position indicated by the point 71, while the candidate on the line segment 65 is selected. Point 67
On the other hand, it is assumed that the new current position A ′ is represented by the position indicated by the point 72. In this case, from the current position A, line segments 73 and 74 which are connected to the line segment 64 and in which the difference between the azimuth and the vehicle azimuth is equal to or less than a predetermined value are taken out, and a new segment is added. From the current position A ′, a line segment 75 which is connected to the line segment 65 and whose difference between the direction and the vehicle direction is equal to or less than a predetermined value is extracted. Then, the distance L from the current position A to the line segments 73 and 74
1 (1) and L1 (2) are calculated, and the distance L2 (1) from the current position A ′ to the line segment 75 is calculated.
Furthermore, the distance calculated in relation to the current position A and the line segment 7
Based on the angles θ1 (1) and θ1 (2) of 3,74, the vehicle direction θcar, and the like, the related error cost, accumulated error cost, and reliability are calculated, as well as the current position A ′. Distance, angle of line segment 75
Based on 2 (1) and the vehicle direction θcar, the related error cost, accumulated error cost, and reliability are calculated.

Furthermore, based on the traveling distance R of the vehicle obtained in step 307 of FIG.
4 and 73, or along line segments 64 and 74,
Alternatively, a position advanced from the candidate point 67 along the line segments 65 and 75 by a length corresponding to the mileage R of the vehicle is calculated, and the points corresponding to this position are respectively set as new candidate points. . FIG. 9 shows the candidate points 81 to 83 newly obtained in this way.

Here, the matching state and the free state of the candidate points will be described. From the current position A for a candidate point located on a certain line segment, the line segment at which the candidate point is located or a line segment connected thereto, and the difference between the direction and the traveling direction of the vehicle is not more than a predetermined value. It is possible that there is no line segment such as. In this case, the current position A itself is treated as the next candidate point calculated from the candidate points. In this embodiment, the candidate points obtained in this way are called free state candidate points. On the contrary,
In other states, that is, from the current position A for a candidate point located on a certain line segment, to the line segment where the candidate point is located or to the line segment connected thereto, the direction and the traveling direction of the vehicle A state in which there is a line segment whose difference is less than or equal to a predetermined value, and as a result, the next candidate point can exist on a particular line segment is called a matching state.

Further, when there is no line segment where the candidate point is located or which is connected to this line segment and the difference between the direction and the traveling direction of the vehicle is less than a predetermined value, Error cost ec to be calculated in step 425
(N) is given a constant value that is larger than the value of the error cost obtained in the matching state.

When obtaining a new candidate point based on this free state candidate point, the virtual current position itself for the free state candidate point is set as a new free state candidate point, and the virtual current position is calculated from the virtual current position. If a line segment whose difference between the direction and the traveling direction of the vehicle is equal to or less than a predetermined value exists within the predetermined range D, the intersection point of the perpendicular line drawn from the virtual current position to the line segment and this line segment. Is also a candidate point.

As easily predicted from the above description,
The processing time for executing the map matching process depends on the density of the road corresponding to the road data, and particularly when the map corresponding to the read map data is an urban area, the road density is high. In some cases, this processing time becomes very large. Therefore, in the present embodiment, as described above with reference to FIG. 4, the processing load is reduced and the processing time is increased by deleting the candidate points having the reliability lower than the reliability of the display candidate points by a predetermined amount or more. Can be prevented. Since the candidate points are deleted based on the reliability, it is possible to reduce the processing load while maintaining the accuracy of map matching.

The display candidate points obtained in step 427 of FIG. 5 are displayed on the screen of the display 17 by the processing based on the flowchart shown in FIG.

This process is a routine of the microprocessor 24 which is activated and executed every one second.

First, it is determined whether or not the switch 14 is instructed to change the map scale by looking at the contents of the parallel I / O 21 (step 1301). If it is pressed (Yes in step 1301), a predetermined scale flag is set correspondingly (step 130).
2).

Next, data indicating the position and direction of the display candidate point is read from a predetermined area of the RAM of the memory 25 (step 1303), and a map of a scale according to the content of the scale flag switched in step 1302 is displayed. It is displayed on the display 17, for example, in the state as shown in FIG. 2 (step 1304).

Then, the position and the direction of the display candidate point are displayed in superimposition on the map by using the arrow symbol "↑" as shown in FIG. 2 described above (step 130).
5). Finally, the north mark indicating north and the distance mark corresponding to the reduced scale are displayed as shown in FIG. 2 so as to be superimposed on them (step 1306).

In this embodiment, the vehicle position and the direction are indicated by using the arrow symbols as described above, but the display form of the vehicle position and the direction indicates that the position and the traveling direction are
The form may be arbitrary as long as the display state is clearly shown. The same applies to the north mark and the like.

According to the present embodiment, the candidate points are deleted based on the reliability of each candidate point, so that the reliability of the map matching processing is maintained and the processing that requires a large processing time is reduced. Therefore, it is possible to reduce the processing time required to calculate the current position.

In the present specification, the term "means" does not necessarily mean physical means, but also includes the case where the function of each means is realized by software. Also,
The function of one means may be realized by two or more physical means, or the functions of two or more means may be realized by one physical means.

[0068]

As described above, according to the present invention, it is possible to provide a current position calculation device capable of reducing the processing load of the map matching without degrading the accuracy thereof.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a current position calculation device according to an embodiment of the present invention.

FIG. 2 is a diagram showing a display example of a map and a current position according to the present embodiment.

FIG. 3 is a flowchart showing a process of calculating a traveling direction and a traveling distance of a vehicle.

FIG. 4 is a flowchart showing a main process executed for each predetermined traveling distance according to the present embodiment.

5 is a flowchart showing detailed processing of one step (map matching processing) shown in FIG.

FIG. 6 is a diagram for explaining an example of road data according to the present embodiment.

FIG. 7 is a diagram for explaining a line segment corresponding to a road, a virtual current position, and a candidate point.

FIG. 8 is a diagram for explaining a line segment corresponding to a road, a virtual current position, and a candidate point.

FIG. 9 is a diagram for explaining a line segment corresponding to a road, a virtual current position, and a candidate point.

FIG. 10 is a flowchart showing a current position display process according to the present embodiment.

[Explanation of symbols]

 10 Current Position Calculation Device 11 Angular Velocity Sensor 12 Geomagnetic Sensor 13 Vehicle Speed Sensor 14 Switch 15 CD-ROM 16 CD-ROM Read Driver 17 Display 18 Controller

Claims (4)

[Claims] 1. A current position calculating device mounted on a vehicle for calculating a current position of the vehicle, comprising: an azimuth detecting means for detecting a traveling azimuth of the vehicle; and a distance calculating means for calculating a traveling distance of the vehicle. Road data storage means for storing road data, and based on the relative displacement obtained based on the traveling direction and the traveling distance, the current position of the vehicle is estimated as a virtual current position, and the virtual current position is stored in the road data storage. By comparing with the road data of the means, a candidate point of the current position on the road is calculated together with the reliability indicating its credibility, and a map matching means for recognizing the candidate point with the highest reliability as the current position is provided, The map matching means also estimates the next virtual current position and calculates the candidate points associated with the candidate points other than the candidate point having the highest reliability, and at that time, the calculated candidate points are calculated. Yoriyukido current position calculating device, wherein a delete candidate point lower than degree of advance. 2. A candidate point having a degree of reliability lower than a predetermined degree is a candidate point having a reliability lower than a reliability of the candidate point having the highest degree of reliability by a predetermined amount or more. Current position calculation device. 3. The map matching means calculates the reliability of a candidate point found on the road based on the distance to the road around the virtual current position and the heading difference between the road heading and the vehicle heading. The current position calculation device according to claim 1. 4. The map matching means, when a candidate point on the road corresponding to the virtual current position outside the road is not found, sets the virtual current position outside the road as a candidate point,
2. The predetermined degree used when determining whether or not to delete the candidate points outside the road is set so as to be easily deleted as compared with the candidate points on the road. 2. The current position calculation device described in 2 or 3.