JPH0763570A - Map matching mthod in navigation device - Google Patents

Abstract

PURPOSE:To estimate the present position by the evaluation of the matching of a traveling orbit obtained from the position measurement data of a GPS receiver with a map data in a navigation device for vehicle. CONSTITUTION:A system is formed of a GPS receiver and a CD-ROM for holding a map data. The data from the GPS receiver is sampled and held for a fixed period, whereby a traveling orbit is provided. The area in which the amount of maximum position measurement error is added to the minimum area including the traveling orbit is determined as the present position existing area. The conformity with the quantized road map is evaluated while moving the traveling orbit in parallel within this area. The position having a high evaluation value is set as the position of a candidate road having the possibility of the present position. Whether the candidate road is conformed with the actual road is evaluated to determine the final position. When the candidate road can not be finally narrowed to one, the minimum area including all candidate roads is displayed. When no candidate road is present, only the present position existing area is shown to an user, whereby reliability is kept.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention estimates a current position in a navigation system by obtaining positioning data from a GPS receiver, matching a running locus obtained from this positioning data with a quantized road map, and evaluating the result. It relates to a map matching method.

[0002]

2. Description of the Related Art In recent years, many navigation devices have been equipped with a plurality of sensors to calculate the current position based on the output data from each sensor and to perform map matching with road data to improve accuracy. .

As a conventional map matching method,
Some are found in Japanese Patent Laid-Open No. 3-26913. The conventional method will be described below.

In a navigation device equipped with an azimuth sensor, a distance sensor, and a GPS receiver, movement components (Δxi, Δyi) in the x and y axis directions are obtained from the azimuth sensor and distance sensor for each unit traveling, and trajectory data is obtained. Is saved as. When a certain distance or more is advanced, it is saved as matching trajectory data, and N pieces of data are gathered to enter the map matching process. When the matching process is completed, the oldest items are discarded.

In addition, from the GPS receiver, the positioning position center (C
x, Cy) is calculated, and this position is set as the search center point. The DOP value RD which is the index value of the error is read from the GPS receiver, and the inside of the circle of radius α × RD centering on the search center point is set as the search area S.

Next, the matching process will be described.
The roads in the search area S are divided for each unit distance, and these divided roads are set as the candidate points of the current vehicle position. After that, the candidate points are narrowed down by the sequential relaxation method, and the strong candidate points are extracted. If there is no strong candidate point, no matching process is performed. Then, the pattern matching process is performed on the assumption that the current vehicle position is present at this strong candidate point. The entire trajectory shape is rotated within a certain range for each unit angle around a strong candidate point. The residual sum of squares of the points corresponding to the locus and the road is calculated for each unit angle rotation, and the minimum point is set as the error value at the influential candidate point. This process is performed for all the promising candidate points. If there is a point where the error value is less than or equal to the threshold, the next processing is performed. The standard deviation σ and the correlation coefficient γ are obtained for the coordinate sequence of the influential candidate points, and then the most suitable candidate point for position correction is obtained. If σ is smaller than a constant value σc, the position is corrected to a strong candidate point that minimizes the residual sum of squares. If σc or more and γ is smaller than a constant value γc,
The position is corrected to an influential candidate point where the difference between the distance between the previously corrected position and the current position and the distance between the previously corrected position and the influential candidate point is the smallest. When σ is σc or less and γ is γc or less, position correction is not performed.

[0007]

However, according to the above-mentioned conventional method, when the positioning data from the GPS receiver is used in a system which is not equipped with a distance sensor and an orientation sensor, the GPS receiver is moved every time a certain distance is traveled. The positioning data is not always obtained from. In addition, since the positioning error of GPS is large, it is difficult for the conventional map matching method to correlate with road data, and the current position cannot be determined in many cases.

The present invention solves the above-mentioned problems of the prior art, and an object of the present invention is to provide a method of map matching between data from a GPS receiver and map data.

[0009]

In order to achieve this object, the present invention constitutes a system using a GPS receiver and map data, and evaluates matching between a traveling locus and map data.

First, the map data is quantized in advance in order to reduce the influence of GPS error. Next, data from the GPS receiver is sampled to obtain a traveling locus.
A function that evaluates the matching between the sampling data and the quantized road map is defined, and the position with a high evaluation value obtained from this function is set as the position of the candidate road where the current position may exist. Next, in order to evaluate whether or not this candidate road matches the actual road, the change of the evaluation value at the position where the travel locus is slightly translated in the direction that does not overlap this position at the position of the candidate road Compare sizes. This determines the final position. When there is no candidate road, reliability is maintained by showing the user only the area that may exist.

Further, a velocity is set so as not to exceed the maximum value of the drift velocity of the position in the fixed point measurement of the GPS receiver, and the distance from the fixed point when moving in one direction at this velocity is the position in the drawing of the map. By calculating a period that does not exceed approximately four times the quantized quantization unit and storing only the information from the GPS receiver sampled within this period, the shape of the traveling locus is maintained and Make it easier to correlate.

[0012]

According to the present invention, the running locus obtained from the GPS receiver is matched with the quantized road map, and the above-mentioned means is used for the evaluation method, so that the current position can be calculated without any other sensor data. You can specify.

Further, by limiting the sampling period for obtaining the running locus, the running locus and the quantized road map can be easily correlated, and the current position can be determined more frequently.

[0014]

【Example】

(Embodiment 1) Hereinafter, Embodiment 1 of the present invention will be described with reference to the drawings. 1 is a diagram showing a configuration of a navigation device according to a first embodiment of the present invention. In FIG. 1, reference numeral 1 is a GPS receiver that measures a current position by capturing signals from GPS satellites. Reference numeral 2 is a CD-ROM that stores map data, 3 is a control device that performs map matching processing, and 4 is a display device that displays the results obtained by the control device 3. Although the CD-ROM is used in the first embodiment, the storage device is not limited to the CD-ROM and may be any storage device that holds the map information.

The operation of the above arrangement will be described. FIG. 2 shows a flow chart of the main program for map matching in the first embodiment. Hereinafter, description will be given according to this flowchart. At step 1, map data including a necessary area is read from the CD-ROM 2. In step 2, the read map data is quantized. In the first embodiment, the quantization method is GP.
A value that does not exceed the maximum value of the positioning error obtained from the S receiver 1 is determined, and the size of about 1/3 of this value is used as the quantization unit. Set up a map to determine the position with this as a unit,
The road read in step 1 is drawn in binary as to whether or not it exists in the unit area of this map. However, the quantization method is not limited to this. Also, the quantization unit when quantizing is not limited to this,
The size should be large enough to distinguish the shapes of roads. next,
In step 3, the information from the GPS receiver 1 is sampled.

A detailed description will be given with reference to FIG. Step 8
Then, the sampling number Ns of data from the GPS receiver 1
Is initialized. What is this sampling number? Step 4
It is the number of data input from the GPS receiver 1 before the subsequent matching process is started. Next, in step 9, GPS
Input data from the receiver and save. The input data (hereinafter, travel data) is the positioning data, error,
It is a velocity vector. Sampling number N in step 10
It is determined whether s is 10 or more. If 10 or more, the process proceeds to step 11. If not, the process returns to step 9 to input the traveling data. In step 11, the sampled data is saved as matching data. The matching traveling data includes the data sampled up to the previous time point so that the shape as the locus can be obtained. In the first embodiment, the running locus of about 60 points is obtained, but the number of points is not limited to this. In step 12, the matching process starts. When the number of matching data (hereinafter referred to as sampling data) exceeds 60, the process proceeds to step 13 and 10 data are discarded in the oldest order. However, the number of data is not limited to this value, and it is sufficient if the number of data is enough to obtain a travel locus. The above steps 8 to 13 are repeated.

In step 4, the minimum area including the travel locus obtained in step 3 is obtained, and the rectangular area is determined by adding the maximum size of the positioning error of the sampling data to this area. It is assumed that the current position exists in this area (hereinafter referred to as the current position existing area). The degree of matching between the shape of the road and the shape of the running locus in the present position existence area is evaluated. This is the same as the conventional technique in that the search area is specified and the current position is searched in the search area.
In step 5, the sampling data is weighted by the error information. The weight is increased as the sampling data has a smaller error. In the first embodiment, the weighting is performed based on the calculation formula shown in the equation 1. However, the method is not limited to this. In step 6, the shape of the traveling locus and the quantized road map are matched and evaluated. In order to find the position that most matches the running locus, the running locus is translated in the current position existence area. The amount of parallel translation at one time is one half of the quantization unit. In step 7, the matching result is evaluated. The evaluation method will be described with reference to FIG. First, in step 14, a threshold Th that serves as a reference for evaluation is set in advance. In the first embodiment, the value of the threshold Th is set to about 80% of the expected evaluation value when all the sampling data match. This value depends on the system design. In the first embodiment, the evaluation value is a value obtained by adding the weights of the sampling data that match the quantized road map. However, it is not limited to this. Next, in step 15, the sum of the weights of the sampling data that matches the road is calculated. This value is the evaluation value E1. In step 16, it is determined whether the evaluation value E1 is greater than or equal to the threshold Th. If E1 ≧ Th, the process proceeds to step 17 and the position of the traveling locus at that point is set as the position of the candidate road. If not, the process moves to the next step 18 as it is, and it is determined whether or not the search for all the roads in the current position existing area is completed. If not completed, the traveling locus is translated in step 19 and the processing from step 15 is repeated at that position.

As described above, from step 14 to step 1
If one or more candidate roads are determined as a result of the process of 8, the final road is determined from the candidate roads. Since it is the evaluation of the traveling locus and the quantized road map, the evaluation value may be high at a position that is not actually a road, and an incorrect position may be a candidate road. A method for identifying such a candidate road and finally determining the correct position will be described with reference to FIG. First, step 19
Then, set the threshold Th2 for evaluation. This value will be described later. In step 20, the direction perpendicular to the shape of the traveling locus is obtained from the velocity vector of each sampling data. However, although the velocity vector exists at each sampling point, in consideration of the velocity error, the velocity is not used unless the velocity is approximately four times or more of the velocity error in the first embodiment. This value depends on the system design. In the case of the traveling locus as shown in FIG. 6, the directions of a and b can be obtained.
Moving to step 21, the traveling locus is slightly translated from the position of the candidate road in the directions of a and b. In the first embodiment, the amount of translation is the quantization unit.

In step 22, the total sum of the sampling data matched with the road map quantized at the translated position is obtained. This value is used as the evaluation value E2. Moving to step 23, the evaluation value E2 is the threshold T determined in step 19
Determine if it is less than or equal to h2. If the sampling points that match the road at the previous position do not match as a result of the parallel movement, the candidate road exists on the actual road. In the first embodiment, the threshold Th2 is set to a value that does not exceed the sum of the weights of the sampling points that do not coincide with the road at the position of the candidate road. If E2.ltoreq.Th2, proceed to step 25 as it is, otherwise move to step 24 and judge that the candidate road does not exist on the actual road and remove it. In step 25, it is determined whether the evaluation has been completed for all the candidate roads. If it has not been completed, the processing from step 20 to step 25 is performed on the next candidate road. In the first embodiment, the directions in which the parallel movement is performed are as described above, but a plurality of directions are determined in advance, the parallel movements are performed in these directions, and it is determined whether the evaluation value decreases in any direction. But it's okay.

The position of the final candidate road obtained as described above is determined as the road where the current position exists. If there are multiple candidate roads at this point, the smallest area including all candidate roads is presented. If the candidate road does not exist, the current position existing area is presented.

As described above, according to the first embodiment, since the positioning error of GPS is large, the influence of the error is reduced by using the means for matching the quantized road map with the running locus. In addition, the candidate road obtained by simply evaluating the shape is translated by a quantization unit in the direction in which there is less overlap with the previous position, and whether the road is actually on the road by comparing the evaluation values at that position. Can be determined. By performing such processing, it is possible to prevent the evaluation value of the matching between the traveling locus obtained from the GPS and the quantized road map from being erroneously recognized as the current position even if the evaluation value becomes high at an erroneous position other than the actual road. You can

(Second Embodiment) A second embodiment of the present invention will be described below with reference to the drawings. The operation of the same configuration as that of the first embodiment will be described. 2 as in the first embodiment.
Match as shown in. The difference from the first embodiment is the method of sampling data from the GPS receiver. The drift velocity of the position in the fixed point measurement of the GPS receiver 1 is obtained in advance, and the period Td in which the distance when traveling in one direction at a constant velocity that does not exceed this velocity is approximately four times the quantization unit is calculated. A period equal to or longer than the period in which the traveling locus is obtained and within Td is determined and is set as the valid period of the matching data. In the second embodiment, it is set to about 60 seconds, but it is not limited to this value.

FIG. 7 shows a flowchart of the sampling process in the second embodiment. Hereinafter, description will be given with reference to FIG. 7. In step 26, the sampling time ts is initialized. This sampling time is the time for which data is input from the GPS receiver before the matching processing in step 4 and subsequent steps is started. Move to step 27
Input data from GPS receiver and save. Step two
At 8, it is determined whether the sampling time ts has passed 10 seconds. If ts ≧ 10, the process moves to step 29 and the sampled data is stored as matching data. If not, the process returns to step 27 and the data of the GPS receiver is input. Next, in step 30, the process returns to the main routine of the matching process, and the process of step 4 starts.
When the processing up to step 7 is completed, then step 31
Then, the data for matching which has passed approximately 60 seconds after the positioning is discarded.

As described above, when the traveling locus is obtained from the positioning data input from the GPS receiver, a period in which the positional relationship is lost at the beginning and the end of the traveling locus due to the drift of the GPS position has passed. By not using, the shape of the traveling locus can be maintained and the correlation with the map data can be easily obtained.

[0025]

As described above, the present invention evaluates the matching between the traveling locus and the quantized road map, and re-evaluates whether the position having a high evaluation value is on the actual road, Reliability is improved without determining an incorrect position as the current position.

Further, the sampling method for obtaining the traveling locus from the information from the GPS receiver does not use the data in which the period where the beginning and the end of the traveling locus become unrelated due to the drift of the GPS position has elapsed. Correlation with map data can be facilitated.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a navigation device according to first and second embodiments of the present invention.

FIG. 2 is a flowchart of a matching main routine according to the first and second embodiments of the present invention.

FIG. 3 is a flowchart for explaining an operation of sampling from the GPS receiver according to the first embodiment of the present invention.

FIG. 4 is a flowchart showing an evaluation method for determining a candidate road in the first and second embodiments of the present invention.

FIG. 5 is a flowchart for evaluating candidate roads according to the first and second embodiments of the present invention.

FIG. 6 is a diagram showing directions of parallel movement for evaluating candidate roads in the first and second embodiments of the present invention.

FIG. 7 is a flowchart for explaining a sampling operation from a GPS receiver according to the second embodiment of the present invention.

[Explanation of symbols]

 1 GPS receiver 2 CD-ROM 3 control device 4 display device

Claims (2)

[Claims] 1. A navigation device having a GPS receiver as a position sensor, a map information storage device for storing map information, a control unit for performing map matching processing, and a display unit for displaying the current position, wherein the GPS Positioning data is obtained from the receiver, a traveling locus is obtained from this positioning data, and a size that does not exceed the size of the area in which the maximum positioning error of the GPS receiver is taken into consideration is a quantization unit,
Providing a map that determines the position with this as a unit, and regarding the road read by the map information storage device on this map,
Whether or not a road exists in this unit area is drawn by a numerical value, a function for evaluating the matching between the drawn road map and the shape of the running path is defined, and the map data and the running path are moved relative to each other. Evaluate the road map and the running locus with the function at the position, select the position with the obtained high evaluation value, and at the selected position, the running locus is parallel to the direction that does not overlap the previous position by the quantization unit. A map matching method in a navigation device, in which evaluation values at moved positions are compared and the current position is determined based on the magnitude of change in the evaluation values. 2. In the fixed point measurement of the GPS receiver, a speed is set so as not to exceed the maximum value of the moving speed obtained when the drift velocity of the position is obtained, and from this fixed point when the vehicle moves in one direction. 2. The map matching method in the navigation device according to claim 1, wherein only the data sampled within a period in which the distance of .gtoreq. .