JP7489002B2 - Autonomous Vehicles - Google Patents

Description

The present invention relates to autonomous vehicles.

For an autonomous vehicle, it is extremely important to know where the vehicle body is located on a map. To date, there are a wide variety of self-location detection methods, including the sensors used and algorithms. On the other hand, accurate self-location estimation is not always possible with only one algorithm. For this reason, there are methods for estimating self-location using multiple sensors, algorithms, etc. In the mobile body disclosed in Patent Document 1, a first estimated position and a first reliability corresponding to the first estimated position are obtained by estimating the position of the mobile body using information about the surroundings of the mobile body acquired by a first sensor. In addition, a second estimated position and a second reliability corresponding to the second estimated position are obtained by estimating the position of the mobile body using information about the surroundings of the mobile body acquired by a second sensor. Furthermore, the first and second estimated positions and the first and second reliability are used to obtain the current position of the mobile body.

JP 2020-17173 A

When using multiple algorithms in combination to estimate self-location, reliability, that is, an index showing which estimation result is most likely, is necessary, but the calculation method and value range of reliability differ for each algorithm. For example, in the reliability calculation method using the first algorithm, the reliability takes a value in the range of 0 to 100, and in the reliability calculation method using the second algorithm, the reliability takes a value in the range of 0 to 1. For this reason, it is difficult to integrate (merge) self-location estimation results using reliability as a weighting.

The autonomous vehicle for solving the above problem comprises a first data acquisition device installed on the vehicle body for estimating the vehicle's own position, a second data acquisition device installed on the vehicle body for estimating the vehicle's own position, a first self-position estimation unit installed on the vehicle body for estimating the vehicle's position coordinates and attitude using data acquired by the first data acquisition device and calculating the reliability of the estimation result, a second self-position estimation unit installed on the vehicle body for estimating the vehicle's position coordinates and attitude using data acquired by the second data acquisition device and calculating the reliability of the estimation result using a reliability calculation algorithm in the first self-position estimation unit, and an integrated self-position calculation unit installed on the vehicle body for calculating the integrated vehicle's position coordinates and attitude based on the estimation result of the vehicle's position coordinates and attitude by the first self-position estimation unit and the reliability of the estimation result of the vehicle's position coordinates and attitude by the second self-position estimation unit.

According to this, the first self-position estimation unit estimates the position coordinates and attitude of the vehicle body from the data acquired by the first data acquisition device, and calculates the reliability of the estimation result. Also, the second data acquisition device estimates the position coordinates and attitude of the vehicle body from the data acquired by the second data acquisition device, and calculates the reliability of the estimation result using a reliability calculation algorithm in the first self-position estimation unit. This makes it possible to calculate the relative reliability, and to calculate the integrated position coordinates and attitude of the vehicle body based on the reliability.

The autonomous vehicle may further include a storage unit installed in the vehicle body and storing map data associated with the position coordinates and attitude of the vehicle body, the first self-position estimation unit estimating the position coordinates and attitude of the vehicle body by comparing the data acquired by the first data acquisition device with the map data and calculating the reliability of the estimation result, and the second self-position estimation unit estimating the position coordinates and attitude of the vehicle body by comparing the data acquired by the second data acquisition device with the map data and calculating the reliability of the estimation result using a reliability calculation algorithm in the first self-position estimation unit.

In addition, in an autonomous vehicle, the reliability may be determined based on the sum of the deviations for each point in a comparison between the point cloud of map data and the point cloud obtained by a data acquisition device.

According to the present invention, it is possible to calculate the relative reliability and calculate the integrated vehicle body position coordinates and attitude based on the reliability.

FIG. 2 is a side view of an autonomous vehicle according to an embodiment. FIG. 1 is a block diagram showing the electrical configuration of an autonomous vehicle. 11 is a flowchart for explaining the operation. A diagram showing the environment and the location of the autonomous vehicle. FIG. 4A is a diagram showing an acquired point cloud, and FIG. 4B is a diagram showing a point cloud in map data. A diagram showing the captured environment and an autonomous vehicle. FIG. 13 is a diagram showing the state in which the acquired point cloud and the point cloud of map data are overlaid. FIG. 13 is a diagram showing the state in which the acquired point cloud and the point cloud of map data are overlaid.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
As shown in Fig. 1, the autonomous vehicle 10 is a four-wheel vehicle and includes a vehicle body 11, drive wheels 12 arranged at the bottom of the vehicle body 11, and steering wheels 13 arranged at the bottom of the vehicle body 11. The autonomous vehicle 10 travels by rotating the drive wheels 12 and is steered by changing the orientation of the steering wheels 13. Here, the direction of travel of the autonomous vehicle 10 is defined as the X direction, and the direction perpendicular to the direction of travel (X direction) (the width direction of the autonomous vehicle 10) is defined as the Y direction. The vertical direction perpendicular to the horizontal X-Y plane is defined as the Z direction.

As shown in FIG. 2, the autonomous vehicle 10 includes a control device 20, motor drivers 21 and 22, a traction motor 23, and a steering motor 24. The control device 20 is capable of controlling the traction motor 23 via the motor driver 21 to drive the drive wheels 12. The control device 20 is capable of controlling the steering motor 24 via the motor driver 22 to drive the steering wheels 13.

2, the control device 20 includes a processing unit 30 and a storage unit 40. The storage unit 40 stores various programs for controlling the autonomous vehicle 10.
The control device 20 may include dedicated hardware, such as an application specific integrated circuit (ASIC), for performing at least some of the various processes. The control device 20 may be configured as a circuit including one or more processors operating according to a computer program, one or more dedicated hardware circuits such as ASICs, or a combination thereof. The processor includes a CPU and memory such as RAM and ROM. The memory stores program code or instructions configured to cause the CPU to execute processes. The memory, i.e., computer-readable medium, includes anything that can be accessed by a general-purpose or dedicated computer.

The control device 20 operates the autonomous vehicle 10 by controlling the travel motor 23 and the steering motor 24 according to the program stored in the memory unit 40. The autonomous vehicle 10 of this embodiment is a vehicle that automatically travels and steers under the control of the control device 20 without being operated by a passenger.

As shown in FIG. 1, the autonomous vehicle 10 is equipped with a Lidar (Light Detection and Ranging or Laser Imaging Detection and Ranging) scanner 50, a road surface camera 60, and a GPS (Global Positioning System) antenna 70 as data acquisition devices for estimating its own position.

The Lidar scanner 50 is installed on the top surface of the vehicle body 11. The Lidar scanner 50 has an irradiation unit that irradiates the surroundings of the autonomous vehicle 10 with a laser, and a light receiving unit that receives the reflected light. The distance to surrounding objects is detected by measuring the time between the irradiation of the laser and the reception of the light. The laser is irradiated to scan a range of 270 degrees, for example, 135 degrees to the right and 135 degrees to the left, based on the direction of travel of the autonomous vehicle 10, and the distance to the object is measured in accordance with that angle.

The Lidar scanner 50 may be a two-dimensional type that detects objects in the horizontal direction, or a three-dimensional type that detects objects in the horizontal direction and the up-down direction.
2, the control device 20 is connected to a Lidar scanner 50. Measurement data obtained by the Lidar scanner 50 is input to the control device 20. This makes it possible to grasp the surrounding environment of the autonomous vehicle 10.

As shown in FIG. 1 , the road surface photographing camera 60 is installed in the center of the underside of the vehicle body 11 and is capable of photographing the road surface Sr below the vehicle body 11 .
As shown in FIG. 1, the autonomous vehicle 10 includes light sources 61 and 62 for illumination. The light sources 61 and 62 are installed on the underside of the vehicle body 11. The light sources 61 and 62 are, for example, light-emitting diodes (LEDs). The light sources 61 and 62 are for irradiating light onto an area photographed by a road surface photographing camera 60 on the road surface Sr. The light sources 61 and 62 are arranged on the underside of the vehicle body 11 so as to surround the periphery of the road surface photographing camera 60. The light sources 61 and 62 are turned on in synchronization with the photographing timing of the road surface photographing camera 60 to irradiate light onto the photographing area of the road surface photographing camera 60.

2 , the autonomous vehicle 10 includes a lighting driver 63. The control device 20 can control the light sources 61 and 62 via the lighting driver 63.
As shown in Fig. 1, the GPS antenna 70 is installed on the top surface of the vehicle body 11. The GPS antenna 70 can receive radio waves from artificial satellites. As shown in Fig. 2, the autonomous vehicle 10 is equipped with a GPS receiver 71. The GPS antenna 70 is connected to the control device 20 via the GPS receiver 71. The control device 20 can import measurement information from the GPS antenna 70.

As shown in FIG. 2 , the processing unit 30 of the control device 20 includes a Lidar self-position estimation unit 31 , a road surface imaging camera self-position estimation unit 32 , and a GPS self-position estimation unit 33 .
The storage unit 40 stores map data 41. The map data 41 includes surrounding environment data. The surrounding environment data is created in association with geographical location information. The map data 41 also includes a map image of the road surface Sr that has been photographed in advance. The map image is created in association with geographical location information.

In this way, the autonomous vehicle 10 is equipped with a memory unit 40 that is installed on the vehicle body 11 and stores map data 41 to which the position coordinates and attitude of the vehicle body 11 are associated.
The Lidar self-position estimation unit 31 can estimate the self-position of the autonomous vehicle by comparing the features (point cloud) extracted from the surrounding environment data measured by the Lidar scanner 50 with the features extracted from the map data 41 in the storage unit 40. The self-position is the position coordinates and attitude of the vehicle body 11 (for example, see the attitude angle θ in FIG. 6). The position coordinates of the vehicle body 11 are coordinates indicating one point on the vehicle body 11, for example, the coordinates at the installation position of the Lidar scanner 50 in the horizontal direction of the vehicle body 11. In detail, a feature point is detected from the surrounding environment (actual surrounding environment) at the present time, and a feature amount for the feature point, that is, a feature amount representing the degree of brightness around the feature point, is detected. Similarly, a feature point is detected from map data obtained in advance, and a feature amount for the feature point, that is, a feature amount representing the degree of brightness around the feature point, is detected. Then, the feature amount of each feature point in the surrounding environment (actual surrounding environment) at the present time is compared with the feature amount of each feature point in the map data obtained in advance, to estimate the self-position of the autonomous vehicle.

When map data (surrounding environment data) 41 is stored in advance in the storage unit 40, the coordinates of the surrounding environment are stored as an environmental map. The environmental map is created by mapping using SLAM (Simultaneous Localization and Mapping). SLAM simultaneously estimates the self-position of a moving body and creates an environmental map, and can create an environmental map in an unknown environment for the moving body. A specific task is performed using the constructed map information. When estimating the self-position, the self-position of the autonomous vehicle is estimated by comparing the surrounding environment acquired by the Lidar scanner 50 with map data acquired in advance.

The road surface photographing camera self-position estimation unit 32 can estimate the self-position of the autonomous vehicle by comparing the features (point cloud) extracted from the image of the road surface Sr photographed by the road surface photographing camera 60 with the features extracted from the map data (map image) 41 in the memory unit 40. The self-position is the position coordinates and attitude of the vehicle body 11. The position coordinates of the vehicle body 11 are coordinates indicating a point on the vehicle body 11, for example, the coordinates at the installation position of the road surface photographing camera 60 in the horizontal direction of the vehicle body 11. In detail, a feature point is detected from an image that is a road surface image (actual image) at the current time, and a feature amount for the feature point, that is, a feature amount representing the degree of brightness of the surrounding pixels relative to the pixel where the feature point is located, is detected. Similarly, a feature point is detected from a map image photographed in advance, and a feature amount for the feature point, that is, a feature amount representing the degree of brightness of the surrounding pixels relative to the pixel where the feature point is located, is detected. The system then estimates the autonomous vehicle's own position by comparing the feature values of each feature point in the current road surface image (actual image) with the feature values of each feature point in a map image captured in advance.

When map data (map image) 41 is stored in advance in storage unit 40, the coordinates of the road surface pattern are stored as an environmental map. The environmental map is created by mapping using SLAM (Simultaneous Localization and Mapping). SLAM simultaneously estimates the self-position of a moving body and creates an environmental map, and can create an environmental map in an unknown environment for the moving body. A specific task is performed using the constructed map information. When estimating the self-position, the self-position of the autonomous vehicle is estimated by comparing a road surface image acquired by a camera with a map image acquired in advance.

The GPS self-position estimation unit 33 can estimate the self-position of the autonomous vehicle from measurement information from the GPS antenna 70. The self-position is the position coordinates and attitude of the vehicle body 11. The position coordinates of the vehicle body 11 are coordinates indicating a point on the vehicle body 11, for example, the coordinates at the installation position of the GPS antenna 70 in the horizontal direction of the vehicle body 11.

The control device 20 can move the autonomous vehicle 10 to a desired position by controlling the driving motor 23 and the steering motor 24 while performing self-position estimation to estimate the position of the autonomous vehicle 10 on a map.

Next, the operation will be described.
FIG. 3 shows the process executed by the control device 20.
As a prerequisite, self-location estimation is performed using three algorithms. That is, self-location estimation is performed using algorithm A and algorithm B. Furthermore, self-location estimation is performed using algorithm C. Then, the position estimation result calculated using algorithm A is set as LA, and the position estimation result calculated using algorithm B is set as LB. Furthermore, the position estimation result calculated using algorithm C is set as LC.

Specifically, algorithm A is self-location estimation by Lidar using a Lidar scanner 50. Furthermore, algorithm B is self-location estimation by road surface images using a road surface imaging camera 60. Furthermore, algorithm C is self-location estimation by GPS positioning using a GPS antenna 70.

In step S10, the control device 20 performs self-position estimation for each algorithm. Specifically, the control device 20 calculates the position coordinates and attitude of the vehicle body 11 by referring to the surrounding environment data and map data from the Lidar scanner 50. The control device 20 also calculates the position coordinates and attitude of the vehicle body 11 by referring to the road surface image captured by the road surface imaging camera 60 and map data. Furthermore, the control device 20 calculates the position coordinates and attitude of the vehicle body 11 from data from artificial satellites captured by the GPS antenna 70.

In step S11, the control device 20 calculates the reliability of the position estimation result when it is LA based on algorithm A. Specifically, the control device 20 calculates the reliability of the position coordinates and attitude of the vehicle body 11 based on the surrounding environment data from the Lidar scanner 50 based on algorithm A.

In step S12, the control device 20 calculates the reliability of the position estimation result when it is LB based on algorithm A. Specifically, the control device 20 calculates the reliability of the position coordinates and attitude of the vehicle body 11 based on the image of the road surface captured by the road surface capturing camera 60 based on algorithm A.

In step S13, the control device 20 calculates the reliability of the position estimation result when it is determined to be LC based on algorithm A. Specifically, the control device 20 calculates the reliability of the position coordinates and attitude of the vehicle body 11 based on data from the satellites via the GPS antenna 70 based on algorithm A.

Through the above processes, the reliability of the position estimation result LA, the position estimation result LB, and the position estimation result LC from the viewpoint of algorithm A can be known.
In step S14, the control device 20 calculates and outputs a merged position estimation result based on the respective reliability. Specifically, for example, the result is a weighted sum, weighted according to the reliability, and the sum is used as the self-position.

An example of a reliability calculation method will be described below.
As a prerequisite, assume that the autonomous vehicle 10 is in an environment and at a position as shown in Fig. 4. In Fig. 4, the surrounding environment includes straight lines L1, L2, L3, L4, L5, L6, L7, and L8 formed by objects around the vehicle. In this case, a point cloud as shown in Fig. 5(a), i.e., points P1 to P36, can be acquired. Also assume that a map is created in advance in the environment of Fig. 4, and the point cloud of the map as shown in Fig. 5(b), i.e., points P101 to P136, is also acquired.

In other words, Figure 4 shows the environment in which the autonomous vehicle 10 is located, as well as its position coordinates and attitude. Figure 5(a) shows the point cloud (points P1 to P36) acquired by the autonomous vehicle 10. Figure 5(b) shows the point cloud (points P101 to P136) of the map held by the autonomous vehicle 10.

An example of calculating the reliability will be described.
As a prerequisite, algorithm A is self-location estimation using the Lidar scanner 50, and algorithm B is self-location estimation using the road surface imaging camera 60.

As a first step, it is assumed that the position coordinates and attitude of the autonomous vehicle 10 are estimated as shown in FIG. 6. FIG. 6 shows the estimated position coordinates and attitude of the autonomous vehicle 10, and in FIG. 6, the angle between the traveling direction (X direction) of the autonomous vehicle 10 and a specified orientation is the attitude angle θ. In FIG. 4, the attitude angle θ = 0.

In the second step, as shown in FIG. 7, the position coordinates and attitude of the autonomous vehicle 10 estimated in the first step are superimposed on the map data available.
In the third step, for the results of the overlapping in the second step, as shown in Fig. 8, for example, a point P135 on the map closest to point P3 is found, the distance ΔL1 is measured, and the score (point number) is calculated as Sc. Similarly, a point P136 on the map closest to point P4 is found, the distance ΔL2 is measured, and the score (point number) is calculated as Sc. Similarly, the score Sc for each point is calculated.

In this way, the reliability calculation method using algorithm A involves calculating the distance between each actual input point and the closest map data point when it is assumed that the Lidar scanner 50 is located at the estimated position.

In the fourth step, the score (points) Sc calculated in the third step is calculated for all points, and the sum is used as the reliability for the position coordinates and attitude of the autonomous vehicle 10. In other words, the distance between the actually input point and the point on the map data is calculated, and the sum of the distances is used as the reliability. This becomes the reliability (score Sc). Specifically, the position estimation result calculated by Lidar is used as LA, and the score Sc at this position is calculated as the reliability.

Similarly, the position estimation result calculated from the road surface image is designated as LB, and the score Sc at this position is calculated as the reliability. The score Sc for the position estimation result LB is calculated by calculating the distance between the input point where the Lidar point cloud data obtained at the same time (data estimated as the position estimation result LA in algorithm A) is aligned with the position estimation result LB, and the closest point in the map data.

Then, an integrated position is calculated based on the reliability of the position estimation result LA calculated by the Lidar and the reliability of the position estimation result LB calculated from the road surface image.
In this way, the Lidar self-position estimation unit 31 estimates the position coordinates and attitude of the vehicle body 11 from the data acquired by the Lidar scanner 50. More specifically, the Lidar self-position estimation unit 31 estimates the position coordinates and attitude of the vehicle body 11 by comparing the data acquired by the Lidar scanner 50 with the map data 41. Meanwhile, the road surface photographing camera self-position estimation unit 32 estimates the position coordinates and attitude of the vehicle body 11 from the data acquired by the road surface photographing camera 60. More specifically, the road surface photographing camera self-position estimation unit 32 estimates the position coordinates and attitude of the vehicle body 11 by comparing the data acquired by the road surface photographing camera 60 with the map data 41.

Furthermore, the Lidar self-position estimation unit 31 calculates the reliability of the estimation result when estimating the position coordinates and attitude of the vehicle body 11 from the data acquired by the Lidar scanner 50. Furthermore, the road surface photographing camera self-position estimation unit 32 calculates the reliability of the estimation result when estimating the position coordinates and attitude of the vehicle body 11 from the data acquired by the road surface photographing camera 60 using a reliability calculation algorithm in the Lidar self-position estimation unit 31.

Then, the control device 20 calculates and outputs the integrated position coordinates and attitude of the vehicle body 11 based on the estimation results and their reliability of the position coordinates and attitude of the vehicle body 11 by the Lidar self-position estimation unit 31 and the estimation results and their reliability of the position coordinates and attitude of the vehicle body 11 by the road surface imaging camera self-position estimation unit 32.

Let's talk about reliability.
The reliability calculation algorithm can be used depending on the installation position of the data acquisition device on the vehicle body 11. Since the data are viewed from the same position, the method of calculating the reliability differs depending on each self-position estimation software, but in this embodiment, the same reliability calculation algorithm is used for self-positions calculated by different methods.

As explained with reference to FIG. 8, the sum of the deviations between the point cloud at the estimated self-position and the point cloud of the map data is the deviation between the point cloud from the Lidar scanner 50 and the point cloud of the map data when the vehicle is estimated to be in this position coordinate and in this attitude. Therefore, the reliability can be calculated using the reliability calculation algorithm described above if the installation position of the road surface imaging camera 60 is the same as the installation position of the Lidar scanner 50.

Conventionally, when self-position estimation is performed using multiple algorithms in combination, the calculation method and value range of the reliability differ for each algorithm. For example, in the reliability calculation method using the first algorithm, the reliability takes a value in the range of 0 to 100, and in the reliability calculation method using the second algorithm, the reliability takes a value in the range of 0 to 1. For this reason, it was difficult to integrate (merge) the self-position estimation results weighted by the reliability. In this embodiment, the road surface camera self-position estimation unit 32 estimates the position coordinates and attitude of the vehicle body 11 using data acquired by the road surface camera 60, and calculates the reliability of the estimation result using a reliability calculation algorithm in the Lidar self-position estimation unit 31. Therefore, the reliability value range for each algorithm is the same, and the relative reliability can be calculated, and the integrated position coordinates and attitude of the vehicle body 11 can be calculated based on the reliability.

According to the above embodiment, the following effects can be obtained.
(1) The autonomous vehicle 10 is configured to include a Lidar scanner 50 installed on the vehicle body 11 as a first data acquisition device for estimating the vehicle's own position, and a road surface photographing camera 60 installed on the vehicle body 11 as a second data acquisition device for estimating the vehicle's own position. The autonomous vehicle 10 is configured to include a Lidar self-position estimation unit 31 installed on the vehicle body 11 as a first self-position estimation unit that estimates the position coordinates and attitude of the vehicle body 11 using data acquired by the Lidar scanner 50 and calculates the reliability of the estimation result. The autonomous vehicle 10 is configured to include a road surface photographing camera self-position estimation unit 32 installed on the vehicle body 11 as a second self-position estimation unit that estimates the position coordinates and attitude of the vehicle body 11 using data acquired by the road surface photographing camera 60 and calculates the reliability of the estimation result using a reliability calculation algorithm in the Lidar self-position estimation unit 31. The control device 20 is installed on the vehicle body 11 and serves as an integrated self-position calculation unit that calculates the integrated position coordinates and attitude of the vehicle body 11 based on the estimation results and reliability of the position coordinates and attitude of the vehicle body 11 by the Lidar self-position estimation unit 31 and the estimation results and reliability of the position coordinates and attitude of the vehicle body 11 by the road surface shooting camera self-position estimation unit 32.

Therefore, the Lidar self-position estimation unit 31 as the first self-position estimation unit estimates the position coordinates and attitude of the vehicle body 11 from the data acquired by the Lidar scanner 50, and calculates the reliability of the estimation result. In addition, the road surface photographing camera self-position estimation unit 32 as the second data acquisition device estimates the position coordinates and attitude of the vehicle body 11 from the data acquired by the road surface photographing camera 60, and calculates the reliability of the estimation result using a reliability calculation algorithm in the Lidar self-position estimation unit 31 as the first self-position estimation unit. This makes it possible to calculate the relative reliability, and to calculate the integrated position coordinates and attitude of the vehicle body 11 based on the reliability.

(2) The vehicle further includes a storage unit 40 that is installed on the vehicle body 11 and stores map data 41 to which the position coordinates and attitude of the vehicle body 11 are associated. The Lidar self-position estimation unit 31 as a first self-position estimation unit estimates the position coordinates and attitude of the vehicle body 11 by comparing the data acquired by the Lidar scanner 50 as a first data acquisition device with the map data 41, and calculates the reliability of the estimation result. The road surface photographing camera self-position estimation unit 32 as a second self-position estimation unit estimates the position coordinates and attitude of the vehicle body 11 by comparing the data acquired by the road surface photographing camera 60 as a second data acquisition device with the map data 41. In addition, the road surface photographing camera self-position estimation unit 32 as the second self-position estimation unit calculates the reliability of the estimation result using a reliability calculation algorithm in the Lidar self-position estimation unit 31 as the first self-position estimation unit. Therefore, it is practical.

(3) The reliability is based on the sum of the deviations for each point in the comparison between the point cloud of the map data 41 and the point cloud obtained by the data acquisition device (Lidar scanner 50, road surface photography camera 60), and is therefore practical.

The embodiment is not limited to the above, and may be embodied as follows, for example.
In the above embodiment, three algorithms, algorithms A, B, and C, are used, but two algorithms may be used. Also, four or more algorithms may be used. In short, two or more algorithms can be used.

For example, in the above embodiment, the autonomous vehicle 10 was equipped with a Lidar scanner 50, a road surface imaging camera 60, and a GPS antenna 70. Instead of the road surface imaging camera 60, an optical element that detects a reflector laid on the floor surface may be provided, and the autonomous vehicle 10 may travel along the reflector.

Any merging method (integration method) based on reliability may be used. For example, a Kalman filter or the like may be used in addition to the weighted sum.
The reliability when the position estimation result is LA is calculated based on algorithm A, the reliability when the position estimation result is LB is calculated based on algorithm A, and the reliability when the position estimation result is LC is calculated based on algorithm A. After that, the self-position estimation results calculated by each algorithm may be verified again (cross-checked) by other algorithms.

10...Autonomous vehicle, 11...Vehicle body, 20...Control device, 31...Self-position estimation unit for Lidar, 32...Self-position estimation unit for road surface imaging camera, 33...Self-position estimation unit for GPS, 50...Lidar scanner, 60...Road surface imaging camera, 70...GPS antenna.

Claims (1)

A first data acquisition device that is installed on a vehicle body and that estimates a vehicle's own position;
A second data acquisition device is installed on the vehicle body and performs self-location estimation;
a first self-position estimation unit that is installed on the vehicle body and estimates a position coordinate and an attitude of the vehicle body based on the data acquired by the first data acquisition device, and calculates a reliability of the result of the estimation;
a second self-position estimation unit that is installed on the vehicle body and estimates a position coordinate and an attitude of the vehicle body based on the data acquired by the second data acquisition device, and calculates a reliability of the estimation result using a reliability calculation algorithm in the first self-position estimation unit;
an integrated self-position calculation unit that is installed on the vehicle body and calculates integrated position coordinates and attitude of the vehicle body based on the estimation results of the position coordinates and attitude of the vehicle body by the first self-position estimation unit and their reliability and the estimation results of the position coordinates and attitude of the vehicle body by the second self-position estimation unit and their reliability ,
A storage unit is provided on the vehicle body and stores map data in which the position coordinates and the attitude of the vehicle body are associated with each other.
the first self-position estimation unit estimates a position coordinate and an attitude of the vehicle body by comparing the data acquired by the first data acquisition device with the map data, and calculates a reliability of the result of the estimation;
The second self-position estimation unit estimates the position coordinates and attitude of the vehicle body by comparing the data acquired by the second data acquisition device with the map data, and calculates the reliability of the estimation result using a reliability calculation algorithm in the first self-position estimation unit, and the reliability is based on the sum of the amount of deviation for each point in comparing the point cloud of the map data with the point cloud acquired by the data acquisition device .