JP7356208B2 - Self-position estimation device, self-position estimation method, and self-position estimation program - Google Patents

Description

The present invention relates to a self-position estimating device for estimating the self-position of a moving body.

In the autonomous driving of moving objects such as vehicles, which is currently under development, the situation around the own vehicle is acquired using various sensors such as lidar (Light Detection And Ranging), and based on the acquired surrounding situation. and estimates its own position (self-position). Specifically, self-position estimation is often performed using predetermined features (particularly road surface information) as the surrounding situation (for example, see Patent Document 1).

JP2017-78607A

When estimating one's own position using the road surface information described above, the information that can be acquired by a sensor such as a rider may change depending on weather conditions such as rain and snow. For example, if water accumulates on a road marking or the like drawn on the road surface due to rain, the road marking may not be detected by a sensor such as a lidar.

As described above, it is desired that moving objects such as self-driving vehicles be able to travel according to the weather.

An example of the problem to be solved by the present invention is to enable a moving object such as an automatic driving vehicle to travel according to the weather.

In order to solve the above problem, the invention according to claim 1 includes a first acquisition unit that acquires an external situation signal indicating the external situation of a mobile object, and a map including information indicating characteristics affected by weather. A second acquisition unit that acquires data, a third acquisition unit that acquires weather information indicating weather conditions, and a self-position estimating unit that estimates the self-position of the mobile object based on the external situation signal. , the self-position estimating unit evaluates the accuracy of the external situation signal based on the map data and the weather information.

The invention according to claim 6 is a self-position estimation method executed by a self-position estimating device for estimating the self-position of a mobile body, the method comprising: a first acquisition method for acquiring an external situation signal indicating an external situation of the mobile body; a second acquisition step of acquiring map data including information indicating features affected by weather; a third acquisition step of acquiring weather information indicating weather conditions; and a self-position estimating step of estimating the self-position of the mobile object, and the self-position estimating step is characterized in that the accuracy of the external situation signal is evaluated based on the map data and the weather information.

The invention according to claim 7 is characterized in that the self-position estimation method according to claim 6 is executed by a computer.

The invention according to claim 8 is a map data structure of map data used for self-position estimation, and the map data includes information indicating characteristics related to weather.

The invention according to claim 10 is characterized in that the map data structure according to claim 8 or 9 is stored.

1 is a schematic configuration diagram of a system having a map data storage device according to an embodiment of the present invention. 2 is a functional configuration diagram of the server device shown in FIG. 1. FIG. 2 is a functional configuration diagram of the vehicle control device shown in FIG. 1. FIG. It is a diagram showing an image of weather characteristic information. 4 is a diagram showing a portion related to weather characteristic information of the map data structure of the map data shown in FIG. 3. FIG. 4 is a flowchart of a self-position estimation method in the control unit shown in FIG. 3. FIG. 7 is a flowchart according to a modification of the self-position estimation method shown in FIG. 6. 8 is a flowchart according to a modification of the self-position estimation method shown in FIG. 7.

A self-position estimating device according to an embodiment of the present invention will be described below. A self-position estimating device according to an embodiment of the present invention includes a first acquisition unit that acquires an external situation signal indicating an external situation of a mobile object, and acquires map data including information indicating characteristics affected by weather. a second acquisition unit that acquires weather information indicating weather conditions; and a self-position estimating unit that estimates the self-position of the mobile object based on an external situation signal. evaluates the accuracy of the external situational signal based on map data and weather information. By doing so, the vehicle can travel while estimating its own position according to the weather indicated by the weather information acquired by the second acquisition unit, using the information included in the map data and indicating the characteristics affected by the weather.

Further, if the evaluation of the external situation signal is lower than a predetermined standard, the self-position estimation unit may not use the signal for self-position estimation. By doing this, it is possible to prevent erroneous self-position estimation by determining that the reliability of self-position estimation has decreased due to the influence of weather and not using external situation signals for self-position estimation. can.

Further, the map data may be set with information indicating characteristics that are affected by weather for each section subdivided into predetermined sizes. By doing this, for example, features that are affected by the weather are set in a subdivided manner, making it possible to create map data that more accurately reflects the features of the region indicated by the section.

Further, the information indicating the characteristics affected by the weather may include at least one of the following information: water tends to accumulate due to rainfall, snow tends to accumulate due to snowfall, and freezing tends to occur. By doing so, it is possible to estimate the self-position in response to cases where the road surface condition is likely to change due to rain, snowfall, or freezing.

The second acquisition unit may also include a route search unit that searches for a route to a predetermined destination based on the map data acquired by the second acquisition unit, and the route search unit may further search for a route based on weather information. . By doing so, it becomes possible to search for a route according to the weather.

Further, the self-position estimation method according to an embodiment of the present invention includes a first acquisition step of acquiring an external situation signal indicating the external situation of the mobile object, and map data including information indicating characteristics affected by weather. a second acquisition step of acquiring weather information; a third acquisition step of acquiring weather information; and a self-position estimation step of estimating the self-position of the mobile object based on an external situation signal, the self-position estimation step comprising: Evaluate the accuracy of external situational signals based on map data and weather information. By doing so, the vehicle can travel while estimating its own position according to the weather indicated by the weather information acquired by the second acquisition unit, using the information included in the map data and indicating the characteristics affected by the weather.

Further, the self-position estimation method described above may be executed by a computer. By doing so, it is possible to use a computer to estimate the self-position according to the weather indicated by the weather information acquired by the third acquisition unit, based on the information indicating the characteristics affected by the weather included in the map data. can.

Further, the map data structure according to an embodiment of the present invention is a map data structure of map data used for self-position estimation, and the map data includes information indicating characteristics affected by weather. By doing this, it is possible to estimate the self-position according to the weather using information indicating the characteristics affected by the weather.

The map data may also include information indicating the characteristics affected by the weather and information as a result of evaluating a signal indicating the external situation of a predetermined mobile object based on the weather information. By doing this, when it is indicated that a feature cannot be detected by a sensor such as a lidar mounted on a moving body such as a vehicle, the result of evaluating whether the situation was affected by the weather or not. can be included. Therefore, it is possible to accumulate a history of weather and evaluation results, and by referring to this history, it is possible to effectively utilize information indicating characteristics affected by the weather.

Further, a map data storage device according to an embodiment of the present invention stores the above-described map data structure. By doing so, it is possible to perform self-position estimation according to the weather using information indicating weather-related characteristics.

A self-position estimating device, map data structure, and map data storage device according to an embodiment of the present invention will be described with reference to FIGS. 1 to 8. As shown in FIG. 1, the vehicle control device 3 as a self-position estimating device installed in the automatic driving vehicle C is capable of communicating with the server device 1 and the server device 2 via a network N such as the Internet. There is.

FIG. 2 shows the functional configuration of the server device 1 as a map data storage device. The server device 1 includes a control section 11, a communication section 12, and a storage section 13.

The control unit 11 functions as a CPU (Central Processing Unit) of the server device 1 and manages the overall control of the server device 1 . In response to a request from the vehicle control device 3 , the control section 11 reads map data of a necessary area from the map data 13 a stored in the storage section 13 and distributes it to the vehicle control device 3 via the communication section 12 .

The communication unit 12 functions as a network interface of the server device 1, and receives request information etc. output from the vehicle control device 3. Further, the control unit 11 transmits map data read from the map data 13a to the vehicle control device 3.

The storage unit 13 functions as a storage device such as a hard disk of the server device 1, and stores map data 13a. The map data 13a is a map that includes detailed information (lane network, location of features, etc.) that allows the automatic driving vehicle C to travel autonomously. The map data 13a also includes weather feature information 13b (see FIG. 5) as information indicating features affected by weather. The weather characteristic information 13b will be described later.

The automatic driving vehicle C includes a vehicle control device 3 and a sensor 4. The vehicle control device 3 causes the automatic driving vehicle C to travel autonomously (automatically drive) based on the result detected by the sensor 4 and the map data 33a that the vehicle control device 3 has.

The functional configuration of the server device 2 is also similar to that of the server device 1. In the server device 2, weather data 2a, which is information indicating weather conditions in various places, is stored in a storage unit 13. The weather data 2a includes not only weather conditions but also information such as precipitation, snowfall, and temperature.

FIG. 3 shows the functional configuration of the vehicle control device 3. The vehicle control device 3 includes a control section 31, a communication section 32, and a storage section 33.

The control unit 31 estimates the self-position of the self-driving vehicle C based on the results detected by the sensor 4, the weather data acquired from the server device 2, and the map data 33a stored in the storage unit 33. Then, the control unit 31 controls the steering wheel (steering device), accelerator, brake, etc. of the automatic driving vehicle C to cause the automatic driving vehicle C to travel autonomously. That is, the control unit 31 acquires the detection result (recognition result) of the sensor 4 as an external world recognition unit. In addition, the control unit 31 requests the server device 1 to distribute map data of an area that is a driving route via the communication unit 32, and stores the map data 33a distributed from the server device 1 in the storage unit 33. The route to the destination is searched.

The communication unit 32 transmits the request information etc. output by the control unit 31 to the server device 1. It also receives map data distributed from the server device 1 and weather data distributed from the server device 2.

The storage unit 33 serving as a map data storage device stores map data 33a. The map data 33a is a map that includes detailed information that allows the automatic driving vehicle C to travel autonomously. Furthermore, the map data 33a includes weather characteristic information 13b, similar to the map data 13a.

The sensor 4 is a sensor for recognizing information about the own vehicle such as the own vehicle position and the surrounding environment (features existing in the surrounding area, etc.), and includes a camera, a lidar, a GPS (Global Positioning System) receiver, etc. . In addition to these sensors, there is also an acceleration sensor that detects the acceleration of the vehicle, a speed sensor that detects the speed of the vehicle, or an inertia sensor that recognizes the posture (orientation, etc.) of the vehicle and corrects the data acquired by other sensors. A measurement device (IMU: Inertial Measurement Unit), a gyro sensor, etc. may be included.

A camera included in the sensor 4 captures an image representing the external environment of the automatic driving vehicle C. The lidar included in the sensor 4 discretely measures the distance to an object existing in the outside world, and recognizes the position, shape, etc. of the object as a three-dimensional point group. The information acquired by this lidar is output as point cloud information. A GPS receiver included in the sensor 4 generates and outputs latitude and longitude position information representing the current position of the vehicle.

Next, a map data structure having the weather characteristic information 13b according to this embodiment will be explained. As shown in FIG. 4, the weather characteristic information 13b included in the map data according to the present embodiment is divided into meshes (sections) M of a predetermined size on the map, and each mesh M is divided into areas where water accumulates due to rainfall. This shows how easy it is. FIG. 4 shows that as the hatching density increases, water tends to accumulate more easily. In addition, in FIG. 4, meshes M also exist in areas without hatching, but are omitted to make the drawing easier to see (that is, areas without hatching indicate areas where water is unlikely to accumulate). Further, the shape of the mesh M is not limited to a square, but may be a rectangle or a polygon such as a hexagon or an octagon. Furthermore, although the size of the mesh M can be arbitrarily determined, it is desirable that the mesh M be of a size that can be used for self-position estimation and route searching of the automatic driving vehicle C.

Further, as shown in FIG. 5, each mesh M has a mesh ID, position information, ease of water accumulation, etc. set as weather characteristic information 13b. The position information may be an absolute position such as latitude and longitude, or may be a relative position (number of rows and columns) from a reference position where the absolute position is set. The ease with which water accumulates is information indicating the ease with which water accumulates in the mesh M during rain. This may be shown in stages such as large, medium, and small as shown in FIG. 5, or may be expressed as a numerical value. In the case of a numerical value, it may be a parameter such as a coefficient that can predict the amount of accumulated water by multiplying it by the amount of precipitation, for example.

The ease with which water accumulates may be set based on past results in the area indicated by the mesh M. Alternatively, the setting may be made in consideration of the characteristics of the region, such as topography and geology (many asphalt roads and buildings, many unpaved roads, quality of drainage, etc.).

When puddles occur due to rain, road markings may not be recognized correctly by a rider or the like, and it may be determined that there are no road markings even though there are actually road markings. Therefore, by having the weather characteristic information 13b as shown in FIGS. 4 and 5, it becomes possible to recognize that the detection result by a rider or the like may be incorrect, as will be described later. Although the ease with which water accumulates due to rainfall has been described in FIGS. 4 and 5, the ease with which water accumulates due to snowfall, the ease with which water accumulates, and the like may be similarly added to the information of the mesh M. Snowfall may not be recognized correctly by a lidar or the like because it hides road markings, and ice may also not be recognized correctly by a lidar or the like due to changes in reflectance or diffused reflection. Therefore, it is useful to set these pieces of information as the weather characteristic information 13b.

Next, a method for estimating the self-position of the automatic driving vehicle C (self-position estimation method) in the vehicle control device 3 having the above-described configuration will be described with reference to the flowchart of FIG. 6. The flowchart shown in FIG. 6 is executed by the control unit 31. Further, the flowchart in FIG. 6 may be configured as a self-position estimation program executed by a computer such as a CPU that constitutes the control unit 31.

First, in step S101, the control unit 31 acquires the result detected by the sensor 4 (sensor information). This sensor information is, for example, point cloud information acquired by a lidar or image information photographed by a camera. That is, in this embodiment, the point cloud information and image information become an external situation signal indicating the external situation of the moving body, and the control unit 31 functions as a first acquisition unit. In the following description, point cloud information will mainly be used as the external situation signal.

Next, in step S102, the control unit 31 acquires the map data 33a from the storage unit 33. The map data acquired in this step may be map data around the current position of the automatic driving vehicle C. That is, the control unit 31 functions as a second acquisition unit that acquires map data including weather characteristic information.

Next, in step S103, the control unit 31 acquires weather data from the server device 2 via the communication unit 32. This weather data may be in a range that corresponds to the map data acquired in step S102.

Next, in step S104, the control unit 31 evaluates the results detected by the sensor 4. This evaluation is, for example, to determine the reliability (accuracy) of road markings, etc. included in the point cloud information acquired by the lidar. For example, if it is determined from the weather data acquired in step S103 that it is raining, weather characteristic information corresponding to the position detected by the sensor 4 is read from the map data 33a. Next, based on the amount of precipitation up to now and the ease with which water accumulates as shown in FIG. 5, it is estimated whether or not puddles are occurring on the road surface. If it is assumed that a puddle has occurred, there is a high possibility that road markings, etc., cannot be detected correctly due to the influence of the puddle, and the results detected by sensor 4 indicate that the road markings included in the point cloud information are not correct. The reliability of signs, etc. is evaluated as low. That is, the control unit 31 evaluates the accuracy of the point cloud information based on the map data 33a and weather data.

To explain using the example of FIG. 5, for example, if the amount of precipitation in the past hour is 10 mm or more and less than 30 mm, and the ease of water accumulation is "large", it is assumed that water accumulation has occurred in the mesh M, It is highly likely that the road markings included in the acquired point cloud information cannot be detected, and the reliability of the road markings included in the point cloud information is evaluated to be low. In other words, in this example, the predetermined criterion for evaluation is whether or not road markings, etc. can be detected, and there is a high possibility that road markings, etc. cannot be detected, and the reliability of road markings, etc. is low. , is considered to be lower than the predetermined standard.

Note that this evaluation may be performed by calculating numerical values. For example, the probability that a road marking or the like can be detected is calculated using a mathematical formula or a graph, and if the probability is less than a predetermined probability, it is evaluated that it cannot be detected.

Next, in step S105, if the result evaluated in step S104 is equal to or higher than a predetermined standard, in step S106, the control unit 31 determines whether the Perform position estimation. In the above example, the road markings included in the point cloud information are evaluated to be highly reliable, and in that case, the road markings and map data 33a included in the sensor information acquired in step S101 are The self-position is estimated by matching the included road markings. That is, the control unit 31 functions as a self-position estimating unit that estimates the self-position of the automatic driving vehicle C based on point cloud information.

On the other hand, in step S105, if the result evaluated in step S104 is less than the predetermined standard, in step S107, self-position estimation is performed using, for example, road markings included in the point cloud information detected by the lidar. do not have. In this case, for example, self-position estimation is performed using features other than road markings. Alternatively, self-position estimation may be performed using other methods such as GPS without performing self-position estimation by the rider, or self-position estimation may not be performed at the relevant position.

Next, in step S108, the control unit 31 determines whether or not the automated driving vehicle C has arrived at the destination. If the automated vehicle C has arrived at the destination, the flowchart ends; if the automated vehicle C has not arrived, the process returns to step S101. Run the flowchart again.

In addition, in the explanation of the flowchart mentioned above, the explanation was given using road markings, but the invention is not limited to road markings, and may be anything provided on the road, such as a manhole.

As is clear from the above description, step S101 functions as a first acquisition process, step S102 functions as a second acquisition process, step S103 functions as a third acquisition process, and step S106 functions as a self-position estimation process.

Next, a modification of the flowchart shown in FIG. 6 will be described. The flowchart in FIG. 6 shows the self-position estimation operation while the autonomous vehicle C is running, but the flowchart shown in FIG. Then, the route search is performed again based on the evaluation results.

The flowchart of FIG. 7 will be explained. First, the control unit 31 performs a route search in step S201. This route search is to search for a route to a destination input by a user or the like using a well-known method based on map data. Note that the map used for route searching may be the map data 33a, or separate map data for route searching may be stored in the storage unit 33, or may be acquired from the server device 1. That is, the control section 31 functions as a route search section.

The next steps S202 and S203 are similar to steps S102 and S103 in the flowchart of FIG. However, the acquired map data is the range of the route searched in step S101. Weather data also covers the route.

Next, in step S204, the control unit 31 evaluates the results detected by the sensor 4. This evaluation is basically performed using the same concept as that explained in step S104, but in the case of the flowchart in FIG. This method is used to predict and judge in advance how reliable (accurate) road markings etc. are in certain cases. For example, if it is found from the weather data acquired in step S203 that it is raining on the searched route, weather characteristic information corresponding to the route is read from the map data 33a. Then, based on the amount of precipitation to date and the ease with which water accumulates as shown in Figure 5, it is estimated whether or not puddles are occurring on the road surface of the route, and if it is assumed that puddles are occurring. evaluates that the result detected by the sensor 4 is likely to be unable to correctly detect road markings etc. due to the influence of water puddles.

Next, in step S205, if the result evaluated in step S204 is equal to or higher than the predetermined standard, the flowchart ends. This means that it has been determined that there is no point on the searched route where there is a high possibility that road markings etc. cannot be detected correctly due to the influence of water puddles, and the driver will drive on the route searched in step S101 without re-searching. Start.

On the other hand, in step S205, if the result evaluated in step S204 is less than the predetermined standard, a route to the destination is searched again in step S206. In other words, the search is performed again while avoiding points where there is a high possibility that road markings and the like cannot be detected correctly due to the influence of puddles.

Note that before the automatic driving vehicle C starts traveling, the process is not limited to the flowchart shown in FIG. 7, but may be a flowchart as shown in FIG. The flowchart shown in FIG. 8 searches for a route efficiently by also performing an evaluation during the route search.

The flowchart in FIG. 8 will be explained. Steps S301 and S302 are similar to steps S202 and S203 in FIG. 7. Step S303 performs evaluation and route search. For example, if the weather data indicates rain, the weather feature information is referred to when searching for a route, and the route is searched so as not to pass through areas that overlap meshes with a "large" water accumulation tendency.

According to the present embodiment, the vehicle control device 3 includes a sensor 4 that acquires point cloud information indicating the external situation of the automated driving vehicle C, a communication unit 32 that acquires weather information, and weather characteristic information 13b. It includes a control unit 31 that acquires map data 33a and estimates the self-position of the automatic driving vehicle C based on point cloud information. Then, the control unit 31 evaluates the accuracy of the point cloud information based on the map data 33a and weather information. By doing so, the automatic driving vehicle C can travel by estimating its own position according to the weather indicated by the weather information acquired by the communication unit 32, using the weather characteristic information 13b included in the map data 33a. .

Furthermore, if there is a high possibility that the road markings included in the point cloud information cannot be detected correctly, the control unit 31 does not perform self-position estimation using the road markings included in the point cloud information. By doing this, it is determined that the reliability of self-position estimation using road markings, etc. is reduced due to the influence of the weather, and the point cloud information is not used for self-position estimation, and the self-position is determined. Erroneous estimation can be prevented.

Furthermore, the map data 33a has weather characteristic information 13b set for each mesh M subdivided into predetermined sizes. By doing this, for example, features that are affected by the weather are set in a subdivided manner, so it is possible to create map data that more accurately reflects the features of the region indicated by the mesh.

The weather characteristic information 13b may also include at least one of the following information: water is likely to accumulate due to rainfall, snow is likely to accumulate due to snowfall, and freezing is likely to occur. By doing so, self-position estimation can be performed in response to cases where the road surface condition is likely to change due to rain, snowfall, or freezing.

The control unit 31 also has a function as a route search unit that searches for a route to a predetermined destination based on the map data 33a, and the control unit 31 further searches for a route based on weather information. Good too. By doing so, it becomes possible to search for a route according to the weather.

Furthermore, the map data structure of the map data 33a is the map data structure of map data used for self-position estimation. This map data 33a includes weather characteristic information 13b. By doing so, it is possible to perform self-position estimation according to the weather using the weather characteristic information 13b.

Furthermore, the storage unit 33 stores map data 33a having the map data structure described above. By doing so, it is possible to perform self-position estimation according to the weather using information indicating weather-related characteristics.

Note that the map data 33a may include information as a result of evaluating point cloud information based on the weather feature information 13b and weather information. By doing this, when the sensor such as a lidar installed in the automated driving vehicle C indicates that a feature such as a road marking cannot be detected, it is possible to determine whether or not the situation was affected by the weather. can include the results of the evaluation. Therefore, a history of weather and evaluation results can be accumulated, and by referring to this history, weather characteristic information can be used more effectively.

Further, the present invention is not limited to the above embodiments. That is, those skilled in the art can implement various modifications based on conventionally known knowledge without departing from the gist of the present invention. Such modifications are, of course, included within the scope of the present invention as long as the self-position estimating device, map data structure, and feature data storage device of the present invention are provided.

1 Server device (map data storage device)
3 Vehicle control device 4 Sensor (lidar)
13 Storage unit 13a Map data (map data structure)
31 Control unit 32 Communication unit 33 Storage unit (map data storage device)
33a Map data (map data structure)

Claims (6)

a first acquisition unit that acquires an external situation signal indicating an external situation of the mobile object;
a second acquisition unit that acquires map data in which information indicating characteristics affected by weather is set for each section subdivided into predetermined sizes;
a third acquisition unit that acquires weather information indicating weather conditions;
a self-position estimation unit that estimates the self-position of the mobile body based on the external situation signal,
The information indicating the characteristics affected by the weather includes a parameter related to at least one of the ease with which water accumulates due to rainfall, the ease with which snow accumulates due to snowfall, and the ease with which it freezes,
The self-position estimation unit performs prediction by multiplying the parameter by a value based on the weather information, and evaluates the accuracy of the external situation signal based on the prediction result .
A self-position estimating device characterized by: The self-position estimating device according to claim 1, wherein the self-position estimation unit does not use the signal for self-position estimation when the evaluation of the external situation signal is lower than a predetermined standard. . Claim characterized in that the information indicating the characteristics affected by the weather is set based on at least one of the track record in the region indicated by the section and the topographical or geological characteristics of the region indicated by the section. The self-position estimating device according to 1 or 2. comprising a route search unit that searches for a route to a predetermined destination based on the map data acquired by the second acquisition unit;
The route search unit further searches for a route based on the weather information.
The self-position estimating device according to any one of claims 1 to 3 . A self-position estimation method executed by a self-position estimation device for estimating the self-position of a moving body, the method comprising:
a first acquisition step of acquiring an external situation signal indicating the external situation of the mobile body;
a second acquisition step of acquiring map data in which information indicating characteristics affected by weather is set for each section subdivided into a predetermined size;
a third acquisition step of acquiring weather information indicating weather conditions;
a self-position estimating step of estimating the self-position of the mobile body based on the external situation signal,
The information indicating the characteristics affected by the weather includes a parameter related to at least one of the ease with which water accumulates due to rainfall, the ease with which snow accumulates due to snowfall, and the ease with which freezing occurs,
The self-position estimation step performs prediction by multiplying the parameter by a value based on the weather information, and evaluates the accuracy of the external situation signal based on the prediction result .
A self-position estimation method characterized by: A self-position estimating program that causes a computer to execute the self-position estimating method according to claim 5 .

Images