Map update interval determination device, determination method and recording medium

Abstract

This invention relates to a map update interval determination device, a determination method, and a computer program. The map update interval determination device includes: an evaluation value calculation unit (42) that calculates an evaluation value based on the switching cost between automatic and manual driving, representing the degree of improvement in driver convenience obtained by generating or updating map information for two or more road intervals selected from road intervals where map information cannot be used; and a determination unit (43) that replaces one or more of the selected two or more road intervals with other unselected road intervals, causes the evaluation value calculation unit (42) to calculate an evaluation value, and determines each of the selected two or more road intervals corresponding to the evaluation value that corresponds to the evaluation value calculated before the evaluation value meets a predetermined end condition, at which the degree of improvement in driver convenience is the highest, as the road interval to be the object of map information generation or updating.

Description

Technical Field

[0001] This invention relates to a map update interval determination device, a map update interval determination method, and a computer program for determining road intervals that are the objects of map information generation or updating. Background Technology

[0002] For autonomous driving systems of vehicles, the high-precision maps referenced for autonomous driving control require accurate representation of information related to objects on or around roads that are associated with vehicle movement. Therefore, a technique has been proposed for collecting data containing information related to objects on roads from vehicles actually traveling on roads (see International Publication No. 2018 / 180097 and Japanese Patent Application Publication No. 2020-71053).

[0003] The server device disclosed in International Publication No. 2018 / 180097 requests a second mobile body to send a status information indicating the status of a location where a first mobile body capable of autonomous driving is moving. Furthermore, if the server device receives status information from the first mobile body indicating that autonomous driving is possible, it suppresses the transmission of the status information request.

[0004] In the map information system disclosed in Japanese Patent Application Publication No. 2020-71053, the onboard device determines whether a transition from automatic driving control to manual driving, i.e., takeover, has occurred during automatic driving control. If takeover has occurred, the onboard device sets an upload target area including the location where takeover occurred. Then, the onboard device reads map information related to the upload target area from a storage device and uploads the read map information to an external device. The external device updates its external map information based on the map information uploaded from the onboard device. Summary of the Invention

[0005] As with the technologies described above, in order to collect data including information related to objects on the ground at a certain location, the vehicle actually drives at that location and generates data containing information related to objects on the ground at that location, and then sends the generated data to a device for generating or updating map information. Therefore, the collection of data including information related to objects on the ground and the generation or updating of map information based on the collected data require various costs, primarily communication costs. On the other hand, the time or budget available for generating or updating map information is limited. Therefore, to maximize the advantages of autonomous driving control for the vehicle's driver within limited time or budget, the road sections to be the objects of map information generation or updating are determined; that is, the road sections to be the objects of data collection including information related to objects on the ground are determined.

[0006] Therefore, the purpose of this invention is to provide a map update interval determination device that determines the road intervals that become the objects of map information generation or updating in each road interval in a way that allows the driver to easily obtain autonomous driving control.

[0007] According to one embodiment, a map update interval determination apparatus is provided. This apparatus includes: a storage unit that stores map maintenance costs, map markers, and switching costs for each of multiple road intervals included in a predetermined area; the map maintenance cost is the cost required to generate or update map information for use by a vehicle in autonomous driving for that road interval; the map marker indicates whether map information is usable in that road interval; and the switching cost indicates the driver's workload when switching between autonomous and manual driving occurs in that road interval and in adjacent road intervals among the multiple road intervals; and an initial selection unit that, referring to the map markers of each of the multiple road intervals, selects two road intervals from those where map information is unusable, such that the sum of the map maintenance costs of the selected road intervals is below a target maintenance cost upper limit. The system includes: a road section; an evaluation value calculation unit that calculates an evaluation value based on the switching cost of each of the two or more selected road sections, the evaluation value representing the degree of improvement in driver convenience obtained by generating or updating map information for the two or more selected road sections; and a determination unit that replaces one or more of the two or more selected road sections with other unselected road sections from the road sections where map information cannot be used, and causes the evaluation value calculation unit to calculate the evaluation value, determines whether the evaluation value meets a predetermined termination condition, and when the predetermined termination condition is met, determines each of the two or more selected road sections corresponding to the evaluation value at which the degree of improvement in driver convenience is the highest among the calculated evaluation values ​​as the road sections to be generated or updated with map information.

[0008] In this map update interval determination device, preferably, the storage unit also stores the traffic volume and autonomous driving cost for each of the multiple road intervals, whereby the autonomous driving cost represents the driver's workload when the vehicle autonomously drives using map information within that road interval. Furthermore, preferably, the evaluation value calculation unit calculates, for each of the two or more selected road intervals, a weighted sum of the autonomous driving cost and switching cost for that road interval using a weighting coefficient corresponding to the traffic volume of that road interval, as the evaluation value.

[0009] Alternatively, in this map update interval determination device, it is preferable that the storage unit also stores, for each of the multiple road intervals, the traffic volume of that road interval, the autonomous driving cost representing the driver's workload when the vehicle uses map information to drive autonomously in that road interval, and the manual driving cost representing the driver's workload when the vehicle drives manually in that road interval. Furthermore, it is preferable that the evaluation value calculation unit calculates, for each of the two or more selected road intervals, a weighted sum of values ​​obtained by subtracting the manual driving cost from the autonomous driving cost and adding the switching cost, using a weighting coefficient corresponding to the traffic volume of that road interval, as the evaluation value.

[0010] Furthermore, in this map update interval determination device, it is preferable that, when the determination unit replaces any one of the two or more selected road intervals with other road intervals, the road interval with the larger value obtained by multiplying the individual interval evaluation value calculated for that road interval by the map maintenance cost is more likely to be replaced with other road intervals.

[0011] Furthermore, in this map update interval determination device, it is preferable that the storage unit also stores the traffic volume of each road interval among the multiple road intervals. Also preferably, the initial selection unit selects two or more road intervals from the multiple road intervals from which map information cannot be used, in descending order of the ratio of traffic volume to map maintenance cost.

[0012] According to another approach, a method for determining map update intervals is provided. This method includes: referring to map markers for each road interval within a predetermined area, indicating whether map information for autonomous driving of a vehicle is available in that road interval; selecting two or more road intervals from those road intervals where map information is unavailable, such that the sum of map maintenance costs required to generate or update map information for the selected road intervals is below a target maintenance cost upper limit; and calculating an evaluation value based on switching costs, representing the improvement in driver convenience gained by generating or updating map information for the selected two or more road intervals, where the switching cost represents the improvement in driver convenience gained by generating or updating map information for the selected road intervals. In two or more selected road sections, and in each of the road sections and the adjacent road sections, the driver's load during the switch between autonomous driving and manual driving is considered. One or more of the selected road sections are replaced with other unselected road sections in the road sections where map information cannot be used. An evaluation value is calculated, and it is determined whether the evaluation value meets a predetermined termination condition. If the predetermined termination condition is met, the road sections of the two or more selected road sections corresponding to the evaluation value that provides the highest improvement in driver convenience are identified as the road sections that become the objects of map information generation or updating.

[0013] According to another approach, a computer program for determining map update intervals is provided. This computer program includes commands to cause a computer to perform the following actions: referring to map markers for each of a plurality of road intervals encompassing a predetermined area, indicating whether map information for autonomous driving of a vehicle is available in that road interval; selecting two or more road intervals from the plurality of road intervals where map information is unavailable, such that the total map maintenance cost required to generate or update map information for the selected road intervals is below a target maintenance cost upper limit; and calculating an evaluation value based on switching costs representing the degree of improvement in driver convenience obtained by generating or updating map information for the selected two or more road intervals. Switching cost represents the driver's workload when switching between autonomous driving and manual driving in each of the selected two or more road intervals and in the road intervals adjacent to the selected road interval. The selected two or more road intervals are replaced with other unselected road intervals in the road intervals that cannot utilize map information, and an evaluation value is calculated. It is determined whether the evaluation value meets the predetermined termination condition. When the predetermined termination condition is met, the road intervals of the selected two or more road intervals corresponding to the evaluation value that provides the highest improvement in driver convenience are identified as the road intervals that become the objects of map information generation or updating.

[0014] The map update interval determination device disclosed herein has the effect of determining the road intervals that become the objects of map information generation or updating in a manner that allows the driver to easily obtain autonomous driving control. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a map update system equipped with a map update interval determination device.

[0016] Figure 2 It is a rough structural diagram of the vehicles included in the map update system.

[0017] Figure 3 This is a hardware structure diagram of the data acquisition device mounted on the vehicle.

[0018] Figure 4 This is a hardware structure diagram of a server, which serves as an example of a map update interval determination device.

[0019] Figure 5 This is a functional block diagram of the processor of the server associated with map update processing, including map update range determination processing.

[0020] Figure 6This is a diagram illustrating the general principles of the evaluation value calculation.

[0021] Figure 7 This is a flowchart of the actions involved in determining the map update interval. Detailed Implementation

[0022] Hereinafter, with reference to the accompanying drawings, the map update interval determination device, the map update interval determination method executed by the map update interval determination device, and the computer program for map update interval determination will be described. The map update interval determination device selects road intervals from a plurality of road intervals encompassed in an area where map information used in autonomous driving control is generated or updated, in a manner that allows multiple drivers to enjoy the advantages of autonomous driving control. At this time, the map update interval determination device selects two or more road intervals such that the sum of the costs required for map maintenance of the selected road intervals is below a target maintenance cost upper limit. The map update interval determination device replaces one of the two or more selected road intervals with another unselected road interval and calculates an evaluation value representing the degree of improvement in driver convenience obtained by generating or updating map information for each of the selected road intervals. Furthermore, when the calculated evaluation value meets a predetermined end criterion, the map update interval determination device determines the road intervals corresponding to the evaluation value with the highest degree of convenience improvement among the evaluation values ​​calculated up to this point as the road intervals to be generated or updated with map information. The map update interval determination device, for the determined road interval, notifies the vehicle capable of generating the ground object data of a collection instruction for collecting data representing ground objects associated with the vehicle's movement (hereinafter referred to as ground object data).

[0023] Figure 1 This is a schematic structural diagram of a map update system equipped with a map update interval determination device. In this embodiment, the map update system 1 includes at least one vehicle 2 and a server 3, which is an example of a map update interval determination device. The vehicle 2 connects to the server 3, for example, by accessing a wireless base station 5 connected to a communication network 4 via a gateway (not shown) and thus via the wireless base station 5 and the communication network 4. It should be noted that in Figure 1 In this diagram, only one vehicle 2 is shown, but the map update system 1 can also have multiple vehicles 2. Similarly, multiple wireless base stations 5 can be connected to the communication network 4. In addition, the server 3 can also be connected to a traffic information server (not shown) that manages traffic information via the communication network 4 to communicate.

[0024] Figure 2This is a schematic diagram of vehicle 2. Vehicle 2 includes a camera 11, a GPS receiver 12, a wireless communication terminal 13, and a data acquisition device 14. The camera 11, GPS receiver 12, wireless communication terminal 13, and data acquisition device 14 are connected to communicate via an in-vehicle network that follows standards such as a controller area network.

[0025] Camera 11 is an example of an imaging unit used to photograph the area around vehicle 2. It has a two-dimensional detector composed of an array of photoelectric conversion elements sensitive to visible light, such as a CCD or C-MOS, and an imaging optical system that images the area to be photographed onto the two-dimensional detector. Furthermore, camera 11 is mounted, for example, inside the passenger compartment of vehicle 2, facing forward of vehicle 2. Camera 11 photographs the area in front of vehicle 2 at a predetermined shooting cycle (e.g., 1 / 30 second to 1 / 10 second), generating an image reflecting that area. The image obtained by camera 11 can be either a color image or a grayscale image. It should be noted that multiple cameras 11 with different shooting directions or focal lengths may also be provided in vehicle 2.

[0026] Each time an image is generated, the camera 11 outputs the generated image to the data acquisition device 14 via the in-vehicle network.

[0027] GPS receiver 12 receives GPS signals from GPS satellites at predetermined intervals and determines the vehicle 2's own position based on the received GPS signals. Additionally, GPS receiver 12 outputs positioning information, representing the vehicle 2's own position based on GPS signals, to data acquisition device 14 via the in-vehicle network at predetermined intervals. It should be noted that vehicle 2 may also have a receiver based on a satellite positioning system other than GPS receiver 12. In this case, that receiver can determine the vehicle 2's own position.

[0028] Wireless communication terminal 13 is a device that performs wireless communication processing according to a predetermined wireless communication standard, for example, by accessing wireless base station 5 and connecting to server 3 via wireless base station 5 and communication network 4. Wireless communication terminal 13 receives downlink wireless signals from wireless base station 5 and sends a collection instruction for ground object data from server 3 contained in the wireless signals to data acquisition device 14. Additionally, wireless communication terminal 13 generates uplink wireless signals containing ground object data received from data acquisition device 14. Wireless communication terminal 13 then sends the ground object data to server 3 by transmitting the uplink wireless signals to wireless base station 5.

[0029] Figure 3This is a hardware structure diagram of the data acquisition device. The data acquisition device 14 performs processing related to the generation of ground object data based on images generated by the camera 11. Therefore, the data acquisition device 14 has a communication interface 21, a memory 22, and a processor 23.

[0030] The communication interface 21 has an interface circuit for connecting the data acquisition device 14 to the in-vehicle network. Specifically, the communication interface 21 connects to the camera 11, the GPS receiver 12, and the wireless communication terminal 13 via the in-vehicle network. Furthermore, each time the communication interface 21 receives an image from the camera 11, it passes the received image to the processor 23. Additionally, each time the communication interface 21 receives location information from the GPS receiver 12, it passes the received location information to the processor 23. Moreover, the communication interface 21 outputs ground object data received from the processor 23 to the wireless communication terminal 13 via the in-vehicle network. Furthermore, the communication interface 21 sends a collection instruction for ground object data received from the server 3 via the wireless communication terminal 13 to the processor 23.

[0031] The memory 22 may be, for example, a volatile semiconductor memory or a non-volatile semiconductor memory. The memory 22 may also include other storage devices such as a hard disk drive. Furthermore, the memory 22 stores various data used in the processing associated with the generation of ground object data executed by the processor 23 of the data acquisition device 14. Such data includes, for example, road maps, vehicle 2 identification information, camera 11 parameters such as setup height, shooting direction, and field of view, and a parameter set for determining an identifier for detecting ground objects from the image. It should be noted that the road map may be, for example, a map used in a navigation device, containing information such as the location, length, and connection relationships of each road section within the area represented by the road map. Additionally, the memory 22 may store images received from the camera 11 and positioning information received from the GPS receiver 12 for a certain period. Moreover, the memory 22 stores information indicating the road sections (hereinafter sometimes referred to as "collection target sections") designated by the ground object data collection instruction as the objects to be generated and collected for ground object data. Furthermore, the memory 22 may also store computer programs and the like for implementing various processes executed by the processor 23.

[0032] The processor 23 has one or more CPUs (Central Processing Units) and their peripheral circuits. The processor 23 may also have other arithmetic circuits such as logic units, numerical processing units, or graphics processing units. Furthermore, the processor 23 stores images received from the camera 11 and positioning information received from the GPS receiver 12 in the memory 22. Moreover, while the vehicle 2 is in motion, the processor 23 performs processing related to the generation of ground object data at predetermined intervals (e.g., 0.1 seconds to 10 seconds).

[0033] As part of the processing associated with the generation of ground object data, the processor 23 determines, for example, whether the vehicle 2's location, as indicated by the positioning information received from the GPS receiver 12, is included within the target collection area. If the vehicle's location is included within the target collection area, the processor 23 generates ground object data based on images received from the camera 11.

[0034] For example, processor 23 may use the image itself received from camera 11 (hereinafter, sometimes referred to as the "overall image") as ground object data. Alternatively, processor 23 may extract a portion of the image from the overall image received from camera 11, including the area representing the road surface, and use this extracted portion as ground object data. It should be noted that information indicating the area presumed to represent the road surface in the overall image may be pre-stored in memory 22. Furthermore, processor 23 may determine the area to be extracted from the overall image by referring to the information representing that area.

[0035] Alternatively, processor 23 may input the entire image or a portion of the image into a pre-learned recognizer that detects ground objects that are the objects to be detected, thereby detecting the ground objects represented by the input entire image or a portion of the image (hereinafter, sometimes simply referred to as the input image). Furthermore, processor 23 may generate information representing the types of detected ground objects as ground object data. As such a recognizer, processor 23 may, for example, use a so-called deep neural network (DNN) pre-learned to detect ground objects shown in the input image. Such a DNN may, for example, use a Single Shot MultiBox Detector (SSD) or a Faster R-CNN, a DNN with a convolutional neural network (CNN) type architecture. In this case, the recognizer calculates the confidence level of the accuracy of representing a ground object in various regions of the input image, according to the type of ground object being detected (e.g., lane markings, pedestrian crossings, temporary stop lines, etc.). The recognizer determines that a ground object of that type is represented in regions where the confidence level for any type of ground object is above a predetermined detection threshold. Furthermore, the recognizer outputs information representing the region on the input image containing the ground object that is the target of detection (e.g., the bounding rectangle of the ground object that is the target of detection, hereinafter referred to as the "object region") and information representing the type of ground object represented in the object region. Therefore, the processor 23 can generate ground object data in a manner that includes information representing the type of ground object shown in the detected object region.

[0036] Furthermore, the processor 23 determines the location represented by the ground object data or the actual spatial position of the ground object, and includes information indicating that position in the ground object data. For example, the processor 23 uses the vehicle 2's position at the time the image used to generate the ground object data is generated as the location represented by the ground object data. In this case, the processor 23 can use the position represented by the positioning information received from the GPS receiver 12 at the time closest to the time the image used to generate the ground object data is generated as the vehicle 2's position. Alternatively, if the vehicle 2's position is estimated by the vehicle 2's electronic control unit (ECU, not shown), the processor 23 can also obtain information indicating the estimated vehicle 2's position from the ECU via the communication interface 21. Furthermore, the processor 23 can obtain information indicating the vehicle 2's direction of travel from the ECU. In addition, the position of each pixel in the image corresponds one-to-one with the orientation from the camera 11 toward the object represented by that pixel. Therefore, when the ground object data is a whole image or a partial image, the processor 23 can also estimate the actual spatial position corresponding to the center of the whole image or partial image as the location represented by the ground object data. In this case, the processor 23 can estimate the location corresponding to the center of the overall image or a portion of the image based on parameters such as the orientation relative to the camera 11, the vehicle 2's own position, the vehicle 2's direction of travel, and the camera 11's shooting direction, field of view, and setting height. Alternatively, if the ground object data includes information indicating the types of ground objects detected from the image, the processor 23 can estimate the location of the ground object represented by the object region based on the orientation relative to the camera 11 corresponding to the center of gravity of the object region, the vehicle 2's position, direction of travel, and the camera 11's parameters. Alternatively, the processor 23 can also estimate the location of the ground object represented by the ground object data using so-called Structure from Motion (SfM). In this case, the processor 23 uses optical flow to correlate object regions representing the same ground object between two images acquired at different times. Furthermore, the processor 23 can estimate the location of the ground object using triangulation based on the vehicle 2's position and direction of travel when the two images were acquired, the camera 11's parameters, and the location of the object region in each image.

[0037] The processor 23 includes the latitude and longitude of the location or position of the object represented by the ground object data as information indicating the location or position of the object represented by the ground object data in the ground object data. Furthermore, the processor 23 refers to a road map to determine the road segment containing the location or position of the object represented by the ground object data, or the road section closest to that location. The processor 23 can also include the identification number of the determined road segment in the ground object data. Additionally, if the ground object data is a complete image or a partial image, the position and direction of travel of the vehicle 2 at the time the image was generated, as well as the parameters of the camera 11, can be included in the ground object data so that the server 3 can estimate the position of the object represented by the complete image or partial image.

[0038] Each time ground cover data is generated, processor 23 outputs the generated ground cover data to wireless communication terminal 13 via communication interface 21. Thus, the ground cover data is sent to server 3.

[0039] Next, server 3, which serves as an example of a map update interval determination device, will be explained.

[0040] Figure 4 This is a hardware structure diagram of server 3, which serves as an example of a map update interval determination device. Server 3 has a communication interface 31, a storage device 32, a memory 33, and a processor 34. The communication interface 31, the storage device 32, and the memory 33 are connected to the processor 34 via signal lines. Server 3 may also have input devices such as a keyboard and mouse, and a display device such as an LCD screen.

[0041] Communication interface 31 is an example of a communication unit, and has interface circuitry for connecting server 3 to communication network 4. Furthermore, communication interface 31 is configured to communicate with vehicle 2 via communication network 4 and wireless base station 5. Specifically, communication interface 31 transmits collection instructions received from processor 34 to vehicle 2 via communication network 4 and wireless base station 5. Additionally, communication interface 31 transmits ground object data received from vehicle 2 via wireless base station 5 and communication network 4 to processor 34.

[0042] Storage device 32 is an example of a storage unit, such as having a hard disk device or an optical recording medium and its access device. Storage device 32 stores various data and information used in the map update interval determination process. For example, storage device 32 stores information indicating areas that are the objects of map information generation or updating, road maps, and information for identifying each road interval and indicating the connection relationships between each road interval. Furthermore, storage device 32 stores map markers, autonomous driving costs, manual driving costs, traffic volume, map maintenance costs, and switching costs with adjacent road intervals for each road interval. Additionally, storage device 32 stores an upper limit value of the target cost in map information generation or updating (hereinafter referred to as the "target maintenance cost upper limit value"). Additionally, storage device 32 stores a table showing the relationship between traffic volume and the weighting coefficients corresponding to traffic volume used in the calculation of evaluation values. Additionally, storage device 32 stores ground feature data received from vehicle 2. Furthermore, storage device 32 may also store a computer program executed on processor 34 for performing the map update interval determination process.

[0043] Each road section can be, for example, defined as a road segment or node represented on the road map used by the navigation device (not shown). That is, a road section is defined as the distance from a point where multiple roads intersect, branch, or merge (hereinafter, for convenience, referred to as an intersection or merging point) to an adjacent intersection or merging point. Furthermore, if such a road section is longer than a predetermined distance, it can be divided into multiple road sections. Furthermore, intersections and merging points are also defined as road sections.

[0044] Map markers indicate whether map information can be used to enable a vehicle to drive autonomously within the road section corresponding to that marker. Specifically, when a map marker has a value indicating that map information can be used (e.g., 1), it means that the map information includes information related to terrain features for the road section corresponding to that marker, to a degree sufficient for the vehicle to perform autonomous driving. Conversely, when a map marker has a value indicating that map information cannot be used (e.g., 0), it means that the map information does not include information related to terrain features necessary for the vehicle to perform autonomous driving within the road section corresponding to that marker. Hereinafter, road sections where map information can be used during autonomous driving will sometimes be referred to as "maintained sections." Additionally, road sections where map information cannot be used during autonomous driving will sometimes be referred to as "unmaintained sections."

[0045] For each road segment, the costs of autonomous driving, manual driving, and switching are set, with the value increasing as the driver's workload increases. The autonomous driving cost represents the driver's workload when the vehicle uses map information to autonomously drive within the corresponding road segment. For example, it is set as the cost per unit time of autonomous driving multiplied by the average time required to traverse the corresponding road segment. It should be noted that since the driver's workload during autonomous driving is relatively low, the autonomous driving cost can also be 0. The manual driving cost represents the driver's workload when the vehicle travels within the corresponding road segment manually. For example, it is set as the cost per unit time of manual driving multiplied by the average time required to traverse the corresponding road segment. Generally, the workload for manual driving is greater than that for autonomous driving, therefore the manual driving cost is greater than the autonomous driving cost for the same road segment. Furthermore, the switching cost represents the driver's workload when switching between autonomous and manual driving between two adjacent road segments. Therefore, when no switch occurs between two consecutive road segments, the switching cost is 0. For road sections including merging and intersection points, map markers, autonomous driving costs, manual driving costs, and switching costs are set for each combination of roads accessible to vehicles connected to that road section. For example, for road sections including intersections, map markers, autonomous driving costs, manual driving costs, and switching costs are set for each road connected to that intersection, for the directions of going straight, turning left, and turning right. However, if any road connected to an intersection prohibits entry in any of the directions of going straight, turning right, or turning left, map markers, autonomous driving costs, manual driving costs, and switching costs may not be set for the prohibited directions. It should be noted that the switching cost can also be set to the same value for all road sections.

[0046] Additionally, map maintenance cost represents the cost required to generate or update map information for a corresponding road segment. The longer the corresponding road segment, or the more lanes it contains, the higher the map maintenance cost will be for road segments that do not include merging or crossing points. Furthermore, for each combination of roads that connect at merging or crossing points and through which vehicles can pass, the map maintenance cost for the road segments including those merging or crossing points will be set to a value proportional to the number of lanes in that combination.

[0047] In addition, traffic volume for each road section is obtained from the traffic information server via communication network 4.

[0048] Memory 33 is another example of a storage unit, such as a non-volatile semiconductor memory and a volatile semiconductor memory. Furthermore, memory 33 temporarily stores various data generated during the execution of the map update interval decision processing.

[0049] Processor 34 is an example of a control unit, having one or more CPUs (Central Processing Units) and their peripheral circuitry. Processor 34 may also have other arithmetic circuitry such as logic units or numerical processing units. Furthermore, processor 34 performs map update processing, including map update interval determination processing.

[0050] Figure 5 This is a functional block diagram of a processor 34 associated with map update processing, including map update interval determination processing. The processor 34 includes an initial selection unit 41, an evaluation value calculation unit 42, a determination unit 43, a collection instruction unit 44, and a map update unit 45. These units of the processor 34 are, for example, functional modules implemented by a computer program that operates on the processor 34. Alternatively, these units of the processor 34 may be dedicated arithmetic circuits provided in the processor 34. It should be noted that the initial selection unit 41, the evaluation value calculation unit 42, and the determination unit 43 of the processor 34 perform processing associated with the map update interval determination processing.

[0051] The initial selection unit 41 selects two or more candidate road sections from among the multiple road sections included in the area where map information is generated or updated. These candidate road sections are the objects of evaluation value calculation, i.e., the objects of surface object data collection. Hereinafter, the candidate road sections that will become objects of surface object data collection will sometimes be referred to as candidate sections. For example, the initial selection unit 41 determines unmaintained sections by referring to the map markers of each road section. Furthermore, from the determined unmaintained sections, the initial selection unit 41 selects two or more unmaintained sections as candidate sections in descending order of the value obtained by multiplying the ratio of traffic volume to map maintenance cost by the difference between manual driving cost and autonomous driving cost. When the total map maintenance cost of the selected candidate sections reaches the target maintenance cost upper limit, the initial selection unit 41 stops selecting candidate sections. It should be noted that if the total map maintenance cost exceeds the target maintenance cost upper limit, the initial selection unit 41 removes the last selected unmaintained section from the candidate sections, thereby ensuring that the total map maintenance cost does not exceed the target maintenance cost upper limit.

[0052] It should be noted that the initial selection unit 41 is not limited to the selection method described above; it can also select candidate intervals using other methods. For example, the initial selection unit 41 can randomly select two or more unmaintained intervals as candidate intervals. Alternatively, the initial selection unit 41 can select two or more unmaintained intervals as candidate intervals in order of proximity to predetermined locations within the area where map information is generated or updated, from closest to furthest. Alternatively, the initial selection unit 41 can also select two or more unmaintained intervals as candidate intervals in order of descending road standards.

[0053] The initial selection unit 41 notifies the evaluation value calculation unit 42 of the information of each candidate interval selected.

[0054] The evaluation value calculation unit 42 calculates evaluation values ​​for each candidate interval when it is notified of information about each candidate interval selected by the initial selection unit 41, or each time it is notified of information about a candidate interval replaced by the determination unit 43. The evaluation value represents the degree of improvement in driver convenience obtained by generating or updating map information.

[0055] In this embodiment, in order to calculate the overall evaluation value of the selected candidate intervals, the individual interval evaluation value Ei is calculated for each candidate interval as shown below.

[0056] Ei=(Ac+Hc) / Tv

[0057] Here, parameter Ac represents the autonomous driving cost for the candidate interval of interest. Parameter Hc represents the total switching cost to each unmaintained interval in the road interval adjacent to the candidate interval of interest. It should be noted that if an unmaintained interval adjacent to the candidate interval of interest is also selected as a candidate interval, this selected unmaintained interval can also utilize map information after collecting ground object data, thus incurring no switching cost. Furthermore, for two adjacent road intervals that switch, the switching cost can be calculated for either of these two road intervals. Furthermore, parameter Tv represents a weighting coefficient related to the traffic volume of the candidate interval of interest. For example, the higher the traffic volume of the candidate interval of interest, the larger the value of parameter Tv is set. It should be noted that, as shown in the above formula, when the autonomous driving cost is set to 0, the individual interval evaluation value Ei is a value calculated based on the switching cost and traffic volume.

[0058] The evaluation value calculation unit 42 sums the individual interval evaluation values ​​Ei calculated for each candidate interval as the overall evaluation value for the selected candidate intervals. Therefore, the more continuous the selected candidate intervals are, or the more they are connected via maintained intervals, the more continuously the vehicle can perform autonomous driving, and the lower the switching costs. Thus, the evaluation value becomes smaller. Furthermore, the higher the traffic volume in the selected candidate intervals, the lower the evaluation value. Therefore, the lower the evaluation value, the greater the improvement in driver convenience obtained through generating or updating map information.

[0059] Figure 6 This is a diagram illustrating the summary of the evaluation value calculation. Figure 6 The example shown depicts five road sections 601-605. Road sections 601-604 are each defined for a single road, and road section 605 includes a cross-shaped intersection connecting the various road sections 601-604. Figure 6 In this example, lines represent the various paths connecting road sections 601 to 604 that can be traversed via the intersection. Road sections 601, 602, and 605 (including the intersection) are unmaintained sections and candidate sections. Road section 603 is a maintained section, and road section 604 is an unmaintained section that is not a candidate section. In this case, a handover occurs when traveling from any of road sections 601 to 603 to road section 604 via road section 605. Although road sections 601, 602, and 605 are unmaintained sections at the current time, they become maintained sections through map maintenance. Therefore, when calculating the evaluation value, traveling from road section 601 to road section 602 via road section 605, or vice versa, can be treated as a case where no handover occurs (i.e., the handover cost is 0). Therefore, if the weight coefficient Tv of each road section is... Figure 6 The values ​​shown indicate that the cost of autonomous driving in each road section is 1, and the switching cost is 5. Therefore, the evaluation value is 1 / 3+1 / 9+(1 / 1+1 / 1+(1+5) / 1+1 / 5+(1+5) / 3+(1+5) / 4)=1093 / 90.

[0060] The evaluation value calculation unit 42 notifies the determination unit 43 of the calculated evaluation value.

[0061] The determination unit 43 replaces one or more of the selected candidate intervals with other unmaintained intervals that were not selected as candidate intervals, and the evaluation value calculation unit 42 calculates an evaluation value. Each time an evaluation value is calculated, the determination unit 43 stores the calculated evaluation value and information used to identify the corresponding candidate intervals in the memory 33. Furthermore, when the evaluation value meets a predetermined termination condition, the determination unit 43 determines the minimum value among the calculated evaluation values. As described above, the lower the evaluation value, the greater the improvement in driver convenience obtained by generating or updating map information. Therefore, when the calculated evaluation value is the minimum, the improvement in driver convenience obtained by generating or updating map information becomes the highest. Therefore, the determination unit 43 determines the candidate intervals corresponding to the minimum calculated evaluation value as the road intervals to be generated or updated for map information.

[0062] Each time the evaluation value is calculated, the determination unit 43 replaces several candidate intervals in the candidate interval with other unmaintained intervals in a manner that reduces the evaluation value. Furthermore, each time the candidate interval is replaced, the evaluation value calculation unit 42 calculates the evaluation value.

[0063] In this embodiment, the determination unit 43 determines the candidate intervals to be replaced using a method based on simulated annealing. That is, the fewer trials required to replace a candidate interval, the more candidate intervals the determination unit 43 can determine. When determining the number of candidate intervals to be replaced, the determination unit 43 decides which candidate intervals to replace according to a predetermined replacement criterion. For example, the determination unit 43 sets an exclusion probability for each candidate interval based on the principle that the larger the value of (individual interval evaluation value × map maintenance cost), the easier it is to exclude it from the evaluation value calculation. Furthermore, the determination unit 43 determines which candidate intervals are actually excluded from the evaluation value calculation based on the set exclusion probability. For example, the determination unit 43 assigns a random number generated according to a predetermined random number generation method to each candidate interval, based on the assigned random number generation probability. Furthermore, the determination unit 43 excludes candidate intervals whose generated probability is lower than the exclusion probability from the evaluation value calculation.

[0064] Furthermore, the determination unit 43 selects unmaintained intervals from those that have never been candidate intervals and adds them to the evaluation value calculation objects as candidate intervals. At this time, the determination unit 43 sets an addition probability for each unmaintained interval, based on the principle that the smaller the value of (individual interval evaluation value × map maintenance cost), the more likely it is to be added as a candidate interval to the evaluation value calculation objects. Then, the determination unit 43 determines the unmaintained intervals from each unmaintained interval that are added as candidate intervals to the evaluation value calculation objects according to the set addition probability. At this time, the determination unit 43 determines the unmaintained intervals to be added to the evaluation value calculation objects through the same steps as the step of determining candidate intervals excluded from the evaluation value calculation objects.

[0065] It should be noted that the determination unit 43 may also set the exclusion probability for each candidate interval in a manner that the larger the evaluation value of a candidate interval, the higher the probability of exclusion. Alternatively, the determination unit 43 may also set the exclusion probability for each candidate interval in a manner that the smaller the difference between the manual driving cost and the autonomous driving cost, the higher the probability of exclusion.

[0066] Alternatively, the determination unit 43 may set an addition probability for each unmaintained interval that is not a candidate interval, in a manner that the larger the difference between the manual driving cost and the autonomous driving cost, the higher the addition probability. Or, the determination unit 43 may set an addition probability for each unmaintained interval that is not a candidate interval, in a manner that the addition probability of an unmaintained interval adjacent to any of the candidate intervals already selected as evaluation targets is higher than the addition probability of other unmaintained intervals.

[0067] The determining unit 43 selects unmaintained intervals to be added to the evaluation value calculation object based on the principle that the sum of the map maintenance costs of all replacement candidate intervals does not exceed the target maintenance cost upper limit. Preferably, the determining unit 43 determines the unmaintained intervals to be added as candidate intervals in such a way that the sum of the map maintenance costs of all replacement candidate intervals is as close as possible to the target maintenance cost upper limit. Therefore, the determining unit 43 can include more unmaintained intervals as road intervals that become objects of map information generation or updating. It should be noted that the total number of candidate intervals that become objects of evaluation value calculation may differ before and after the replacement.

[0068] The determination unit 43 determines whether a predetermined termination condition is met each time an evaluation value is calculated. The predetermined termination condition can be, for example, set as the difference between the absolute values ​​of the calculated evaluation values ​​in the most recent predetermined calculation being less than or equal to a predetermined value. Alternatively, the predetermined termination condition can be set as the evaluation value being less than or equal to a termination baseline value. Alternatively, the predetermined termination condition can be set as the number of trial calculations of the evaluation value being greater than or equal to a predetermined number of termination determinations. When any of these termination conditions is met, the determination unit 43 terminates the replacement of candidate intervals and the calculation of evaluation values ​​performed by the evaluation value calculation unit 42. Alternatively, the determination unit 43 may terminate the replacement of candidate intervals and the calculation of evaluation values ​​performed by the evaluation value calculation unit 42 when two or more of these termination conditions are met. Furthermore, the determination unit 43 determines the candidate intervals corresponding to the minimum value among the evaluation values ​​calculated so far as the road intervals that become the objects of map information generation or updating.

[0069] The determination unit 43 notifies the collection instruction unit 44 of the information of the road section used to identify the road section that has been determined to be the object of map information generation or updating.

[0070] The collection instruction unit 44 generates a collection instruction for collecting ground feature data for road sections that are the targets of map information generation or updating, as notified by the determination unit 43. Specifically, the collection instruction unit 44 generates the collection instruction in a manner that includes information for identifying road sections that are the targets of map information generation or updating. Furthermore, the collection instruction unit 44 sends the generated collection instruction to the vehicle 2 via the communication interface 31.

[0071] The map updating unit 45 adds information related to the ground features represented by the collected ground feature data to the map information read from the storage device 32 for each road section that is the object of map information generation or updating, thereby generating or updating the map information. For example, when the ground feature data is a whole image or a partial image, the map updating unit 45 performs the same processing as the data acquisition device 14 mounted on the vehicle 2, detecting ground features and their types from the whole image or partial image, and estimating the location of the detected ground features. Then, for ground features of the same type located within a predetermined range, the map updating unit 45 determines the location of the ground feature by the average of the location of the ground feature contained in the collected ground feature data or the location of the ground feature estimated as described above. Then, for each ground feature whose location has been determined, the map updating unit 45 updates the map information by including information indicating the type of ground feature and the determined location in the map information.

[0072] Figure 7This is the flowchart for the map update interval determination process in server 3. The processor 34 of server 3 executes the map update interval determination process according to the flowchart shown below.

[0073] The initial selection unit 41 of the processor 34 selects two or more candidate unmaintained intervals from among the multiple unmaintained intervals included in the area where map information is generated or updated, which are the objects of evaluation value calculation and are used for collecting ground feature data (step S101). The evaluation value calculation unit 42 of the processor 34 calculates an evaluation value representing the degree of improvement in driver convenience obtained by generating or updating map information for the selected candidate intervals (step S102).

[0074] The determination unit 43 of the processor 34 determines whether the calculated evaluation value meets the predetermined termination condition (step S103). If the termination condition is not met (step S103 - "No"), the determination unit 43 replaces several candidate intervals in the candidate intervals with other unmaintained intervals (step S104). At this time, as described above, the fewer times the evaluation value is calculated, the more candidate intervals the determination unit 43 replaces. Then, the processor 34 repeats the processing after step S102.

[0075] On the other hand, if the termination condition is met (step S103 - "Yes"), the determining unit 43 determines each candidate interval corresponding to the minimum value among the evaluation values ​​calculated so far as the road interval to be the object of map information generation or updating (step S105). Then, the processor 34 ends the map update interval determination process. After the map update interval determination process is completed, the collection instruction unit 44 of the processor 34 generates a collection instruction for collecting ground feature data for the road intervals that are the objects of map information generation or updating, and sends the generated collection instruction to the vehicle 2 via the communication interface 31. In addition, when a predetermined amount or more of ground feature data for the road intervals that are the objects of map information generation or updating is collected, the map updating unit 45 of the processor 34 adds information related to the ground features represented by the collected ground feature data to the map information for each road interval that is the object of map information generation or updating.

[0076] As explained above, the map update interval determination device sequentially replaces candidate intervals selected from each unmaintained interval within the region where map information is to be generated or updated, and calculates an evaluation value. Furthermore, when the calculated evaluation value meets a predetermined termination condition, the map update interval determination device determines the candidate intervals corresponding to the evaluation value that provides the highest improvement in driver convenience among the evaluation values ​​calculated so far as the road intervals to be generated or updated with map information. Therefore, the map update interval determination device can determine the road intervals to be generated or updated with map information in a way that allows the driver to easily obtain the advantage of autonomous driving control. In particular, the map update interval determination device considers the switching costs of each candidate interval when calculating the evaluation value. Therefore, the map update interval determination device can select as many consecutive road intervals as possible as the map information to be generated or updated with map information, minimizing switching costs. As a result, the road intervals to be generated or updated with map information are selected in a way that increases the total length of road intervals where autonomous driving can be continuously applied. Therefore, the map update interval determination device can determine the road intervals to be generated or updated with map information in a way that allows the driver to obtain the maximum advantage of autonomous driving control. In addition, the map update interval determination device can determine the road intervals that will be used for map information generation or updating even without calculating the evaluation value of all road interval combinations, thus reducing the hardware resources used to determine the road intervals.

[0077] According to a modified example, the evaluation value calculation unit 42 can also calculate the individual interval evaluation value Ei for each of the selected candidate intervals according to the following formula.

[0078] Ei=(Ac-Mc+Hc) / Tv

[0079] As described above, parameter Ac represents the autonomous driving cost for the candidate section of interest. Additionally, parameter Mc represents the manual driving cost for the candidate section of interest. Furthermore, parameter Hc represents the sum of the respective switching costs to unmaintained sections within the road sections adjacent to the candidate section of interest. Finally, parameter Tv represents a weighting coefficient corresponding to the traffic volume for the candidate section of interest.

[0080] In this way, by defining individual interval evaluation values, the larger the difference between the cost of autonomous driving and the cost of manual driving, the smaller the individual interval evaluation value. Therefore, road intervals that reduce the driver's workload more by maintaining map information are more likely to be selected as the target intervals for map information generation or updating. Thus, according to this variation, the map update interval determination device can determine the road intervals that will be the target intervals for map information generation or updating in a way that makes it easier for the driver to gain the advantage of autonomous driving control.

[0081] In addition, the evaluation value calculation unit 42 can also calculate the individual interval evaluation value Ei for each of the selected candidate intervals according to the following formula.

[0082] Ei=(Mc-Ac-Hc)*Tv

[0083] It should be noted that, as shown in the above formula, when the cost of autonomous driving is set to 0, the individual interval evaluation value Ei of this variant becomes a value calculated based on the cost of manual driving, switching cost, and traffic volume.

[0084] In this case, the higher the evaluation value, the greater the improvement in driver convenience obtained by generating or updating map information. Therefore, the determination unit 43 can set one of the predetermined end conditions to an evaluation value higher than the end reference value. Furthermore, when the end condition is met, the determination unit 43 determines each candidate interval corresponding to the maximum value among the evaluation values ​​calculated up to this point as the road interval to be used for generating or updating map information. In this modified example, it is preferable that the determination unit 43 replaces the road intervals used for calculating the evaluation value with those having higher evaluation values. Therefore, the determination unit 43 sets an exclusion probability for each candidate interval in a way that candidate intervals with smaller (individual interval evaluation value / map maintenance cost) values ​​are more likely to be excluded from the evaluation value calculation objects. Conversely, the determination unit 43 sets an addition probability for each unmaintained interval in a way that unmaintained intervals with larger (individual interval evaluation value / map maintenance cost) values ​​are more likely to be added as candidate intervals to the evaluation value calculation objects.

[0085] In this variation, the map update interval determination device can also determine the road intervals that become the objects of map information generation or updating in a way that makes it easier for the driver to gain the advantage of autonomous driving control.

[0086] Computer programs based on the above embodiments or variations, used to enable a computer to perform the processing of various parts executed by the processor 34 of the server 3, may also be recorded in a semiconductor memory device, a magnetic recording medium, or an optical recording medium and distributed.

[0087] As described above, those skilled in the art can make various modifications within the scope of this invention and in accordance with the manner in which it is implemented.

Claims

1. A map update interval determination device, wherein, The map update interval determination device has the following features: The storage unit stores map maintenance costs, map markers, and switching costs for each of the multiple road sections included in a predetermined area. The map maintenance cost is the cost required to generate or update map information for the road section to enable vehicles to drive autonomously. The map markers indicate whether the map information is usable. The switching cost represents the driver's workload when switching between autonomous driving and manual driving occurs in the road section and the road sections adjacent to the road section. The initial selection unit, referring to the map markers of each of the plurality of road sections, selects two or more road sections from the road sections from which the map information cannot be used, such that the sum of the map maintenance costs of the selected road sections is below the target maintenance cost upper limit. The evaluation value calculation unit calculates an evaluation value based on the switching cost of each of the two or more selected road sections, the evaluation value representing the degree of improvement in driver convenience obtained by generating or updating map information for the two or more selected road sections; and The determination unit replaces one or more of the selected road intervals with other unselected road intervals that cannot utilize the map information, and the evaluation value calculation unit calculates the evaluation value, determines whether the evaluation value meets a predetermined end condition, and when the predetermined end condition is met, determines each of the selected road intervals corresponding to the evaluation value that provides the highest improvement in driver convenience among the calculated evaluation values ​​as the road intervals that become the objects of generating or updating the map information.

2. The map update interval determination device according to claim 1, wherein, The storage unit also stores the traffic volume and autonomous driving cost for each of the multiple road sections. The autonomous driving cost represents the driver's workload when the vehicle uses the map information to drive autonomously in that road section. The evaluation value calculation unit calculates the sum of the values ​​obtained by weighting the sum of the autonomous driving cost and the switching cost of each of the two or more selected road sections using a weighting coefficient corresponding to the traffic volume of that road section, and uses this sum as the evaluation value.

3. The map update interval determination device according to claim 1, wherein, The storage unit also stores, for each of the plurality of road sections, the traffic volume of that road section, the autonomous driving cost representing the driver's workload when the vehicle uses the map information to drive autonomously in that road section, and the manual driving cost representing the driver's workload when the vehicle drives manually in that road section. The evaluation value calculation unit calculates, for each of the two or more selected road sections, the sum of the values ​​obtained by weighting the value obtained by subtracting the value of the manual driving cost from the value of the automatic driving cost plus the value of the switching cost using a weighting coefficient corresponding to the traffic volume of that road section, as the evaluation value.

4. The map update interval determination device according to any one of claims 1 to 3, wherein, When the determining unit replaces any one of the two or more selected road intervals with the other road intervals, the road interval with the larger value obtained by multiplying the individual interval evaluation value calculated for that road interval by the map maintenance cost is more likely to be replaced with the other road interval.

5. The map update interval determination device according to claim 1, wherein, The storage unit also stores the traffic volume for each of the plurality of road sections. The initial selection unit selects two or more road sections from the multiple road sections whose map information cannot be used, in descending order of the ratio of traffic volume to map maintenance cost.

6. A method for determining map update intervals, wherein, The methods for determining the map update interval include: Referring to map markers indicating whether map information for vehicles to drive autonomously within a predetermined area is available for each of the multiple road sections, two or more road sections are selected from those road sections where the map information is unavailable, in a manner that the total map maintenance cost required to generate or update the map information for the selected road sections is below a target maintenance cost upper limit. The switching cost is an evaluation value representing the improvement in driver convenience achieved by generating or updating map information for two or more selected road sections. This switching cost represents the driver's workload when switching between autonomous and manual driving occurs in each of the selected road sections and in adjacent road sections. Replace one or more of the selected road intervals with other unselected road intervals from the road intervals where the map information cannot be used, and calculate the evaluation value. Determine whether the evaluation value meets the predetermined termination condition. When the predetermined termination condition is met, each of the two or more selected road sections corresponding to the evaluation value that provides the highest improvement in driver convenience among the calculated evaluation values ​​is determined as the road section to be generated or updated in the map information.

7. A recording medium recording a computer program for determining map update intervals, wherein, The map update interval is determined by a computer program that causes the computer to perform the following actions: Referring to map markers indicating whether map information for vehicles to drive autonomously within a predetermined area is available for each of the multiple road sections, two or more road sections are selected from those road sections where the map information is unavailable, in a manner that the total map maintenance cost required to generate or update the map information for the selected road sections is below a target maintenance cost upper limit. The switching cost is an evaluation value representing the improvement in driver convenience achieved by generating or updating map information for two or more selected road sections. This switching cost represents the driver's workload when switching between autonomous and manual driving occurs in each of the selected road sections and in adjacent road sections. Replace one or more of the selected road intervals with other unselected road intervals from the road intervals where the map information cannot be used, and calculate the evaluation value. Determine whether the evaluation value meets the predetermined termination condition. When the predetermined termination condition is met, each of the two or more selected road sections corresponding to the evaluation value that provides the highest improvement in driver convenience among the calculated evaluation values ​​is determined as the road section to be generated or updated in the map information.

Images