JP7736602B2 - Root Calculator - Google Patents

Description

The present invention relates to a route calculation device that calculates a target route for mobility.

An example of a road surface condition estimation device that estimates road surface conditions is described in Japanese Patent Application Laid-Open No. 2020-13537. This road surface condition estimation device determines whether abnormal conditions are met based on specific behaviors that the vehicle is expected to exhibit if it encounters a road surface abnormality, based on vehicle behavior information, and estimates road surface conditions based on the determination results.

Japanese Patent Application Laid-Open No. 2020-13537

It is desirable that the target route traveled by mobility (especially autonomous mobility) be a safe route that avoids rough roads where the vehicle is likely to get stuck. Therefore, information about road surface conditions is important when selecting a route. As with the road surface condition estimation device described above, road surface conditions are generally estimated based on the detection results of various sensors installed on the mobility (here, the vehicle). However, sensors vary depending on the vehicle manufacturer and model, making it difficult to estimate road surface conditions in a unified and statistical manner. Furthermore, in places such as mines where there are unpaved roads where large heavy machinery and passenger vehicles (e.g., pickup trucks) travel together, it is difficult for road surface condition estimation devices, which are designed for paved roads and passenger vehicles, to accurately estimate road surface conditions.

The object of the present invention is to provide a route calculation device that can calculate a target route based on uniformly estimated road surface conditions, even in locations where there are unpaved roads on which multiple different vehicle types travel.

The route calculation device of the present invention includes a map data storage unit that stores map data divided into a plurality of areas, a driving history storage unit that stores driving histories of mobility vehicles for each area based on location information of the mobility vehicles, an impact storage unit that stores, for each of a plurality of mobility vehicles, the impact of changes in road surface conditions caused by the driving of one mobility vehicle on the driving of the other mobility vehicles, a driving risk calculation unit that calculates a target driving risk value that represents the driving risk when a target mobility vehicle, which is a mobility vehicle that is the target of calculation of a target route, travels through each of the areas based on the driving history and the impact on the target mobility vehicle, and a target route calculation unit that calculates the target route for the target mobility vehicle based on the map data, location information of the target mobility vehicle, destination information, and the target driving risk value. Note that the route calculation device may be configured to perform machine learning using the calculation results for the target route and the driving results of the target mobility vehicle as input data, and update the impact of each mobility vehicle.

According to the present invention, the driving risk (target driving risk value) for a target mobility is calculated for each area on map data based on the mobility's driving history and the degree of impact on the target mobility. A target route for the target mobility is calculated based on this target driving risk value. The target driving risk value is calculated based on the driving history for each area and the degree of impact for each mobility, regardless of the sensors of each mobility. In other words, according to the present invention, the target driving risk value for each area is calculated uniformly for each mobility, so road surface conditions can be estimated without being affected by differences in sensor configuration. According to the present invention, a target route can be calculated based on uniformly estimated road surface conditions, even in places where there are unpaved roads on which multiple different vehicle types travel.

FIG. 1 is a configuration diagram of a root calculation device according to an embodiment of the present invention. FIG. 2 is a conceptual diagram illustrating an example of map data according to the embodiment. 4 is a time chart for explaining a target driving risk value according to the present embodiment. FIG. 3 is a conceptual diagram illustrating an example (first calculation example) of map data according to the embodiment. FIG. 10 is a conceptual diagram illustrating an example of map data (second calculation example) according to the present embodiment. FIG. 10 is a conceptual diagram for explaining a target driving risk value for each driving direction (angle) according to the present embodiment. FIG. 10 is a conceptual diagram illustrating an example of map data (third calculation example) according to the present embodiment. FIG. 10 is a conceptual diagram for explaining a crossing risk value according to the present embodiment. FIG. 10 is a conceptual diagram for explaining a crossing risk value according to the present embodiment. FIG. 10 is a conceptual diagram for explaining a crossing risk value according to the present embodiment. FIG. 10 is a conceptual diagram for explaining a crossing risk value according to the present embodiment. FIG. 10 is a conceptual diagram for explaining an example of resetting the target driving risk value according to the present embodiment. FIG. 10 is a conceptual diagram for explaining an example of resetting the target driving risk value according to the present embodiment. 10 is a flowchart illustrating an example of a flow of calculation of a target route according to the present embodiment.

The following describes in detail, with reference to the drawings, a root calculation device 1, which is one embodiment of the present invention. In addition to the following example, the present invention can be implemented in a variety of forms, including various modifications and improvements based on the knowledge of those skilled in the art.

The route calculation device 1 of this embodiment is a computer or electronic control unit (ECU) equipped with at least one processor and at least one memory. Various programs and data are stored in the memory. One or more processors read and execute programs from the memory to perform various calculations. The route calculation device 1 is configured to be able to communicate with each mobility via a communications network. The route calculation device 1 is incorporated into, for example, a central control system, and transmits information on the determined target route to the mobility for which the target route is to be calculated. For example, if the target mobility is an autonomous vehicle, the target mobility performs autonomous driving based on the received target route.

As shown in FIG. 1, the route calculation device 1 has the following functions: a map data storage unit 11, a driving history storage unit 12, an impact storage unit 13, a driving risk calculation unit 14, and a target route calculation unit 15. The route calculation device 1 realizes each function through the operation of a processor, memory, etc.

The map data storage unit 11 stores map data divided into multiple areas. The route calculation device 1 recognizes multiple pre-defined areas on the map data. The areas are divided, for example, by straight lines in a grid pattern at a predetermined interval, and can also be called grids or cells. Areas may be set for the entire map data, or may be set only for parts of the map data where roads (including paved and unpaved roads) exist. The characteristics of the areas (size, shape, number, etc.) can be set appropriately for the map data.

The route calculation device 1 determines whether a road is paved or unpaved based on, for example, road data included in map data, accumulated information on estimated road surface conditions based on the mobility's travel, or information on road surface conditions estimated based on the mobility's behavior while traveling. Furthermore, the route calculation device 1 may be configured to recognize roads as, for example, "unpaved roads only (e.g., route calculation in a mine)" or "paved roads only (e.g., route calculation on public roads)" depending on the user's settings.

The driving history storage unit 12 stores the driving history of each mobility for each area based on the mobility's location information. The mobility's location information can be obtained, for example, from GPS (Global Positioning System) data. For example, each mobility is equipped with a GPS receiver, determines its own location based on the GPS data, and transmits its own location data to the route calculation device 1 via wireless communication. Based on the GPS data of each mobility, the route calculation device 1 can determine and store which mobility traveled where and when. The driving history storage unit 12 can be said to store driving history for each mobility (for each type). For example, the type of traveling mobility, the driving area, the driving direction, and time series data for these can be read from the driving history.

For travel by a vehicle larger than a single area (such as a large dump truck with a tire diameter of 2m or more), the travel history storage unit 12 updates the travel history for areas where the tires of the vehicle are estimated to have passed, based on the vehicle's position and direction of travel. The travel history storage unit 12 stores the travel history by treating the positions of the vehicle's two tires (the positions of the left and right wheels) as the vehicle's position.

The route calculation device 1 stores information such as the size and number of wheels of a large dump truck, calculates the tire position based on this information, GPS data, and driving direction, and updates the driving history of the area the tire has passed through. If a large dump truck has two front wheels and two rear wheels, the driving of the large dump truck along one route will create two tire trajectories, i.e., two driving routes, and the driving history of the areas passed through on the two driving routes will be updated (see the wheel ruts in Figure 5).

The impact memory unit 13 stores, for each of multiple mobility vehicles, the degree of impact that the driving of one mobility vehicle has on the driving of other mobility vehicles. For example, if the road is unpaved, ruts of dirt or mud are likely to form on the road when a mobility vehicle is driving. Also, for example, if the road is paved and snow is present, ruts of snow may be formed on the road when a mobility vehicle is driving. The impact memory unit 13 stores the degree to which changes in road surface conditions (ruts, unevenness, etc.) caused by the driving of one mobility vehicle affect the driving of other mobility vehicles as the impact on the driving of the other mobility vehicles.

The impact memory unit 13 of this embodiment stores the impact for each type (kind) of mobility. That is, for each of multiple types (kinds) of mobility, the impact memory unit 13 of this embodiment stores the impact that the driving of one type of mobility has on the driving of other types of mobility. The impact memory unit 13 may also store the impact that the driving of one type of mobility has on the driving of the same type of mobility. The impact memory unit 13 stores, for example, pickup trucks (small freight vehicles), medium-sized trucks, and large dump trucks (large heavy equipment) as types of mobility. The impact memory unit 13 stores, for example, the impact of a pickup truck's driving on the driving of medium-sized trucks and the impact of a large dump truck's driving on the driving of pickup trucks and the impact of a medium-sized truck's driving on the driving of pickup trucks and the impact of a large dump truck's driving on the driving of pickup trucks and the impact of a large dump truck's driving on the driving of pickup trucks and the impact of a large dump truck's driving on the driving of pickup trucks and the impact of a medium-sized ... driving of medium-sized trucks. The impact memory unit 13 also stores the impact on unpaved roads and on paved roads with snow.

As an example, if there is a driving history of a medium-sized truck driving through the first area, the impact level when a pickup truck drives through the first area is set relatively high, and the impact level when a large dump truck drives through the first area is set relatively low (or 0). If the mobility to be driven is smaller than the mobility in the driving history of the area, the impact level on driving will be high. On the other hand, if the mobility to be driven has a size (tires, etc.) equal to or larger than the mobility in the driving history of the area, the impact level on driving will be low.

The impact level can be thought of as the degree of driving difficulty that takes into account the degree of roughness of the road surface and the relative mobility. For example, there is a high possibility that large ruts will be formed on an unpaved road after a large vehicle has driven on it, and the next time a small vehicle drives on that unpaved road, the road surface (large ruts) is likely to have a high impact on the vehicle's driving. Furthermore, when a vehicle of the same type or larger size that has previously driven on that unpaved road drives on it, the impact of the road surface (large ruts) on the vehicle's driving is likely to be lower than when a small vehicle drives on that unpaved road, since it is a vehicle of the same size or larger than the previous time. The impact level is set based on this thinking.

(Calculation of driving risk)
The driving risk calculation unit 14 calculates a target driving risk value, which is the driving risk when a target mobility, which is a mobility to be calculated for the target route, travels through each area, based on the driving history and the impact on the target mobility. The target driving risk value increases as the impact on the target mobility in the travel area increases. For example, the target driving risk value increases the more times a mobility larger than the target mobility has traveled, i.e., the more times a mobility with a high impact has traveled. Furthermore, for example, if a mobility of the same size or smaller than the target mobility has traveled through a given area, the target driving risk value for the given area for the target mobility is reduced or reset.

As an example of the calculation of the target driving risk value, with reference to Figures 2 and 3, the calculation of the target driving risk value for area X1 when the target mobility is a pickup truck will be described. As shown in Figure 2, multiple areas divided into a grid are set on the map data. In this example, all roads are unpaved, and multiple areas are set on unpaved roads in the map data so as to divide the unpaved roads into multiple areas. The target driving risk value is calculated for each area.

As shown in Figure 3, the target driving risk value for area X1 for pickup trucks (hereinafter simply referred to as the "target driving risk value" in the explanation of this example) changes depending on the driving history of other mobility vehicles. The target driving risk value from time t0 to t1 is "0." The state of area X1 from time t0 to t1 is either that nothing has yet been driven there, or that other pickup trucks of the same model (kind) have only driven there.

At time t1, a medium-sized truck passes through area X1, causing the target driving risk value to increase according to the impact. In this example, the impact of a medium-sized truck's passage on the driving of a pickup truck is "1 (first predetermined value)." Therefore, the target driving risk value from time t1 to t2 is "1." Similarly, the impact is added to the target driving risk value each time a medium-sized truck passes through area X1, as at times t2, t3, and t4, so the target driving risk value becomes "2" from time t2 to t3, "3" from time t3 to t4, and "4" from time t4 to t5. At time t4, the cumulative number of passes of the medium-sized truck through area X1 is four, and the impact of each pass is "1," so the target driving risk value at time t4 is a cumulative value of "4." In this way, the target driving risk value is calculated based on driving history and impact.

At time t5, a pickup truck passes through area X1, causing the target driving risk value to become "0." In this example, the target driving risk value is reset (or reduced) in areas where a mobility of the same type or smaller than the target mobility passes through. In this example, because a mobility of the same type or smaller than the target mobility was able to travel without getting stuck, the driving risk is deemed low (none), and the target driving risk value is reset. At time t6, a medium-sized truck passes through area X1, causing the target driving risk value to become "1."

At time t7, the target driving risk value becomes "3" due to the passage of a large dump truck through area X1. In the settings of this embodiment, the impact of the passage of a large dump truck on the driving of a pickup truck is "2 (second predetermined value)" (second predetermined value > first predetermined value). Therefore, the target driving risk value is 3, which is the sum of 1 and 2. At time t8, the target driving risk value becomes "5" due to the passage of a large dump truck through area X1. Note that each value (e.g., first predetermined value, second predetermined value, etc.), rules for reducing and resetting the target driving risk value, and the amount of reduction when reducing the target driving risk value can be changed as appropriate.

In this way, the driving risk calculation unit 14 increases the target driving risk value for areas where a mobility larger than the target mobility, i.e., a mobility with a greater impact on the road surface, has traveled, based on the degree of impact. The driving risk calculation unit 14 also reduces or resets the target driving risk value in areas where a mobility of the same type as the target mobility, a mobility smaller than the target mobility, or a mobility with lower off-road capability than the target mobility has traveled.

The driving risk calculation unit 14 may create and store a driving risk map that calculates the target driving risk value for each area for each type of mobility. In this case, the driving risk calculation unit 14 may calculate and store, for example, a first driving risk map that represents the target driving risk value for each area when a pickup truck is the target mobility, and a second driving risk map that represents the target driving risk value for each area when a medium-sized truck is the target mobility. The driving risk calculation unit 14 can also calculate and store a third driving risk map that represents the target driving risk value for each area when a large dump truck is the target mobility. The driving risk calculation unit 14 updates each driving risk map based on the driving history and impact level of each mobility.

(Calculation of target route)
The target route calculation unit 15 calculates a target route for the target mobility based on map data, location information of the target mobility, destination information, and the target driving risk value. That is, when calculating the target route, the target route calculation unit 15 takes into account the driving risk (target driving risk value) of each area for the target mobility. For example, the target route calculation unit 15 may calculate the target route so as not to travel (avoid) areas where the target driving risk value is equal to or greater than a predetermined driving threshold.

The target route is the target route for the target mobility to reach its destination, and includes information about the road to be traveled and the travel position within the road (target trajectory). For example, when traveling on a wide road, there may be multiple routes on a single road. For example, in mines, there may be unpaved roads with widths of 20 meters or more.

The target route is calculated according to predetermined conditions. The predetermined conditions may be, for example, a route with the shortest distance, a route incorporating a clothoid curve, or a route with the shortest driving time. Furthermore, as described above, for example, the predetermined conditions may include a "route that does not travel through areas where the target driving risk value is equal to or greater than a predetermined driving threshold."

As an example, the target route calculation unit 15 is configured to be able to execute tentative route calculation processing, route risk calculation processing, and target route determination processing. The tentative route calculation processing is processing that calculates a tentative route, which is a tentative target route, based on map data, location information of the target mobility, and destination information. The route risk calculation processing is processing that calculates a route risk value for the tentative route based on the target driving risk values of the areas that the target mobility will pass through if it travels along the tentative route. The target route determination processing is processing that determines the target route based on the route risk values. If the route risk value is less than a predetermined route threshold in the target route determination processing, the target route calculation unit 15 determines the tentative route as the target route, and if the route risk value is equal to or greater than the route threshold, executes the tentative route calculation processing again. Note that multiple target routes may be determined, in which case the user may be allowed to select a target route.

(First Example of Calculation of Route Risk Value)
In the route risk calculation process, a route risk value is calculated based on the sum of the target driving risk values of the areas that the target mobility will pass through if it travels along the tentative route. For example, as shown in Figure 4, the areas that the target mobility will travel along tentative route R1 on the diagram are areas X1, X2, X3, X4, X5, X6, and X7 (hereinafter abbreviated as "X1 to X7"). For tentative route R1, assume that the starting point (current location) is X1 and the ending point (destination) is X7. The sum of the target driving risk values of areas X1 to X7, i.e., the route risk value of tentative route R1, is, for example, "20."

If the specified conditions for target route determination include "select as the target route a tentative route whose route risk value is less than a specified route threshold," in other words, "do not select as the target route a tentative route whose route risk value is equal to or greater than a specified route threshold," the route risk value of the tentative route is compared with the route threshold in the target route determination process. For example, if the route threshold is "20," the route risk value of tentative route R1 is "20," so tentative route R1 is not selected as the target route. In this case, the target route calculation unit 15 calculates another tentative route again in the tentative route calculation process. In the tentative route calculation process, the tentative route is calculated based on specified conditions (e.g., the shortest route), just like in the first calculation.

As shown in Figure 4, in the recalculated tentative route R2, some of the areas traveled by the target mobility are different from those in the tentative route R1. That is, in tentative route R2, the target mobility travels through areas X11 and X12 instead of area X1. If the sum of the target driving risk value of area X11 and the target driving risk value of area X12 is lower than the target driving risk value of area X1, the route risk value of tentative route R2 will be less than the route threshold. In this case, the target route calculation unit 15 determines tentative route R2 as the target route in the target route determination process.

(Second Example of Calculation of Route Risk Value)
In calculating the route risk value, in addition to the calculation of the first calculation example (calculation of the sum of the target driving risk values), the following calculation may be performed: That is, in the route risk calculation process, the target route calculation unit 15 increases the route risk value based on the difference between the target driving risk values of adjacent areas on the tentative route.

For example, on tentative route R1 in Figure 4, if the difference between the target driving risk value of area X1 and the target driving risk value of area X2 is equal to or greater than a predetermined difference threshold, a predetermined crossing risk value is added to the total target driving risk value when calculating the route risk value. A situation in which there is a large difference in the target driving risk values of adjacent areas is considered to be relatively likely to be a situation in which one area has ruts and the other does not. This situation can be estimated as a situation in which one area has ruts and the other does not have ruts, i.e., a situation in which a rut is being crossed (crossed).

A situation in which a mobility vehicle crosses a rut is considered to pose a higher risk of the mobility vehicle getting stuck (for example, getting stuck, rolling over, or falling off) than a situation in which the mobility vehicle is driving along the rut. Based on this idea, a "crossing risk value" is added to the route risk value depending on the difference in the target driving risk values between adjacent areas. The crossing risk value is set to a higher value when the target mobility vehicle drives across a rut than when it drives along the rut. As an example, if the difference in the target driving risk values is equal to or greater than the difference threshold, the vehicle is considered to be driving across a rut, and the route risk value is increased by a predetermined crossing risk value. In the second calculation example, if the difference in the target driving risk values is less than the difference threshold, the route risk value is not increased.

As shown in Figure 5, the target driving risk value is relatively high in areas X2 and X4 on the tentative route, where a large dump truck traveled. The driving areas on the tentative route are areas X1, X2, X3, X4, and X5 on the drawing. If the difference in the target driving risk values between areas X1 and X2, the difference in the target driving risk values between areas X2 and X3, the difference in the target driving risk values between areas X3 and X4, and the difference in the target driving risk values between areas X4 and X5 are each equal to or greater than the difference threshold, the route risk value of the tentative route on the drawing is the sum of the target driving risk values of areas X1 to X5 plus the crossing risk values for four times. If the route risk value of this tentative route is equal to or greater than the route threshold, the target route calculation unit 15 recalculates the tentative route. On the other hand, if the route risk value of this tentative route is less than the route threshold, the target route calculation unit 15 determines this tentative route as the target route.

The crossing risk value may be set in stages. For example, if the crossing risk value when the difference between adjacent target driving risk values (hereinafter also referred to as the "adjacent difference") is equal to or greater than the difference threshold is referred to as the first crossing risk value, and the crossing risk value when the adjacent difference is less than the difference threshold and each target driving risk value is equal to or greater than a predetermined value is referred to as the second crossing risk value, the first crossing risk value and the second crossing risk value may be set to different values. Comparing the risk of mobility crossing a rut with the risk of mobility driving along a rut, the former is considered to be higher than the latter. According to this concept, the first crossing risk value is set to a value higher than the second crossing risk value (first crossing risk value > second crossing risk value).

For example, a tentative route passing through areas X1 and X6 in Figure 5 passes through an area with a relatively high target driving risk value, but the adjacent difference is small, so the tentative route can be estimated to be a route that follows ruts, and the second crossing risk value (a medium value) is applied as an additional value. On the other hand, for example, a tentative route passing through areas X1 and X2, the adjacent difference is large, so the tentative route can be estimated to be a route that crosses ruts, and the first crossing risk value (a high value) is applied as an additional value. Examples of crossing risk values will be described further below.

(Third Example of Calculation of Route Risk Value)
The target driving risk value for each area may be calculated to take on different values depending on the driving direction of the target mobility. For example, as shown in FIG. 6 , the target driving risk value may be calculated for each angle (at predetermined angle intervals) between a predetermined reference direction and the driving direction of the target mobility. In other words, for one area, a target driving risk value is calculated according to the driving direction of the target mobility. Note that in FIG. 6 , the angle of the driving direction relative to the reference direction (hereinafter also referred to as the "driving angle") is expressed as 0 to 360 degrees (deg), but it may also be expressed as, for example, 0 to 180 degrees (deg). In this case, the target driving risk value when traveling from one end to the other end on a line segment is the same as the target driving risk value when traveling from the other end to the one end. For example, in a north-south direction, the direction from south to north and the direction from north to south have the same target driving risk value. Furthermore, in the example of FIG. 6 , the target driving risk value is set every 20 degrees of driving angle, but the set interval can be set arbitrarily.

As shown in Figure 7, if the reference direction is from north to south in the north-south direction and a large dump truck travels from north to south in the north-south direction, the driving angle of the large dump truck in the driving history is 0 degrees. The driving history storage unit 12 stores, as driving history, information about the traveled mobility (e.g., model information) as well as the driving direction (driving angle) of the mobility. The driving direction may be defined, for example, as the direction from the start point to the end point of the driving history (driving trajectory) within the area.

According to tentative route R3 in Figure 7, the angle between the driving direction of the large dump truck and the driving direction of the target mobility in area X20 is approximately 90 degrees (or 270 degrees). On the other hand, according to tentative route R4, the angle between the driving direction of the large dump truck and the driving direction of the target mobility in area X20 is approximately 0 degrees (parallel).

An example of how target driving risk values are calculated according to driving direction is described below. If a large dump truck drives once in a certain area and the driving angle of the large dump truck is 0 degrees, the target driving risk value will be maximum when the driving angle of the target mobility is 90 degrees and 270 degrees, and minimum when the angle is 0 degrees and 180 degrees (see, for example, Figure 6). This is based on the idea that, assuming that ruts are formed in a north-south direction, driving the mobility in an east-west direction (90 degrees or 270 degrees), i.e., driving across the ruts, is relatively high risk, while driving the mobility in a north-south direction (0 degrees or 180 degrees), i.e., driving along the ruts, is relatively low risk. Therefore, in the example of Figure 7, the target driving risk value for area X20 on tentative route R3 will be maximum, and the target driving risk value for area X20 on tentative route R4 will be minimum.

As an example of calculation in this case, the driving risk calculation unit 14 calculates the target driving risk value when the driving angle is 90 degrees and 270 degrees based on the value obtained by multiplying the influence degree by a first coefficient (e.g., first coefficient > 1). The driving risk calculation unit 14 also calculates the target driving risk value when the driving angle is 0 degrees and 180 degrees based on the value obtained by multiplying the influence degree by a second coefficient smaller than the first coefficient (e.g., second coefficient < 1). The driving risk calculation unit 14 also calculates the target driving risk value when the driving angle is between 0 degrees and 90 degrees and between 180 degrees and 270 degrees so that it gradually increases as the driving angle increases. The driving risk calculation unit 14 calculates the target driving risk value when the driving angle is between 90 degrees and 180 degrees and between 270 degrees and 360 degrees so that it gradually decreases as the driving angle increases. The distribution of the target driving risk values for each driving angle calculated in this way is, for example, as shown in Figure 6. The driving risk calculation unit 14 may calculate the target driving risk value based on a predetermined relational expression (e.g., a function) between the driving angle and the target driving risk value.

The driving angles at which the target driving risk value reaches its maximum or minimum value are not limited to the above and can be set as appropriate, and can be changed depending on how the risk of the mobility getting stuck, falling, or rolling over is considered. For example, if the mobility is prone to rolling over when its approach angle to a rut is 45 degrees, then the calculation rules can be set so that the driving angle corresponding to an approach angle of 45 degrees reaches the maximum value of the target driving risk value. For example, if it is desired to increase the risk value when the approach angle is 0 degrees, the calculation rules can be set using a similar approach.

Furthermore, the target driving risk value may be calculated by adding a value based on the degree of influence only to driving angles where the target mobility vehicle has a high risk of rollover, etc., relative to the driving of other mobility vehicles. For example, in an area where a large dump truck has driven at a driving angle of 0 degrees, the target driving risk value for mobility vehicles smaller than large may be increased only for a predetermined driving angle range estimated to be high risk. In this case, the target driving risk value is increased only for some driving angles due to the driving of the mobility vehicle.

The target driving risk value is accumulated for each driving angle (driving direction). For example, if a large dump truck drives twice through an area at the same driving angle (driving direction), the target driving risk value for each driving angle in that area will be twice that of the first drive. Furthermore, if the driving direction for the second drive differs by 90 degrees from the driving direction for the first drive, the target driving risk value after the second drive may be the same for each driving angle, for example, by adding the maximum value for the second drive to the minimum value for the first drive. Note that, depending on the settings, the driving risk calculation unit 14 may determine that only the ruts from the second drive remain, assuming that the second drive overwrites the first drive. In this case, for example, the target driving risk value for the first drive may be set to a constant value according to the degree of impact regardless of the driving direction, and a value that takes into account the driving direction of only the second drive history may be added to the target driving risk value.

The driving risk calculation unit 14 may also be configured so that each time a mobility travels through an area, the target driving risk values for all driving angles in that area are increased by a fixed value (a value based on the degree of impact), and each time a mobility of the same type or smaller than the target mobility travels through the area, the target driving risk values for the corresponding driving direction (driving angle) in that area are reduced or reset by a fixed value.

The target route calculation unit 15 calculates the sum of the target driving risk values (values corresponding to the driving direction) for all areas through which the tentative route passes, with the extension direction of the tentative route being the driving direction of the target mobility, as in the above calculation example. Furthermore, as in the second calculation example, the target route calculation unit 15 may add a crossing risk value to this sum. As in the above calculation example, the target route calculation unit 15 determines the target route based on the calculated route risk value.

In this way, in the third calculation example, the driving risk calculation unit 14 calculates the target driving risk value for each driving direction of the target mobility in each area based on the driving direction of the mobility in the driving history.The target route calculation unit 15 then calculates the target route of the target mobility based further on the driving direction of the target mobility.

(Examples of cross-sectional risk values)
Below, an example of calculating the crossing risk value for two adjacent areas on the tentative route will be described. In the route risk calculation process, the target route calculation unit 15 sets a predetermined value that can be compared with the target driving risk value as a guide to the roughness of the road surface. If the target driving risk value is higher than the predetermined value, it can be determined that the road surface is rough, and if the target driving risk value is equal to or lower than the predetermined value, it can be determined that the road surface has almost no impact on driving. This predetermined value can be set to any value, but in this example, the predetermined value is set to the initial value of the target driving risk value for each area (the target driving risk value when the driving history is 0).

The driving risk calculation unit 14 increases the target driving risk value when a larger vehicle than the target mobility vehicle is traveling, and decreases the target driving risk value when a vehicle of the same type as the target mobility vehicle or smaller than the target mobility vehicle is traveling. Therefore, the target driving risk value may be less than the initial value (predetermined value). Below, we will explain an example of calculation that takes into account the risk of crossing a rut for two adjacent areas on a tentative route.

(Third crossing risk value)
As shown in Figure 8, if the driving angle of the target mobility is 90 degrees with respect to two adjacent areas X91 and X92 on the tentative route, the target route calculation unit 15 refers to the 90-degree target driving risk value of each area X91 and X92 in the route risk calculation process.

In the example of Figure 8, the target driving risk value for both areas X91 and X92 at 90 degrees is less than a predetermined value. In this way, when the target driving risk values of two adjacent areas X91 and X92 are less than a predetermined value (which may be set to be equal to or less than the predetermined value), the target route calculation unit 15 adds the third traversal risk value to the route risk value in the route risk calculation process, regardless of the difference between the two target driving risk values (adjacent difference). The third traversal risk value is set to a value lower than the second traversal risk value (first traversal risk value > second traversal risk value > third traversal risk value ≧ 0).

When a target mobility vehicle travels through two adjacent areas where the target driving risk value is less than a predetermined value, it can be assumed that "a mobility vehicle of the same model or smaller has already traveled there at the same driving angle without rolling over, etc.", and the route risk value will be small. The route risk value (X1, X2, 90deg) of a tentative route with area X91 as the start point and area X92 as the end point is, for example, the sum of the target driving risk value of area X91 and the target driving risk value of area X92, plus the third traversing risk value. Note that the third traversing risk value may, for example, be 0.

In addition, the target route calculation unit 15 may refer to the driving history, and if a larger vehicle than the target vehicle has recently driven through area X91 and/or area X92, it may add a crossing risk value different from the third crossing risk value to the route risk value, even if the target driving risk value in both areas X91 and X92 is less than a predetermined value (which may be set to be equal to or less than the predetermined value).

(Second crossing risk value)
As shown in Figure 9, if the driving angle of the target mobility is 90 degrees with respect to two adjacent areas X93 and X94 on the tentative route, the target route calculation unit 15 refers to the 90-degree target driving risk value of each area X93 and X94 in the route risk calculation process.

In the example of Figure 9, the 90-degree target driving risk value of area X93 is less than a predetermined value, and the 90-degree target driving risk value of area X94 is higher than the predetermined value. Furthermore, the difference between the two target driving risk values (adjacent difference) is less than the difference threshold. In this case, the target route calculation unit 15 adds the second traversal risk value to the route risk value in the route risk calculation process (first traversal risk value > second traversal risk value > third traversal risk value ≥ 0). In the settings of this embodiment, even if the 90-degree target driving risk values of both areas X93 and X94 are higher than the predetermined value, if the adjacent difference is less than the difference threshold, the value added to the route risk value is the second traversal risk value. In other words, in this embodiment, if the target driving risk value of at least one of the adjacent areas is higher than the predetermined value and the adjacent difference is less than the difference threshold, the second traversal risk value is added to the route risk value.

In the situation shown in Figure 9, a larger vehicle than the target vehicle was traveling in at least one area, but because the adjacent difference is small, it can be inferred that "the tentative route is unlikely to cross a large rut." Therefore, the second traversal risk value, a medium risk value higher than the first traversal risk value, is added to the route risk value. The route risk value (X93, X94, 90deg) for a tentative route starting from area X93 and ending at area X94 is, for example, the sum of the target driving risk value for area X93 and the target driving risk value for area X94, plus the second traversal risk value. Note that if the target driving risk value for 90 degrees in both areas is higher than the specified value and the adjacent difference is 0 or close to 0, it can be inferred that a 90-degree driving angle follows the rut, so a traversal risk value smaller than the second traversal risk value may be added to the route risk value.

(First crossing risk value)
As shown in Figure 10, if the driving angle of the target mobility is 90 degrees with respect to two adjacent areas X95 and X96 on the tentative route, the target route calculation unit 15 refers to the 90-degree target driving risk value of each area X95 and X96 in the route risk calculation process.

In the example of Figure 10, the 90-degree target driving risk value of area X95 is less than a predetermined value, and the 90-degree target driving risk value of area X96 is higher than the predetermined value. Furthermore, the difference between the two target driving risk values (adjacent difference) is greater than or equal to the difference threshold. In this case, the target route calculation unit 15 adds the first traversal risk value to the route risk value in the route risk calculation process. In this embodiment, if the target driving risk value of at least one of the adjacent areas is higher than the predetermined value and the adjacent difference is greater than or equal to the difference threshold, the first traversal risk value is added to the route risk value (first traversal risk value > second traversal risk value > third traversal risk value ≧ 0).

In a situation like that shown in Figure 10, a larger vehicle than the target vehicle was traveling in at least one area, and because the difference between adjacent areas is large, it can be inferred that "the tentative route is likely to cross a large rut." Therefore, the first traversal risk value, which is a higher risk value than the second traversal risk value, is added to the route risk value. The route risk value (X95, X96, 90deg) of a tentative route starting from area X95 and ending at area X96 is, for example, the sum of the target driving risk value of area X95 and the target driving risk value of area X96, plus the first traversal risk value.

In this way, in the route risk calculation process, the target route calculation unit 15 increases the route risk value based on the difference in the target driving risk values of adjacent areas (hereinafter also referred to as adjacent areas) on the tentative route. Furthermore, if the target driving risk value of at least one of two adjacent areas on the tentative route is higher than a predetermined value and the difference between the two adjacent areas is equal to or greater than a predetermined difference threshold, the target route calculation unit 15 adds a traversing risk value (first traversing risk value) to the route risk value. Furthermore, if the target driving risk value of at least one of two adjacent areas is higher than a predetermined value and the difference between the two adjacent areas is less than the difference threshold, the target route calculation unit 15 adds a medium traversing risk value (second traversing risk value) that is smaller than the traversing risk value (first traversing risk value) to the route risk value. Furthermore, if the target driving risk values of both adjacent areas are less than or equal to the predetermined value, the target route calculation unit 15 adds a small traversing risk value (third traversing risk value) that is smaller than the medium traversing risk value (second traversing risk value) to the route risk value.

(Other examples of calculating crossing risk values)
As shown in Figure 11, if the driving angle of the target mobility is 90 degrees with respect to two adjacent areas X97 and X98 on the tentative route, the target route calculation unit 15 refers to the 90-degree target driving risk value of each area X97 and X98 in the route risk calculation process.

In this example, the initial value and the predetermined value are set to different values (initial value < predetermined value). Furthermore, the calculation rule for the target driving risk value is set so that the target driving risk value is increased only for certain driving angles (here, driving angles between 0 and 20 degrees) due to mobility driving. For example, if a large dump truck passes through area X97 at a driving angle of 0 degrees, the target driving risk value for pickup trucks or medium-sized trucks is increased only in the area around 0 degrees (for example, 0 to 20 degrees) of area X97. Furthermore, when a mobility of the same model or smaller than the target mobility passes through, a relatively small value is added to the target driving risk value for the corresponding driving angle. In other words, in this example, a driving angle for which the target driving risk value is the initial value means that the mobility has never driven at that driving angle.

In the example of Figure 11, the target driving risk value for 90 degrees is set to the initial value for both areas X97 and X98. Therefore, in this example, the mobility has not traveled even once at a driving angle of 90 degrees in areas X97 and X98. In such a case, the target route calculation unit 15 estimates the degree of roughness of the road surface in areas X97 and X98 by referring to the target driving risk values for other driving angles. As an example, the target route calculation unit 15 calculates the average value (or, for example, the maximum value) of the target driving risk values for all driving angles in area X97 as the target driving risk value for area X97. Similarly, the target route calculation unit 15 calculates the average value (or, for example, the maximum value) of the target driving risk values for all driving angles in area X98 as the target driving risk value for area X98.

In the example of Figure 11, the average value of the target driving risk value in area X97 is below a predetermined value (or may be "less than the predetermined value"). On the other hand, the average value of the target driving risk value in area X98 is higher than the predetermined value. Furthermore, the adjacent difference, i.e., the difference between the average values of the two target driving risk values, is less than the difference threshold. This corresponds to a case where the target driving risk value of at least one of the adjacent areas is higher than the predetermined value and the adjacent difference is less than the difference threshold, and a second traversal risk value (risk value: medium) is added to the route risk value.

If the adjacent difference is equal to or greater than the difference threshold, the first crossing risk value (high risk) is added to the route risk value. If both average values are equal to or less than a predetermined value, the third risk value (low risk) is added to the route risk value. If both average values are at their initial values, the crossing risk value may be set not to be added to the route risk value.

(Resetting the target driving risk value)
When the driving risk calculation unit 14 acquires information that a road has been leveled, it resets (initializes) the target driving risk values of all mobility vehicles in the leveled area to their initial values, as shown in FIG. 12 . Also, as shown in FIG. 13 , if there is a driving angle (90 degrees in this example) at which no mobility vehicle has been traveling for a predetermined period of time or more, the driving risk calculation unit 14 may reset the target driving risk value for that driving angle for each mobility vehicle (each driving risk map). The predetermined period of time is set, for example, to a fairly long period of time. In this way, when the driving risk calculation unit 14 acquires information that an area has been leveled, it resets the target driving risk value for the leveled area. Also, the driving risk calculation unit may reset the target driving risk value for an area at which no mobility vehicle has been traveling for a predetermined period of time or more.

(An example of the overall calculation flow)
An example of the overall flow of calculation of a target route performed by the route calculation device 1 will be described with reference to Fig. 14. First, the route calculation device 1 collects mobility travel information (travel history), such as the type of mobility traveled, the area traveled, the travel direction, and the number of travels (S1). The route calculation device 1 also checks whether there is a reason for resetting, such as the acquisition of leveled information (S2).

The route calculation device 1 updates a driving risk map showing the target driving risk value for each area created by mobility type (category) based on the driving information and the presence or absence of a reset reason (S3). For example, if there is no leveling information and a large dump truck has driven once through an area, the target driving risk value for the corresponding area on the first driving risk map corresponding to pickup trucks and the target driving risk value for the corresponding area on the second driving risk map corresponding to medium-sized trucks are updated.

In a tentative route calculation process, the route calculation device 1 generates a tentative route for map data based on the mobility's current location information and destination information (S4). Next, in a route risk calculation process, the route calculation device 1 references a driving risk map corresponding to the target mobility (S5) and calculates a route risk value for the tentative route (S6). For example, in calculating the route risk value of the tentative route, the route calculation device 1 calculates the route risk value for each two adjacent areas and finally calculates the sum of these values. For example, in the case of a tentative route such as that shown in FIG. 5, the tentative route is divided into a sub-tentative route starting from area X1 and ending at area X2, a sub-tentative route starting from area X2 and ending at area X3, a sub-tentative route starting from area X3 and ending at area X4, and a sub-tentative route starting from area X4 and ending at area X5. Then, the route risk value (total target driving risk + crossing risk value) of each small tentative route is calculated, and the route risk values of all small tentative routes are summed to calculate the route risk value of the entire tentative route.

Next, the route calculation device 1 determines whether the route risk value of the tentative route is less than a predetermined route threshold (S7). If the route risk value is less than the route threshold (S7: Yes), the route calculation device 1 determines the tentative route as the target route (S8). On the other hand, if the route risk value is equal to or greater than the route threshold (S7: No), the route calculation device 1 executes the tentative route calculation process again to generate another tentative route (S4). The series of calculation processes ends when the target route is determined. Note that if a route threshold (small route threshold) for the route risk value of the small tentative route has been set, the route calculation device 1 may be configured to recalculate the tentative route if the route risk value of the small tentative route is equal to or greater than the small route threshold. In other words, the length (size) of the tentative route whose route risk value is to be calculated and the corresponding route threshold can be set as appropriate.

(Effects of this embodiment)
According to this embodiment, the risk of driving for the target mobility (target driving risk value) is calculated for each area on the map data based on the mobility's driving history and the impact on the target mobility. A target route for the target mobility is calculated based on this target driving risk value. The target driving risk value is calculated based on the driving history for each area and the impact for each mobility, without depending on the sensors of each mobility. In other words, according to this embodiment, the target driving risk value for each area is calculated uniformly, so that road surface conditions are estimated without being affected by differences in sensor configuration. According to the present invention, even in a location where there are unpaved roads on which multiple different vehicle types travel, a target route can be calculated based on the estimated road surface conditions uniformly for each mobility.

Furthermore, according to this embodiment, a target driving risk value is calculated according to the driving direction of the target mobility, enabling risk calculation that is more in line with road surface conditions. Furthermore, according to this embodiment, a target route is determined based on the route risk value of the tentative route, making it possible to select a low-risk route as the target route by taking into account the risk of the entire driving area. Furthermore, according to this embodiment, a route risk value is calculated taking into account the crossing risk values of two adjacent areas, improving the accuracy of avoiding risks such as rollover due to crossing ruts in the target route calculation results. Furthermore, according to this embodiment, the target driving risk value and route risk value are calculated taking into account ground leveling, etc., improving the accuracy of risk calculation.

(others)
The present invention is not limited to the above embodiment. For example, the target route calculation unit 15 may calculate multiple tentative routes and select, as the target route, a tentative route among the multiple tentative routes whose route risk value is less than the route threshold. In this case, if multiple target routes are selected, the user may be prompted to select one target route. Alternatively, the target route calculation unit 15 may calculate multiple tentative routes and determine, as the target route, a tentative route among the multiple tentative routes whose route risk value is less than the route threshold and is the smallest.

Furthermore, the route risk value does not have to be a value based on the sum of the target driving risk values in multiple areas traveled on the tentative route, but may be a value based on, for example, the average value of the target driving risk values. Furthermore, the route risk value may be the sum of the crossing risk values (values based on adjacent differences) corresponding to the tentative route, rather than the sum of the target driving risk values. The sum of the crossing risk values alone can distinguish between large and small route risk values for the tentative route. Depending on the settings, the route risk value can be calculated, for example, as "route risk value = sum of target driving risk values," "route risk value = sum of target driving risk values + sum of crossing risk values," "route risk value = sum of crossing risk values," or "route risk value = average value of target driving risk values," "route risk value = average value of target driving risk values + average value of crossing risk values," etc.

The influence memory unit 13 may also store the influence for each driving angle. Furthermore, the mobility's position information may be estimated based on the detection results of a surroundings monitoring device, such as a LiDAR (LiDAR) and a camera, installed on the mobility. The mobility's position information may also be GSNN information. Furthermore, while this embodiment describes an example in which the road is unpaved, the present invention is also applicable to roads that are paved (e.g., paved and snowy). If the road of the tentative route is paved, the target route calculation unit 15 may, for example, multiply the calculated route risk value by a correction coefficient less than 1. Furthermore, for the same mobility's travel, the influence of unpaved roads may be set to a value greater than the influence of paved roads.

Furthermore, for example, the impact level and crossing risk value may be appropriately updated through machine learning using AI technology. For example, if a mobility travels a calculated target route and becomes stuck or rolls over, the route calculation device 1 may determine that the increase in the route risk value based on the driving history is insufficient based on the driving results, and learn the impact level and crossing risk value of each mobility. The route calculation device 1 may perform machine learning using the calculation results for the target route (e.g., the target driving risk value and/or route risk value) and the driving results of the target mobility (e.g., whether or not the vehicle becomes stuck, etc.) as input data, and may be configured to update the impact level and/or crossing risk value of each mobility, for example. In other words, the route calculation device 1 uses, for example, the route risk value of the target route, whether or not the vehicle becomes stuck, etc., the target driving risk value of the driving area, and the driving history as input data, and learns what values the impact level and/or crossing risk value of each mobility should be set to in order to calculate a target route in which the vehicle will not become stuck, etc. The route calculation device 1 may be configured to use, for example, the route risk value of the target route, whether or not the vehicle is stuck, the target driving risk value of the driving area, and driving history as input data, and to review the settings of each risk value or find patterns in the risk of the vehicle being stuck or the like.

1... Route calculation device, 11... Map data storage unit, 12... Driving history storage unit, 13... Impact storage unit, 14... Driving risk calculation unit, 15... Target route calculation unit.

Claims (17)

a map data storage unit that stores map data divided into a plurality of areas;
a travel history storage unit that stores the travel history of the mobility for each area based on the location information of the mobility;
an impact storage unit that stores, for each of a plurality of mobility vehicles, an impact level of a change in road surface conditions caused by the travel of one mobility vehicle on the travel of the other mobility vehicles;
a travel risk calculation unit that calculates a target travel risk value, which is a travel risk when a target mobility, which is a mobility to be calculated for a target route, travels through each of the areas, based on the travel history and the influence on the target mobility;
a target route calculation unit that calculates the target route of the target mobility based on the map data, location information of the target mobility, destination information, and the target driving risk value;
A route calculation device comprising: The driving risk calculation unit increases the target driving risk value for the area in which a mobility that has a greater impact on a road surface than the target mobility has traveled, based on the impact degree.
The route calculation device according to claim 1 . The driving risk calculation unit reduces or resets the target driving risk value in the area where a mobility of the same type as the target mobility, a mobility smaller than the target mobility, or a mobility with lower running capability than the target mobility has traveled.
3. The route calculation device according to claim 1 or 2. The driving risk calculation unit calculates the target driving risk value for each driving direction of the target mobility in each of the areas based on the driving direction of the mobility in the driving history,
The target route calculation unit further calculates the target route of the target mobility based on a traveling direction of the target mobility.
The route calculation device according to any one of claims 1 to 3. When the target driving risk value corresponding to the driving direction of the target mobility in the area is an initial value, the target route calculation unit calculates the target driving risk value of the area based on the target driving risk values of all driving directions in the area.
The route calculation device according to claim 4. the target route calculation unit calculates the target route so that the target mobility does not travel through the area where the target traveling risk value is equal to or greater than a predetermined traveling threshold value;
The route calculation device according to any one of claims 1 to 5. The target route calculation unit
a tentative route calculation process that calculates a tentative route, which is a tentative target route, based on the map data, the position information of the target mobility, and the information of the destination;
a route risk calculation process that calculates a route risk value for the tentative route based on the target traveling risk value of the area through which the target mobility passes when traveling along the tentative route;
a target route determination process for determining the target route based on the route risk value;
configured to run
The route calculation device according to any one of claims 1 to 6. The target route calculation unit calculates the route risk value based on the sum of the target traveling risk values of the areas that the target mobility will pass through when traveling along the tentative route in the route risk calculation process.
The route calculation device according to claim 7 . the target route calculation unit, in the target route determination process, determines the tentative route as the target route when the route risk value is less than a predetermined route threshold, and executes the tentative route calculation process again when the route risk value is equal to or greater than the route threshold.
9. The route calculation device according to claim 7 or 8. the target route calculation unit increases the route risk value based on a difference between the target traveling risk values of the adjacent areas on the tentative route in the route risk calculation process;
The route calculation device according to any one of claims 7 to 9. the target route calculation unit adds a traversing risk value to the route risk value when the target traveling risk value of at least one of two adjacent areas that are adjacent to each other on the tentative route is higher than a predetermined value and the difference between the two adjacent areas is equal to or greater than a predetermined difference threshold.
The route calculation device according to claim 10. the target route calculation unit adds a medium crossing risk value that is smaller than the crossing risk value to the route risk value when the target traveling risk value of at least one of the two adjacent areas is higher than the predetermined value and the difference between the two adjacent areas is less than the difference threshold.
The route calculation device according to claim 11. the target route calculation unit adds a small crossing risk value that is smaller than the medium crossing risk value to the route risk value when the target driving risk values of the two adjacent areas are both less than the predetermined value or equal to or less than the predetermined value;
The route calculation device according to claim 12. When the travel risk calculation unit acquires information that the area has been leveled, the travel risk calculation unit resets the target travel risk value of the area that has been leveled.
The route calculation device according to any one of claims 1 to 13. The travel risk calculation unit resets the target travel risk value for the area in which mobility has not traveled for a predetermined time or more.
The route calculation device according to any one of claims 1 to 14. The driving risk calculation unit creates a driving risk map in which the target driving risk value of each area is represented for each mobility or each type of mobility.
The route calculation device according to any one of claims 1 to 15. The machine learning is performed using the calculation result regarding the target route and the driving result of the target mobility as input data, and the influence degree of each mobility is updated.
The route calculation device according to any one of claims 1 to 16.