Energy consumption prediction method, energy consumption prediction system, and information processing apparatus

By utilizing a grid map with linked grid points and directions, the method addresses inefficiencies in data collection for power consumption prediction, achieving accurate energy consumption forecasts along vehicle routes.

WO2026239605A1PCT designated stage Publication Date: 2026-11-19NISSAN MOTOR CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2025/017908
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-05-16
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Existing power consumption prediction methods for vehicles require extensive data collection from start to end points, leading to a significant amount of data needed for learning, which is inefficient and impractical.

Method used

A method that collects travel data from a start to an end point, using a grid map configured with grid points linked to coordinates and directions, allowing for learning and prediction of energy consumption based on this data, without continuous data collection along the entire route.

Benefits of technology

Improves the efficiency of data collection and achieves highly accurate energy consumption predictions by learning energy consumption rates for each grid point and direction, enabling precise predictions along planned routes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2025017908_19112026_PF_FP_ABST
    Figure JP2025017908_19112026_PF_FP_ABST
Patent Text Reader

Abstract

This energy consumption prediction method is a method for predicting the energy consumption of a vehicle (20) in an energy consumption prediction system (1) provided with the vehicle (20) and a server (10). The method comprises: collecting travel data when the vehicle (20) travels from a travel start point to a travel end point, the travel data including a travel start point (Ss), a travel end point (Ds), a travel distance, and energy consumption; and performing grid map learning processing for predicting the energy consumption of the vehicle (20) on the basis of the travel data. The grid map is configured by associating each of a plurality of grid points (G1) correlated with the coordinates of map information with a plurality of directions and energy consumption when the vehicle (20) travels from the grid point (G1) toward each direction.
Need to check novelty before this filing date? Find Prior Art

Description

Energy consumption prediction method, energy consumption prediction system, and information processing apparatus

[0001] The present invention relates to an energy consumption prediction method, an energy consumption prediction system, and an information processing apparatus.

[0002] Patent Document 1 discloses a power consumption prediction method for an electric vehicle. The power consumption prediction method includes first and second steps. The first step acquires target vehicle information indicating information related to the usage state of a first vehicle. The second step corrects, for each link connecting virtually set nodes on the planned travel route of the first vehicle, actual power consumption values collected from a plurality of second vehicles using the target vehicle information, thereby calculating a predicted power consumption value for each link of the first vehicle.

[0003] Japanese Unexamined Patent Application Publication No. 2019-86430

[0004] However, the technique described in Patent Document 1 collects actual power consumption data for each fixed section on the travel route along which the vehicle travels. In order to obtain power consumption data for a certain travel route, data must be continuously collected from the travel start point to the travel end point. For this reason, there is a problem that the amount of data required for learning power consumption becomes enormous.

[0005] The present invention has been made in view of the above problems, and an object thereof is to realize highly accurate prediction of energy consumption while improving the efficiency of collecting learning data.

[0006] An energy consumption prediction method according to one aspect of the present invention is a method for predicting the energy consumption of a vehicle in a system including the vehicle and a server. This method collects travel data when the vehicle travels from a travel start point to a travel end point, the travel data including the travel start point, the travel end point, a travel distance, and energy consumption, and performs learning processing of a grid map for predicting the energy consumption of the vehicle based on the travel data. The grid map is configured such that each of a plurality of grid points associated with coordinates of map information is linked with a plurality of azimuth directions and energy consumption when the vehicle travels from the grid point in each of the azimuth directions.

[0007] According to the present invention, it is possible to improve the efficiency of collecting training data while achieving highly accurate predictions of energy consumption.

[0008] Figure 1 is a block diagram illustrating the energy consumption prediction system according to this embodiment. Figure 2 is a diagram illustrating a grid map. Figure 3 is a flowchart illustrating the learning process related to the grid map. Figure 4 is a diagram illustrating an overview of the learning process. Figure 5 is a diagram illustrating an overview of the learning process. Figure 6 is a diagram illustrating an overview of the learning process. Figure 7 is a diagram illustrating an overview of the learning process. Figure 8 is a flowchart illustrating the prediction and update processes using the grid map. Figure 9 is a diagram illustrating an overview of the prediction process. Figure 10 is a diagram illustrating an overview of the prediction process. Figure 11 is a diagram illustrating an overview of the update process. Figure 12 is a diagram illustrating an example of the update process.

[0009] The energy consumption prediction method, energy consumption prediction system, and information processing device according to this embodiment will be described below with reference to the drawings.

[0010] As shown in Figure 1, the energy consumption prediction system 1 consists of a server 10 and a vehicle 20 that runs using energy such as fossil fuels or electricity. The energy consumption prediction system 1 predicts the amount of energy consumed by the vehicle 20. In this embodiment, the energy consumption rate, which is the amount of energy consumed by the vehicle 20 per unit distance, is used as an indicator of the energy consumption of the vehicle 20. The vehicle 20 is, for example, an electric vehicle that runs using electricity, in which case the energy consumption rate is the electricity consumption rate (Wh / km).

[0011] Server 10 is an information processing device that acquires driving data from vehicle 20 and performs learning processing to learn about a grid map based on the driving data of vehicle 20. The grid map is a data set in which energy consumption rates are linked to coordinates and directions on map information, and shows the energy consumption rate of vehicle 20 heading in a specific direction at a specific coordinate. Details of the grid map will be described later. In addition, server 10 acquires route data from vehicle 20, including the planned route that vehicle 20 will travel. Based on the route data and the grid map, server 10 performs prediction processing to predict the energy consumption rate of vehicle 20 along its planned route.

[0012] The server 10 includes a controller 11, a storage device 12, and a communication device 13.

[0013] The controller 11 is composed of a computer equipped with a CPU, memory, input / output interfaces, etc. The CPU reads various computer programs stored in memory and executes various instructions contained in the programs. By executing programs, the controller 11 realizes multiple functions provided by the server 10.

[0014] In this embodiment, the controller 11 performs learning and prediction processing using a grid map. As shown in Figure 2, the grid map is configured such that each of the multiple grid points is associated with multiple directions and the energy consumption rate when the vehicle 20 travels in each direction from the grid point. A grid point is a point that corresponds to the coordinates of the map information. Specifically, a grid point is a location identified by the coordinates (latitude, longitude) of the intersection of each grid line drawn in the latitude and longitude directions. As shown in Figure 2, each grid point (coordinate) is assigned a location ID to identify the grid point. The grid map contains multiple grid points that exist within a certain area (for example, Japan). The direction is the direction in which the vehicle 20 travels. In the example shown in Figure 2, four directions are shown: east, west, south, and north, and each direction is assigned a direction ID to identify that direction. Note that the directions that make up the grid map are not limited to four directions, but can be multiple directions such as eight directions or sixteen directions.

[0015] The storage device 12 stores a grid map. A grid map is prepared for each vehicle type. The communication device 13 communicates with the vehicle 20. The controller 11 can communicate with the vehicle 20 via the communication device 13. The communication device 13 may be a device equipped with mobile communication functions such as 4G / LTE or 5G, or a device equipped with Wi-Fi communication functions.

[0016] Vehicle 20, in one aspect, collects driving data necessary for learning processing. In another aspect, vehicle 20 requests server 10 to predict the energy consumption rate along the planned driving route. Vehicle 20 is equipped with a controller 21, a communication device 22, a display device 23, and a storage device 24.

[0017] The controller 21 is a microcomputer mainly composed of a CPU, memory, and input / output interface. The controller 21 collects driving data and transmits it to the server 10. The driving data is data from when the vehicle 20 travels from the starting point to the ending point. The driving data includes the vehicle type, the starting point, the ending point, the distance traveled from the starting point to the ending point, and the energy consumption of the vehicle 20 during the journey from the starting point to the ending point. The distance traveled is not the straight-line distance from the starting point to the ending point, but the actual distance traveled by the vehicle 20. The controller 21 also calculates the planned route from the starting point to the destination based on the information of the starting point and destination. The controller 21 requests the server 10 to predict the energy consumption rate along the planned route by transmitting route data including the vehicle type, planned route, and current remaining energy amount to the server 10.

[0018] The communication device 22 communicates with the server 10. The controller 21 can communicate with the server 10 via the communication device 22. The communication device 22 may be a device equipped with mobile communication functions such as 4G / LTE or 5G, or a device equipped with Wi-Fi communication functions. The display device 23 is a device that displays information to the user of the vehicle 20, and has an integrated input device such as a touch panel. The storage device 24 stores map information necessary for creating the planned driving route. The storage device 24 also records driving data collected by the controller 21.

[0019] Some or all of the controller 21, communication device 22, display device 23, and storage device 24 of the vehicle 20 may be replaced by a user terminal such as a smartphone used by the user.

[0020] The vehicle 20 that collects driving data and the vehicle 20 that is the target of prediction processing may be the same vehicle or different vehicles. The vehicle 20 that collects driving data may be more than one vehicle. Similarly, the vehicle 20 that is the target of prediction processing may be more than one vehicle. Hereinafter, the vehicle 20 that collects driving data will be referred to as the collection vehicle 20, and the vehicle 20 that requests prediction processing from the server 10 will be referred to as the prediction vehicle 20, as necessary.

[0021] Referring to Figure 3, the learning process according to this embodiment will be described. The process shown in the flowchart of Figure 3 is executed by the controller 11 of the server 10.

[0022] First, the controller 11 acquires driving data from the collection vehicle 20 (S10).

[0023] The controller 11 calculates the learning energy consumption rate based on the distance traveled and the energy consumption included in the driving data (S11). Specifically, the controller 11 calculates the learning energy consumption rate by dividing the energy consumption by the distance traveled.

[0024] Next, the controller 11 limits the learning range in the grid map (S12). Specifically, the controller 11 extracts the grid map corresponding to the vehicle type in the driving data based on the vehicle type included in the driving data. Next, as shown in Figure 4, the controller 11 identifies the direction (orientation) of the vector Vc from the starting point Ss to the ending point Ds based on the driving start point Ss and the driving end point Ds included in the driving data. Then, based on the orientation of the vector Vc, the controller 11 identifies the learning target orientation, which is the orientation to be learned from among the four orientations provided by the grid map. For example, the controller 11 sets the range of positive and negative reference angles centered on the orientation of the vector Vc as the extraction range, and identifies the orientations included in the extraction range among the four orientations provided by the grid map as the learning target orientations. For example, east is defined as 0 degrees, north as 90 degrees, west as 180 degrees, and south as 270 degrees, and the target range is set to plus or minus 45 degrees centered on the orientation of the vector Vc. If the vector's orientation is 30 degrees, the controller 11 identifies east as the direction to be learned. Figure 5 shows the actual route Rt traveled by the collection vehicle 20.

[0025] Next, the controller 11 extracts multiple grid points from the grid map that correspond to the learning target direction. Figure 5 conceptually shows multiple grid points G1 extracted corresponding to the learning target direction, east. In Figure 5, the X axis corresponds to the longitude of the grid point coordinates, and the Y axis corresponds to the latitude of the grid point coordinates (the same applies to Figures 6 and 7). The numbers shown for each grid point G1 are the energy consumption rates associated with that grid point G1. If the grid map has not been trained, an initial value is set for the energy consumption rate of each grid point G1. In the example shown in Figure 5, the energy consumption rate is 150 [Wh / km].

[0026] The controller 11 limits the learning range Ob at multiple grid points corresponding to the learning target direction (S12). Specifically, as shown in Figure 6, the controller 11 assigns the learning range Ob at multiple grid points G1 corresponding to the learning target direction. The learning range Ob has a rectangular shape and is set so as to be tangent to the starting point Ss and ending point Ds of travel, that is, so that the vector Vc is the diagonal. The controller 11 extracts the grid points G1 included in the learning range Ob as the learning target grid points G2.

[0027] As shown in Figure 3, the controller 11 calculates the learning rate α and the energy consumption rate after learning (S14).

[0028] The learning rate α is a parameter that weights the learning of the grid map, and is calculated based on the angular difference θ between the orientation of vector Vc and the orientation to be learned. For example, the learning rate α is given by Equation 1, where F is the maximum value of the learning rate α and is set in advance. (Equation 1) α = F × cosθ

[0029] As can be seen from Equation 1, the smaller the angle difference θ between the direction of vector Vc and the direction of the target to be learned, the greater the weight given to learning. For example, if the angle difference θ is 30 degrees and the maximum learning rate F is 0.2, the learning rate α becomes 0.173 (α = 0.2 × cos 30°).

[0030] The energy consumption rate Rec after learning is given by Equation 2. Here, Rce1 is the current energy consumption rate associated with the learning target grid point G2. Rce2 is the energy consumption rate for learning. (Equation 2) Rec = (1 - α) × Rec1 + α × Rce2

[0031] For example, as shown in Figures 5 and 6, let's assume that the energy consumption rate Rec1 associated with the learning target grid point G2 is 150 [Wh / km]. Also, let's assume that the learning energy consumption rate Rec2 is 100 [Wh / km]. The energy consumption rate Rec after learning will be 141 [Wh / km] (Rce = (1 - 0.173) × 150 + 0.173 × 100).

[0032] As shown in Figure 7, the controller 11 updates the energy consumption rate of the target grid point G2 by associating the learned energy consumption rate Rec with the target grid point G2 (S15).

[0033] Furthermore, if the learning range Ob includes multiple learning target grid points G2, the controller 11 performs steps S14 and S15 for each of the multiple learning target grid points G2. Also, if multiple learning target directions are identified, the controller 11 performs steps S12, S14 and S15 for each of the multiple learning target directions.

[0034] Through this series of processes, the controller 11 learns a grid map based on the driving data collected by the collection vehicles 20. By learning the grid map using driving data obtained from multiple collection vehicles 20, it is possible to learn a wide range of energy consumption rates associated with each grid point G1 and each direction. This improves the accuracy of predicting energy consumption rates using the grid map. Furthermore, since a grid map is prepared for each vehicle type, it is possible to appropriately learn the energy consumption rates specific to each vehicle type.

[0035] Next, with reference to Figure 8, the prediction and update processes according to this embodiment will be described. The processes shown in the flowchart of Figure 8 are executed by the controller 11 of the server 10, triggered by the acquisition of route data. Route data is transmitted from the prediction vehicle 20 to the server 10 when the prediction vehicle 20 requests the server 10 to predict the energy consumption rate. The route data includes the type of vehicle of the prediction vehicle 20 and the planned route. As shown in Figure 9, the planned route is a series of routes from the departure point Sp to the destination Dp that the prediction vehicle 20 will travel. The route data also includes the current remaining energy amount in the prediction vehicle 20. For example, the remaining energy amount is shown as a percentage. Note that the route data only needs to include at least the departure point Sp, destination Dp, vehicle type, and current remaining energy amount. The server 10 may acquire the planned route by calculating the planned route itself.

[0036] First, the controller 11 sets parameters a and b (S20). These parameters a and b determine the resolution when predicting the energy consumption rate. Increasing parameters a and b increases the prediction resolution, while decreasing parameters a and b decreases the prediction resolution. For example, parameters a and b are stored in the controller 11 as pre-set fixed values. However, parameters a and b may also be values ​​specified by the user through operation of the display device 23. Here, parameter b is a larger value than parameter a.

[0037] As shown in Figure 8, the controller 11 extracts a grid map corresponding to the vehicle type included in the driving data, and then performs a prediction process to predict the trend of energy consumption along the planned driving route based on the extracted grid map and the planned driving route (S21-S23).

[0038] Specifically, the controller 11 divides the planned route into intervals a, for example, every 1 km, and assigns multiple division points on the planned route (S21). In the example shown in Figure 9, the planned route is divided into six parts, and five points Vp1, Vp2, Vp3, Vp4, and Vp5 are assigned between the departure point Sp and the destination Dp. The controller 11 recognizes the departure point Sp and the five points Vp1, Vp2, Vp3, Vp4, and Vp5 as multiple prediction target points. The multiple prediction target points are defined by coordinates on the map information. Then, the controller 11 identifies the prediction target direction for each of the multiple prediction target points (S22). The prediction target direction is the direction corresponding to the direction of travel of the vehicle 20 when passing through the prediction target point, and is identified by referring to the planned route. For example, the controller 11 identifies the tangential direction to the planned route as the prediction target direction.

[0039] The controller 11 performs a mapping process to associate the energy consumption rate with each of the multiple prediction target points (S22). Specifically, the controller 11 identifies the prediction target grid point that corresponds to the prediction target point from among the multiple grid points included in the grid map. The prediction target grid point that corresponds to the prediction target point means not only the prediction target grid point that is at coordinates that exactly match the prediction target point, but also the prediction target grid point that is in the closest positional relationship to the prediction target point. The controller 11 refers to the grid map and identifies the energy consumption rate associated with the prediction target grid point and the prediction target direction. The controller 11 identifies the identified energy consumption rate as the energy consumption rate at the prediction target point.

[0040] The controller 11 identifies the energy consumption rate for each of the multiple prediction target points and then performs an energy forecast based on the identification results (S22). First, the controller 11 predicts the trend of the energy consumption rate along the planned route. Specifically, the controller 11 arranges the multiple prediction target points in the order in which the vehicle 20 will travel along the planned route. The controller 11 plots multiple energy consumption rates according to the order in which the multiple prediction target points are arranged. By connecting the multiple energy consumption rates with lines, the controller 11 predicts the trend of the energy consumption rate along the planned route, as shown in the upper part of Figure 10. Next, the controller 11 predicts the trend of the remaining energy amount when the predicted vehicle 20 travels along the planned route, based on the current remaining energy amount, interval a, and the predicted trend of the energy consumption rate. Then, as shown in the lower part of Figure 10, the controller 11 predicts the trend of the remaining energy amount (%).

[0041] In addition, the controller 11 transmits prediction data including the transition prediction of the energy consumption rate and the transition prediction of the remaining energy amount to the vehicle 20, in order to display information to the user of the vehicle 20. The controller 21 of the vehicle 20 controls the display device 23 based on the received prediction data, and causes the display device 23 to display the transition prediction of the energy consumption rate and the transition prediction of the remaining energy amount. Note that the method by which the controller 21 notifies the user of information is not limited to notification by display, and may be notification by voice (the same applies hereinafter).

[0042] Next, the controller 21 of the predicted vehicle 20 aggregates the measured values of energy consumption when traveling in a travel section of distance b. Then, when the travel in the travel section is completed, the controller 21 transmits the aggregated measured value data of the energy consumption (measured value data) to the server 10. The controller 11 performs an update process of updating the transition prediction of the remaining energy amount based on the aggregation result of the measured values every time the vehicle travels a travel section (S24-S30).

[0043] As shown in FIG. 8, the controller 11 determines whether or not the predicted vehicle 20 has traveled a distance b, for example, 10 km (S24). The controller 11 determines whether or not the predicted vehicle 20 has traveled the distance b based on whether or not the measured value data has been acquired from the predicted vehicle 20. When the predicted vehicle 20 has traveled the distance b (S24: YES), the controller 11 performs the process of step S25. When the predicted vehicle 20 has not traveled the distance b (S24: NO), the controller 11 performs the process of step S31 described later.

[0044] The controller 11 aggregates the energy consumption rate based on the energy consumption data (S25). Specifically, the controller 11 aggregates the measured value of the energy consumption rate by the predicted vehicle 20 by dividing the energy consumption aggregated by the predicted vehicle 20 by the distance b.

[0045] As shown in the upper part of Figure 11, the controller 11 compares the predicted energy consumption rate with the measured energy consumption rate (S26). In the upper part of Figure 11, circles indicate the predicted energy consumption rate for each travel section, and squares indicate the measured energy consumption rate for each travel section. The controller 11 compares the predicted energy consumption rate with the measured energy consumption rate for the currently traveled travel section. Figure 12 shows the changes in the predicted and measured energy consumption rates at the end of the fifth travel section.

[0046] The controller 11 learns a prediction correction value to correct the discrepancy between the predicted value and the measured value based on the comparison results of energy consumption rates (S27). For example, the controller 11 learns the prediction correction value Va according to equation 3 shown below. Here, β is the learning rate, a parameter that weights the learning of the prediction correction value. β is a fixed value, for example, 0.3. Vap is the prediction correction value calculated in the previous driving section. ΔE is the value obtained by subtracting the predicted energy consumption rate from the measured energy consumption rate in the currently driven driving section (difference) (see Figure 12). (Equation 3) Va = (1 - β) × Vap + β × ΔE

[0047] The middle section of Figure 11 shows the trend of the predicted correction value Va calculated for each driving section. When the measured value is smaller than the predicted value, the predicted correction value Va will be a negative value, and when the measured value is larger than the predicted value, the predicted correction value Va will be a positive value (see Figure 12).

[0048] The controller 11 updates the transition prediction of the remaining energy amount based on the predicted correction value Va (S28). In the lower part of FIG. 11, Rea represents the measured value of the remaining energy amount, and Rer represents the transition prediction of the remaining energy amount. Reb represents the transition prediction of the remaining energy amount updated by the predicted correction value Va. The point Cp is the current location of the predicted vehicle 20. The controller 11 corrects the transition prediction of the remaining energy amount from the current location Cp to the destination Dp (the transition prediction located on the right side of point C) such that the remaining energy amount from the current location Cp to the destination Dp is adjusted by the predicted correction value Va for each travel section. Then, the controller 11 updates the transition prediction of the remaining energy amount by connecting the corrected transition prediction Reb of the remaining energy amount to the measured value Rea of the remaining energy amount.

[0049] The controller 11 transmits updated prediction data including the updated transition prediction of the remaining energy amount to the vehicle 20 in order to display information to the user of the vehicle 20 (S28). The controller 21 of the vehicle 20 controls the display device 23 based on the received updated prediction data, and displays the updated transition prediction of the remaining energy amount on the display device 23.

[0050] The controller 11 predicts the remaining energy amount A (see FIG. 11) when the predicted vehicle 20 arrives at the destination Dp, based on the updated transition prediction of the remaining energy amount (S29). In addition, the controller 11 transmits information on the remaining energy amount A at the destination Dp to the vehicle 20 in order to display information to the user of the vehicle 20 (S29). The controller 21 of the vehicle 20 controls the display device 23 based on the received updated prediction data, and displays the remaining energy amount A at the destination Dp on the display device 23.

[0051] Meanwhile, if the predicted remaining energy level reaches a predetermined lower limit B before the predicted vehicle 20 reaches its destination Dp, the controller 11 identifies the point on the planned route where the predicted remaining energy level reaches the lower limit B (S29). The controller 11 then transmits warning data to the vehicle 20, including information about the point where the predicted remaining energy level reaches the lower limit B, in order to display the information to the user of the vehicle 20 (S29). The controller 21 of the vehicle 20 controls the display device 23 based on the received warning data to display a warning and also displays the point where the predicted remaining energy level reaches the lower limit B.

[0052] The controller 11 updates the travel section (S30). Upon updating the travel section, the controller 11 recognizes the new section for the update process and resets the distance counter, which is used for comparison with distance b in step S24, to zero.

[0053] The controller 11 determines whether the predicted vehicle 20 has arrived at the destination Dp (S31). The controller 11 can determine whether the predicted vehicle 20 has arrived at the destination Dp by communicating with the predicted vehicle 20. If the predicted vehicle 20 has arrived at the destination Dp (S31: YES), the controller 11 terminates this process. If the predicted vehicle 20 has not arrived at the destination Dp (S31: NO), the controller 11 repeats the process in step S24.

[0054] Thus, in this embodiment, the energy consumption prediction method includes collecting driving data including a starting point, an ending point, driving distance, and energy consumption, and performing a grid map learning process for predicting the energy consumption of the predicted vehicle 20 based on the driving data. The grid map is configured such that each of the multiple grid points G1 associated with the coordinates of the map information is linked to multiple directions and the energy consumption when the predicted vehicle 20 travels in each direction from the grid point G1.

[0055] This method allows for learning the energy consumption rate based on the start and end points of a journey, without continuously collecting data between the start and end points. Furthermore, the energy consumption rate can be learned for each direction at each grid point G1. This improves the efficiency of data collection for training while achieving accurate predictions of energy consumption.

[0056] Furthermore, grid maps may be created for each distance traveled. For example, grid maps may be created for groups with short distances, medium distances, and long distances. Grid maps using short distances can appropriately learn energy consumption due to changes in terrain. Grid maps using long distances, on the other hand, can learn energy consumption without being affected by changes in terrain. This can improve prediction accuracy.

[0057] In this embodiment, the learning process includes calculating a learning energy consumption rate based on the distance traveled and energy consumption included in the travel data, and identifying a target direction to be learned from among multiple directions based on the direction of the vector Vc from the starting point Ss to the ending point Ds of the journey. The learning process also includes extracting a plurality of grid points G1 associated with the target direction to be learned, assigning a learning range Ob to encompass the starting point Ss and the ending point Ds of the journey at the extracted plurality of grid points G1, extracting a target grid point G2 which is a grid point included in the learning range Ob, and updating the energy consumption rate associated with the target grid point G2 based on the learning energy consumption rate.

[0058] This method allows for the appropriate learning of energy consumption rates for each grid point G1 and each direction. This enables effective learning necessary for predicting energy consumption.

[0059] In this embodiment, the energy consumption prediction method further includes acquiring route data including the planned route of the predicted vehicle 20 from the departure point Sp to the destination Dp, and performing a prediction process to predict the trend of energy consumption along the planned route based on the route data and a grid map.

[0060] This method allows for accurate prediction of energy consumption trends along a planned route by using route data and grid maps.

[0061] In this embodiment, the prediction process includes assigning a plurality of prediction target points on the planned travel route, identifying a prediction target direction which corresponds to the direction of travel of the predicted vehicle 20 for each of the plurality of prediction target points, and performing a mapping process to associate the energy consumption rate for each of the plurality of prediction target points. The mapping process includes identifying a prediction target grid point which corresponds to the prediction target point from among a plurality of grid points included in the grid map, and identifying the energy consumption rate associated with the prediction target direction at the prediction target grid point as the energy consumption rate at the prediction target point.

[0062] According to this method, the grid points and directions to be predicted are identified in accordance with the movement of the predicted vehicle 20 along the planned route. This makes it possible to accurately predict the changes in energy consumption rate along the planned route.

[0063] In this embodiment, the prediction process includes predicting the change in the amount of remaining energy when the predicted vehicle 20 travels along the planned route, based on the prediction of the change in the energy consumption rate along the planned route.

[0064] This method allows for accurate prediction of changes in remaining energy levels by referring to forecasts of energy consumption trends.

[0065] In this embodiment, the energy consumption prediction method further includes notifying the user of the results of the prediction process.

[0066] This method allows the user to be notified of the predicted change in remaining energy and the predicted change in energy consumption rate along the planned route, which is the basis for the prediction, before the predicted vehicle 20 starts traveling.

[0067] In this embodiment, the energy consumption prediction method further includes aggregating the measured energy consumption rate for each predetermined driving section when the prediction vehicle 20 actually travels along the planned route, and performing an update process to update the predicted change in remaining energy amount based on the aggregated measured values ​​each time the vehicle travels through a driving section.

[0068] This method allows for updating the prediction of residual energy trends using a grid map based on actual measurements. This enables the identification of more accurate residual energy levels.

[0069] In this embodiment, the update process includes calculating a prediction correction value Va to correct the discrepancy between the predicted change and the measured value, and updating the predicted change in the amount of remaining energy from the current location Cp to the destination Dp of the predicted vehicle 20 based on the prediction correction value Va.

[0070] This method allows for updating the predicted change in remaining energy by utilizing the predicted correction value Va and the actual energy consumption trend of the predicted vehicle 20. This improves the accuracy of the predicted change in remaining energy.

[0071] In this embodiment, the energy consumption prediction method further includes notifying the user of the updated predicted trend of the remaining energy amount. This method allows the user to understand the updated predicted trend of the remaining energy amount. As a result, the user can accurately recognize the remaining energy amount of the predicted vehicle 20.

[0072] In this embodiment, the correction process includes predicting the amount of remaining energy when the predicted vehicle 20 arrives at the destination Dp, based on the updated prediction of the change in the amount of remaining energy. This method makes it possible to accurately predict the amount of remaining energy when the predicted vehicle 20 arrives at the destination Dp.

[0073] In this embodiment, the energy consumption prediction method further includes informing the user of the remaining energy amount when the prediction vehicle 20 arrives at the destination Dp. With this method, the user can accurately recognize the remaining energy amount when the prediction vehicle 20 arrives at the destination Dp.

[0074] In this embodiment, if the predicted remaining energy level reaches the lower limit B before the predicted vehicle 20 reaches its destination Dp, the update process identifies the point on the planned route where the predicted remaining energy level reaches the lower limit B. This method makes it possible to accurately identify the point where the predicted vehicle 20's remaining energy level reaches the lower limit B.

[0075] In this embodiment, the energy consumption prediction method notifies the user of an alarm and the location if the predicted change in remaining energy reaches the lower limit B before the prediction vehicle 20 reaches the destination Dp. This method makes it possible to notify the user that the remaining energy will reach the lower limit B and the location of that event before the prediction vehicle 20 reaches the destination Dp.

[0076] This embodiment is not limited to an energy consumption prediction method, but also includes an energy consumption prediction system 1. This energy consumption prediction system 1 comprises a vehicle 20 that runs using energy, and a server 10 that can communicate with the vehicle 20 and predicts the energy consumed by the vehicle 20. This energy consumption prediction system 1 executes the energy consumption prediction method described above. With this energy consumption prediction system 1, energy consumption can be learned based on the starting point and ending point of the journey without continuously collecting data from the starting point to the ending point along the journey path. Furthermore, energy consumption can be learned for each direction at each grid point G1. This makes it possible to improve the efficiency of collecting learning data while achieving accurate prediction of energy consumption.

[0077] Furthermore, this embodiment also includes the following information processing device. This information processing device includes a controller that predicts the energy consumed by the vehicle 20. The controller executes the energy consumption prediction method described above. With this information processing device, energy consumption can be learned based on the starting point and ending point of travel without continuously collecting data from the starting point to the ending point of travel along the travel route. In addition, energy consumption can be learned for each direction at each grid point G1. This makes it possible to improve the efficiency of collecting learning data while achieving accurate prediction of energy consumption. In this embodiment, the server 10 implements the function of an information processing device, but the vehicle 20 may also implement the function of an information processing device.

[0078] As described above, embodiments of the present invention have been presented, but the statements and drawings that constitute part of this disclosure should not be understood as limiting the invention. Various alternative embodiments, examples, and operational techniques will become apparent to those skilled in the art from this disclosure.

[0079] 1: Energy consumption forecasting system, 10: Server, 11: Controller, 12: Memory device, 13: Communication device, 20: Vehicles (collection vehicle, forecasting vehicle), 21: Controller, 22: Communication device, 23: Display device, 24: Memory device

Claims

1. An energy consumption prediction method for a system comprising a vehicle that runs using energy and a server capable of communicating with the vehicle, comprising: collecting driving data when the vehicle has traveled from a starting point to an ending point, the driving data including the starting point, the ending point, the distance traveled from the starting point to the ending point, and the amount of energy consumed by the vehicle between the starting point and the ending point; and performing a grid map learning process for predicting the energy consumption of the vehicle based on the driving data, wherein the grid map is configured such that each of a plurality of grid points associated with coordinates of map information is associated with a plurality of directions and the amount of energy consumed when the vehicle travels in each direction from the grid point.

2. The energy consumption prediction method according to claim 1, comprising: the learning process calculating a learning energy consumption rate based on the driving distance and energy consumption included in the driving data; identifying a learning target direction, which is the direction to be learned from among the plurality of directions, based on the direction of the vector from the driving start point to the driving end point; assigning a learning range to include the driving start point and the driving end point at the plurality of grid points associated with the learning target direction; extracting a learning target grid point, which is a grid point included in the learning range; and updating the energy consumption rate as the energy consumption associated with the learning target grid point based on the learning energy consumption rate.

3. The energy consumption prediction method according to claim 2, further comprising: acquiring route data including the planned route of the vehicle from the departure point to the destination; and performing a prediction process to predict the trend of energy consumption along the planned route based on the route data and the grid map.

4. The energy consumption prediction method according to claim 3, wherein the prediction process includes assigning a plurality of prediction target points on the planned travel route, identifying a prediction target direction which corresponds to the direction of travel of the vehicle for each of the plurality of prediction target points, and performing a mapping process to associate the energy consumption rate for each of the plurality of prediction target points, the mapping process includes identifying a prediction target grid point which corresponds to the prediction target point from among the plurality of grid points included in the grid map, and identifying the energy consumption rate associated with the prediction target direction at the prediction target grid point as the energy consumption rate at the prediction target point.

5. The energy consumption prediction method according to claim 3 or 4, wherein the prediction process includes predicting the change in the amount of remaining energy when the vehicle travels along the planned travel route, based on the prediction of the change in the energy consumption rate along the planned travel route.

6. The energy consumption prediction method according to any one of claims 3 to 5, further comprising notifying the user of the results of the prediction process.

7. The energy consumption prediction method according to claim 5, comprising: aggregating the measured values ​​of the energy consumption rate when the vehicle actually travels the planned route for each predetermined travel section; and performing an update process to update the predicted change in the remaining energy amount based on the aggregated results of the measured values ​​each time the vehicle travels the travel section.

8. The energy consumption prediction method according to claim 7, further comprising: the update process calculating a correction value to correct the discrepancy between the measured value and the predicted trend; and updating the predicted trend of the remaining energy amount from the vehicle's current location to the destination based on the correction value.

9. The energy consumption prediction method according to claim 7 or 8, further comprising notifying the user of the updated forecast of the remaining energy amount.

10. The energy consumption prediction method according to claim 7 or 8, wherein the update process includes predicting the amount of remaining energy when the vehicle arrives at the destination, based on the updated prediction of the trend of the remaining energy.

11. The energy consumption prediction method according to claim 10, further comprising informing the user of the amount of remaining energy when the vehicle arrives at the destination.

12. The energy consumption prediction method according to claim 10, wherein the update process includes identifying a point on the planned route where the predicted change in the remaining energy amount reaches a predetermined lower limit if the vehicle reaches the destination.

13. The energy consumption prediction method according to claim 12, wherein if the predicted change in the remaining energy amount reaches the lower limit before the vehicle reaches the destination, an alarm is issued to the user and the location is notified to the user.

14. An energy consumption prediction system comprising: a vehicle that runs using energy; and a server that can communicate with the vehicle and predicts the energy consumption of the vehicle, wherein the vehicle collects driving data when the vehicle has traveled from a starting point to an ending point, the driving data including the starting point, the ending point, the distance traveled from the starting point to the ending point, and the amount of energy consumed by the vehicle between the starting point and the ending point; transmits the driving data to the server; and the server performs a learning process for a grid map for predicting the energy consumption of the vehicle based on the driving data, wherein the grid map is configured such that each of a plurality of grid points associated with coordinates of map information is associated with a plurality of directions and the amount of energy consumed when the vehicle travels in each direction from the grid point.

15. An information processing device comprising a controller for predicting the energy consumption of a vehicle, wherein the controller acquires driving data when the vehicle has traveled from a starting point to an ending point, the driving data including the starting point, the ending point, the distance traveled from the starting point to the ending point, and the amount of energy consumed by the vehicle between the starting point and the ending point, and performs a learning process for a grid map for predicting the energy consumption of the vehicle based on the driving data, wherein the grid map is configured such that each of a plurality of grid points associated with coordinates of map information is associated with a plurality of directions and the amount of energy consumed when the vehicle travels in each direction from the grid point.