JP7338652B2 - Power consumption prediction device - Google Patents

Description

The present invention relates to a power consumption prediction device for predicting power consumption of an electric vehicle.

Techniques for predicting the power consumption of electric vehicles are known. Patent Literature 1 discloses a technique for estimating power consumption per unit traveling distance to be consumed from now on based on the power consumption per unit distance of an electric vehicle up to now.

WO2014/046071

However, since the power consumption of electric vehicles changes according to the slope of the road, predicting power consumption on future roads based on current power consumption per unit distance will reduce power consumption prediction accuracy. There was something I ended up doing. For example, if power consumption on a downhill slope is predicted based on the power consumption per unit distance of an electric vehicle traveling on an uphill slope, the predicted power consumption will be larger than the actual power consumption.

Therefore, the present invention has been made in view of these points, and an object of the present invention is to improve the prediction accuracy of power consumption.

In an aspect of the present invention, a plurality of pieces of travel data are acquired that associate a travel locus including height information on which the electric vehicle has traveled in the past and actual power consumption of the electric automobile when the electric automobile traveled the travel locus. a driving data acquisition unit, and when an arbitrary route is input using the plurality of driving data acquired by the driving data acquisition unit as teacher data, prediction of power consumption when the electric vehicle travels on the arbitrary route. A model creation unit that creates a power consumption prediction model that outputs a result, a planned travel route acquisition unit that acquires a planned travel route including height information on which the electric vehicle is scheduled to travel, and the above created by the model creation unit a power consumption prediction unit that predicts power consumption when the electric vehicle travels along the planned travel route by inputting the planned travel route acquired by the planned travel route acquisition unit into a power consumption prediction model; A power consumption prediction device with

The travel data acquisition unit acquires a two-dimensional travel trajectory, which is a past travel trajectory of the electric vehicle and does not include height information, and three-dimensional map information including height information, and obtains the three-dimensional map information. By acquiring height information corresponding to each of the plurality of points included in the two-dimensional running trajectory from the above, and associating each of the plurality of points with height information corresponding to each of the plurality of points, A three-dimensional running trajectory including height information may be generated.

The planned travel route acquisition unit acquires a planned travel route that is a route that the electric vehicle travels and does not include height information, and three-dimensional map information that includes height information. Obtaining height information corresponding to each of a plurality of points included in the two-dimensional planned travel route from map information, and associating each of the plurality of points with height information corresponding to each of the plurality of points. may generate a three-dimensional planned travel route including height information.

The planned travel route acquisition unit acquires the current position of the electric vehicle that is running, sequentially acquires a remaining planned route that is a route from the current position to the end point of the planned travel route, and the power consumption prediction unit and inputting the remaining scheduled route to the power consumption prediction model each time the scheduled driving route acquisition unit acquires the remaining scheduled route, thereby calculating the power consumption when the electric vehicle travels on the remaining scheduled route. You can predict.

The travel data acquisition unit acquires the travel data in which the travel locus and the actual power consumption are further associated with history information indicating the state of the electric vehicle when the electric vehicle travels along the travel locus, The model creation unit uses the travel data that associates the travel trajectory, the actual power consumption, and the history information as training data to create an arbitrary route and a predicted state of the electric vehicle when traveling on the arbitrary route. is input, the power consumption prediction model may be created for outputting a prediction result of power consumption when the electric vehicle in the predicted state travels on the arbitrary route.

The planned travel route acquisition unit acquires the current position of the electric vehicle that is running, sequentially acquires a remaining planned route that is a route from the current position to the end point of the planned travel route, and the power consumption prediction device and a prediction information acquisition unit that sequentially acquires a prediction state predicted as a state of the electric vehicle traveling on the remaining scheduled route acquired by the scheduled travel route acquisition unit, wherein the power consumption prediction unit Each time the prediction information acquiring unit acquires the prediction state, the remaining scheduled route and the prediction state corresponding to the remaining scheduled route are input to the power consumption prediction model, thereby obtaining the prediction state as the prediction state. Power consumption when the electric vehicle travels the remaining scheduled route may be predicted.

The travel data acquisition unit obtains the travel data in which the travel locus and the actual power consumption are further associated with history information indicating road conditions when the electric vehicle travels on the road corresponding to the travel locus. The model creation unit uses the travel data in which the travel locus, the actual power consumption, and the history information are associated with each other as teacher data to create an arbitrary route and the predicted situation of the road corresponding to the arbitrary route. The power consumption prediction model may be created for outputting a prediction result of power consumption when the electric vehicle travels the arbitrary route that is the predicted situation when the prediction information is input.

The planned travel route acquisition unit acquires the current position of the electric vehicle that is running, sequentially acquires a remaining planned route that is a route from the current position to the end point of the planned travel route, and the power consumption prediction device a prediction information acquisition unit configured to sequentially acquire predicted road conditions predicted as conditions when the electric vehicle travels on the remaining scheduled route acquired by the planned travel route acquisition unit; and the power consumption prediction unit. each time the prediction information acquiring unit acquires the prediction situation, by inputting the remaining scheduled route and the prediction situation corresponding to the remaining scheduled route to the power consumption prediction model, Power consumption may be predicted when the electric vehicle travels along the planned travel route.

ADVANTAGE OF THE INVENTION According to this invention, there exists an effect that the prediction accuracy of power consumption can be improved.

It is a figure which shows typically the structure of the power consumption prediction apparatus which concerns on embodiment. It is a figure which shows a two-dimensional running locus typically. FIG. 4 is a diagram schematically showing changes in altitude corresponding to the position of an electric vehicle; 4 is a flowchart showing an example of the flow of processing for creating a power consumption prediction model; 4 is a flow chart showing an example of the flow of processing for estimating power consumption when an electric vehicle runs.

[Configuration of power consumption prediction device 1 according to embodiment]
FIG. 1 is a diagram schematically showing the configuration of a power consumption prediction device 1 according to an embodiment. A power consumption prediction device 1 is, for example, a server that predicts the power consumption of an electric vehicle. The power consumption prediction device 1 has a communication unit 11 , a storage unit 12 and a control unit 13 .

The communication unit 11 is a communication module for transmitting/receiving information to/from another device via a communication line. For example, the communication unit 11 transmits and receives information to and from the electric vehicle via wireless communication. In addition, the communication unit 11 can transmit and receive information to and from a terminal of an administrator who manages operation of the electric vehicle, a terminal of a driver who drives the electric vehicle, and other servers.

The storage unit 12 is a storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), a hard disk, and the like. The storage unit 12 stores programs executed by the control unit 13 . The storage unit 12 stores three-dimensional map information including height information. Three-dimensional map information is a map in which each of a plurality of points included in a two-dimensional map is associated with, for example, altitude as height information.

The control unit 13 is a computing resource including a processor such as a CPU (Central Processing Unit). By executing the programs stored in the storage unit 12, the control unit 13 functions as a travel data acquisition unit 131, a model creation unit 132, a planned travel route acquisition unit 133, a prediction information acquisition unit 134, and a power consumption prediction unit 135. Realize the function.

The travel data acquisition unit 131 acquires a travel locus traveled by the electric vehicle. For example, the travel data acquisition unit 131 acquires from the electric vehicle a two-dimensional travel trajectory that is a past travel trajectory of the electric vehicle and does not include height information. Specifically, the travel data acquisition unit 131 acquires the two-dimensional travel locus created by the electric vehicle when the electric vehicle finishes traveling. The travel data acquisition unit 131 may also acquire a two-dimensional travel locus created by sequentially acquiring the current position of the electric vehicle while the electric vehicle is running.

The two-dimensional travel trajectory is, for example, a trajectory in which the positions of the electric vehicle are recorded at predetermined time intervals. FIG. 2 is a diagram schematically showing a two-dimensional travel locus. Positions P1 to P5 indicated by white circles in FIG. 2 indicate positions of the electric vehicle obtained at different times. It is assumed that the positions P1 to P5 are acquired in ascending numerical order. The position of the electric vehicle is indicated by latitude and longitude, for example. The electric vehicle obtains its own position at predetermined intervals, associates the obtained position with the time when the position was obtained, and notifies the power consumption prediction device 1 of the obtained position. The predetermined interval may be set as appropriate, and a specific value is, for example, 10 seconds.

The travel data acquisition unit 131 acquires from the electric vehicle two-dimensional travel data that associates the two-dimensional travel locus with the actual power consumption of the electric vehicle when traveling the two-dimensional travel locus. The actual power consumption is, for example, the power consumed when the electric vehicle travels from one point on the two-dimensional travel locus to the next point. As an example, two-dimensional travel data is acquired by associating actual power consumption when traveling from position P1 to position P2 with position P2. In addition, the travel data acquisition unit 131 acquires the remaining capacity of the battery mounted on the electric vehicle at each position included in the two-dimensional travel trajectory, and obtains the remaining capacity at a certain point and the remaining capacity at the point next to the certain point. You may acquire the difference with a capacity|capacitance as power consumption.

The travel data acquisition unit 131 acquires a three-dimensional planned travel route based on the two-dimensional planned travel route. Specifically, first, the travel data acquisition unit 131 acquires the three-dimensional map information stored in the storage unit 12, and from the three-dimensional map information, each of the positions of the plurality of electric vehicles included in the two-dimensional travel locus. Get altitude as corresponding height information. More specifically, the travel data acquisition unit 131 identifies a position (coordinates) on the three-dimensional map corresponding to the position (coordinates) included in the two-dimensional travel locus, and obtains the identified position (coordinates) on the three-dimensional map. ) is acquired as height information of points on the two-dimensional running locus.

FIG. 3 is a diagram schematically showing changes in altitude H corresponding to the position of the electric vehicle. The horizontal axis of FIG. 3 indicates the distance R, and the vertical axis indicates the altitude H. As shown in FIG. Positions P1 to P5 indicated by white circles in FIG. 3 indicate positions of the electric vehicle including height information. As an example, the travel data acquisition unit 131 acquires from the three-dimensional map information the altitude H corresponding to the position P4 (see FIG. 2) acquired while traveling on the mountainside, and associates it with the position P4. As shown in FIG. 3, it can be seen that the position P4 is higher than the other positions.

The traveling data acquisition unit 131 generates a three-dimensional traveling locus including height information by associating each of the plurality of positions included in the two-dimensional traveling locus with an altitude corresponding to each of the plurality of positions. Then, the travel data acquisition unit 131 acquires three-dimensional travel data that associates the three-dimensional travel locus with the actual power consumption. Specifically, the travel data acquisition unit 131 obtains three-dimensional travel data that associates the three-dimensional travel locus with the actual power consumption when the electric vehicle travels the two-dimensional travel locus corresponding to the three-dimensional travel locus. get.

The travel data acquisition unit 131 may acquire three-dimensional travel data in which the travel locus and the actual power consumption are further associated with history information when the electric vehicle travels along the travel locus. The history information is information indicating the state of the electric vehicle when the electric vehicle travels along the three-dimensional travel locus. The state of the electric vehicle includes, for example, the speed and acceleration of the electric vehicle when the electric vehicle travels on a road corresponding to the three-dimensional travel trajectory, the weight of the vehicle, the weight of the load, the operating state of the auxiliary equipment installed, and the state of the body. state, remaining capacity of the battery, and the like.

The history information may be information indicating road conditions when the electric vehicle travels on the road corresponding to the three-dimensional travel locus. The condition of the road corresponding to the three-dimensional travel locus is the condition of the road corresponding to the three-dimensional route at the time or time zone when the electric vehicle travels along the three-dimensional travel locus. The road conditions are, for example, environmental information at the time or time zone when the electric vehicle travels on the road. The environmental information is, for example, the weather, the weather, the condition of the road surface (dry or wet), the wind direction, and the wind speed. In addition, road conditions may include road congestion degree and congestion information. Note that the history information may include both information indicating the state of the electric vehicle and road conditions.

The travel data acquisition unit 131 acquires three-dimensional travel data each time the electric vehicle finishes traveling, and stores the acquired three-dimensional travel data in the storage unit 12 . The storage unit 12 stores the three-dimensional travel data acquired by the travel data acquisition unit 131 in association with vehicle identification information that identifies the electric vehicle.

The model creating unit 132 creates a power consumption prediction model for predicting the power consumption of the electric vehicle. For example, the model creation unit 132 creates a power consumption prediction model when the accumulated number of travel data stored in the storage unit 12 reaches or exceeds a threshold value. The threshold value is a number at which it can be considered that sufficient travel data has been acquired, for example, the number of travel data for half a year.

When an arbitrary route is input using a plurality of three-dimensional travel data stored in the storage unit 12 as teaching data, the model creation unit 132 calculates power consumption when the electric vehicle travels along the input arbitrary route. Create a power consumption prediction model that outputs the prediction result of The model creating unit 132 can create a power consumption prediction model using a known machine learning model. Any algorithm can be used for the machine learning model, but linear multiple regression, for example, can be used. Also, the algorithm may use a regression method using a decision tree or a regression method using a random forest. The algorithm may be appropriately set according to the three-dimensional travel data acquired by the travel data acquisition unit 131 and the specifications of the vehicle. In this way, the model creation unit 132 creates a power consumption prediction model using a three-dimensional travel trajectory that includes height information when the electric vehicle actually traveled in the past. A corresponding power consumption prediction model can be created.

When the travel data that associates the travel locus, the actual power consumption, and the state of the electric vehicle as history information is acquired, the model creation unit 132 generates the travel data that associates the travel locus, the actual power consumption, and the state of the electric vehicle. is used as training data to create a power consumption prediction model. For example, when an arbitrary route and information indicating the state of the electric vehicle are input, the model creation unit 132 predicts the power consumption when the electric vehicle in the input state travels the arbitrary route. Create a power consumption prediction model to output. Specifically, the model creation unit 132 uses driving data that associates the driving trajectory, the actual power consumption, and the vehicle speed of the electric vehicle as the state of the electric vehicle with the driving data as training data, and selects an arbitrary route and the arbitrary route. Create a power consumption prediction model that outputs a prediction result of power consumption when traveling at a vehicle speed of .

When the travel data that associates the travel locus, the actual power consumption, and the road conditions as history information is acquired, the model creation unit 132 creates the travel data that associates the travel locus, the actual power consumption, and the road conditions. Create a power consumption prediction model as training data. For example, when an arbitrary route and information indicating road conditions are input, the model creation unit 132 predicts the power consumption when an electric vehicle travels on the arbitrary route, which is the input road condition. Create a power consumption prediction model that outputs Specifically, the model creation unit 132 uses driving data that associates the driving trajectory, the actual power consumption, and the past weather when the electric vehicle traveled the driving trajectory as road conditions as teacher data, and uses any When a route and arbitrary weather (for example, rain) are input, a power consumption prediction model is created that outputs a prediction result of power consumption when an electric vehicle travels on the arbitrary route when the weather is rainy.

Note that the model creating unit 132 creates a power consumption prediction model using driving data that associates the travel locus, the actual power consumption, the state of the electric vehicle, and the state of the road as teaching data. For example, the model creation unit 132 creates a power consumption prediction model using driving data that associates the travel locus, the actual power consumption, the vehicle speed of the electric vehicle, and the weather on the road when the vehicle travels along the travel locus as teaching data. By doing so, the model creation unit 132 can create a power consumption prediction model that takes into consideration the state of the electric vehicle and the state of the road when the vehicle travels along the travel locus.

The planned travel route acquisition unit 133 acquires a planned travel route along which the electric vehicle is scheduled to travel. For example, the planned travel route acquisition unit 133 acquires a two-dimensional planned travel route that is a route that the electric vehicle is scheduled to travel and does not include height information. Specifically, the planned travel route acquisition unit 133 acquires the planned two-dimensional travel route input to the terminal by the manager who manages the operation of the electric vehicle or the driver who drives the electric vehicle. The terminal may be any device capable of inputting a planned two-dimensional travel route, and may be, for example, a personal computer, a smart phone, or a car navigation system installed in an electric vehicle, but is not limited to this.

The two-dimensional planned travel route is, for example, a route that indicates the position of the electric vehicle for each predetermined distance. The planned two-dimensional travel route is a route in which the position of the electric vehicle is indicated by latitude and longitude, similar to the two-dimensional travel locus of FIG. Also, the planned two-dimensional travel route may be, for example, a road represented as a plurality of nodes. In this case, each node has the relative position coordinates of the next node with respect to a certain node, cumulative travel distance, two-dimensional coordinates indicated by latitude and longitude, node attributes (intersections, T-junctions, railroad crossings, tunnels), Information such as direction of travel (straight ahead, right turn, left turn, U-turn), legal speed between nodes, etc. may be included.

Next, the planned travel route acquisition unit 133 acquires a planned three-dimensional travel route based on the planned two-dimensional travel route. Specifically, first, the planned travel route acquisition unit 133 acquires the 3D map information stored in the storage unit 12, and from the 3D map information, the positions of the plurality of electric vehicles included in the planned 2D travel route are determined. Altitude H is acquired as height information corresponding to each. Next, the planned travel route acquisition unit 133 associates each of the positions of the plurality of electric vehicles with height information corresponding to each of the positions of the plurality of electric vehicles, thereby obtaining a three-dimensional travel plan including height information. Generate a route. Then, the planned travel route acquisition unit 133 acquires the generated planned three-dimensional travel route as a planned three-dimensional travel route including height information on which the electric vehicle is expected to travel.

The prediction information acquisition unit 134 acquires prediction information corresponding to the planned three-dimensional travel route when the model creation unit 132 has created a power consumption prediction model considering history information. The prediction information is prediction state information that predicts the state of the electric vehicle when it travels along the planned three-dimensional travel route. The predicted state information includes, for example, the speed of the electric vehicle when the electric vehicle travels along the planned three-dimensional travel route (hereinafter referred to as the "planned travel speed"), acceleration, weight of the load carried by the electric vehicle, and the number of auxiliary equipment. They are the operating state, the state of the bodywork, the remaining capacity of the battery, etc. For example, the prediction information acquisition unit 134 obtains the planned travel speed when the electric vehicle travels along the planned three-dimensional travel route and the weight of the cargo to be loaded, from the electric vehicle. Acquired from a server that manages vehicle operation.

Also, the prediction information may be prediction situation information that predicts the road situation corresponding to the planned three-dimensional travel route. The predicted situation information is information that predicts the situation of the road corresponding to the planned three-dimensional travel route when the electric vehicle travels along the planned three-dimensional travel route. The planned traveling route acquisition unit 133 receives, as the predicted situation information, external information such as environmental information, predicted road congestion degree, or predicted traffic congestion information at the time or time period when the electric vehicle is scheduled to travel on the planned three-dimensional traveling route. Obtained from a device (eg, a server managed by a weather forecasting agency). Specifically, the planned travel route acquisition unit 133 acquires, as prediction information, the weather of the planned three-dimensional travel route at the time or time period when the electric vehicle travels the planned three-dimensional travel route. Note that the prediction information acquisition unit 134 may acquire prediction information including prediction state information that predicts the state of the electric vehicle and prediction situation information that predicts the road situation.

The power consumption prediction unit 135 predicts power consumption when the electric vehicle travels the planned three-dimensional travel route acquired by the planned travel route acquisition unit 133 . Specifically, the power consumption prediction unit 135 inputs the planned three-dimensional travel route to the power consumption prediction model created by the model creation unit 132, thereby calculating the power consumption when the electric vehicle travels along the planned three-dimensional travel route. Predict power. In this way, the power consumption prediction unit 135 predicts power consumption by inputting a planned three-dimensional travel route including changes in gradient and elevation into a power consumption prediction model that considers changes in gradient and elevation. As a result, the power consumption prediction unit 135 uses a power consumption prediction model based on changes in elevation and slope of roads on which the electric vehicle has traveled in the past, even if the electric vehicle has not traveled in the past. Since it is possible to analogize the power consumption due to changes in the altitude and gradient of the planned three-dimensional travel route, it is possible to improve the prediction accuracy of the power consumption.

When a power consumption prediction model is created in consideration of historical information indicating the state of the electric vehicle in the past and road conditions, the power consumption prediction unit 135 predicts a three-dimensional travel route and a three-dimensional travel plan corresponding to the three-dimensional travel route. Prediction information is input to a power consumption prediction model to predict power consumption. By doing so, the power consumption prediction unit 135 can predict the power consumption when the electric vehicle travels the planned three-dimensional travel route in the state indicated by the prediction information, for example.

For example, when a planned travel speed is acquired as prediction information, the power consumption prediction unit 135 inputs the planned three-dimensional travel route and the planned travel speed corresponding to the planned three-dimensional travel route to the power consumption prediction model. is used to predict the power consumption when the electric vehicle travels along the planned three-dimensional traveling route at the planned traveling speed. Further, when the predicted weather of the planned three-dimensional travel route is acquired as the prediction information, the power consumption prediction unit 135 inputs the planned three-dimensional travel route and the predicted weather of the planned three-dimensional travel route. As a result, it is possible to predict the power consumption when the electric vehicle travels along the three-dimensional planned travel route in the predicted weather. By doing so, the power consumption prediction unit 135 can predict the power consumption in consideration of the state of the electric vehicle indicated by the prediction information and the state of the road, so that the prediction accuracy of the power consumption can be improved.

The power consumption prediction unit 135 predicts the power consumption of the remaining scheduled routes excluding the route already traveled by the electric vehicle while the electric vehicle is running. In this case, the planned travel route acquisition unit 133 acquires the current position of the running electric vehicle, and obtains the remaining planned route, which is the route from the current position to the end point of the planned three-dimensional travel route (hereinafter referred to as “destination”). are obtained sequentially. The interval at which the planned travel route acquisition unit 133 acquires the current position of the electric vehicle may be determined as appropriate, and is, for example, 10 seconds. Next, the power consumption prediction unit 135 predicts the power consumption when the electric vehicle travels on the remaining scheduled route by inputting the remaining scheduled route into the power consumption prediction model each time the remaining scheduled route is obtained. By doing so, it is possible to sequentially predict the power consumption when traveling from the current position to the destination until the electric vehicle reaches the destination.

By the way, the state of electric vehicles and the state of roads change from moment to moment. Therefore, by predicting the power consumption in consideration of the state of the electric vehicle and road conditions that change from moment to moment, the accuracy of the prediction of the power consumption of the electric vehicle can be improved. Therefore, the power consumption prediction device 1 acquires prediction information corresponding to the remaining scheduled route while the electric vehicle is running, and predicts power consumption based on the acquired prediction information. The process of predicting the power consumption by the power consumption prediction device 1 based on the prediction information will be specifically described below.

The prediction information acquisition unit 134 sequentially acquires the prediction information of the remaining planned route acquired by the planned travel route acquisition unit 133 . For example, the prediction information acquisition unit 134 acquires environmental information in a time period during which the electric vehicle is expected to travel the remaining scheduled route as prediction information from the external device. The prediction information acquisition unit 134 also acquires, as prediction information, the expected running speed of the electric vehicle when traveling the remaining planned route from the server that manages the operation of the electric vehicle. Note that the prediction information acquisition unit 134 may acquire prediction information corresponding to the remaining scheduled route (for example, the degree of congestion based on the number of vehicles traveling on the remaining scheduled route) from an external device. Prediction information corresponding to the remaining scheduled route may be acquired from information detected by the sensor (for example, the degree of congestion based on the number of other vehicles existing around the own vehicle).

The power consumption prediction unit 135 predicts the power consumption by inputting the remaining scheduled route and the prediction information corresponding to the remaining scheduled route into the power consumption prediction model each time the remaining scheduled route and the prediction information are acquired. . For example, the power consumption prediction unit 135 predicts the power consumption when the electric vehicle in the state indicated by the prediction information travels the remaining scheduled route in the road condition indicated by the prediction information. For example, when the remaining scheduled route and the weather (for example, rain) of the remaining scheduled route as prediction information are acquired, the power consumption prediction unit 135 consumes the remaining scheduled route and the weather (rain) of the remaining scheduled route. The data is input to a power prediction model to predict power consumption when an electric vehicle travels a remaining scheduled route with rainy weather. By doing so, the power consumption prediction unit 135 can acquire a prediction result of power consumption according to the state of the electric vehicle and road conditions that change from moment to moment.

[Processing for updating the power consumption prediction model]
FIG. 4 is a flowchart illustrating an example of the flow of processing for creating a power consumption prediction model. First, when the electric vehicle finishes running, the running data acquisition unit 131 acquires the running data of the electric vehicle (step S1). Specifically, the travel data acquisition unit 131 obtains three-dimensional travel data that associates the three-dimensional travel locus traveled by the electric vehicle with the actual power consumption of the electric vehicle when traveling the three-dimensional travel locus. get.

Next, the model creation unit 132 determines whether or not the accumulated number of acquired travel data is equal to or greater than a threshold value (step S2). When the accumulated number of acquired travel data is less than the threshold (No in step S2), the model creation unit 132 does not update the power consumption prediction model until the accumulated number of acquired travel data reaches or exceeds the threshold.

When the cumulative number of acquired travel data is equal to or greater than the threshold (Yes in step S2), the model creation unit 132 updates the power consumption prediction model (step S3). Specifically, the model creation unit 132 creates a new power consumption prediction model by using the number of travel data equal to or greater than the acquired threshold as teacher data. Then, after creating the power consumption model, the model creating unit 132 sets the cumulative number of travel data to 0 (step S4). At this time, the model creation unit 132 deletes the acquired travel data.

By doing so, the model creating unit 132 can periodically (every six months) update the power consumption prediction model based on the travel data for a predetermined period (for example, half a year). As a result, it is possible to create a power consumption prediction model that corresponds to changes in the state of the electric vehicle over time.

[Processing to predict power consumption]
FIG. 5 is a flowchart showing an example of the flow of processing for estimating power consumption when an electric vehicle runs. It is assumed that a power consumption prediction model has already been created.

First, the planned travel route acquisition unit 133 acquires the planned travel route of the electric vehicle (step S11). Specifically, the planned travel route acquisition unit 133 acquires the planned two-dimensional travel route of the electric vehicle, and calculates the altitude H corresponding to each of the positions of the plurality of electric vehicles included in the planned two-dimensional travel route. Acquired from map information. The planned traveling route acquisition unit 133 associates each of the positions of the plurality of electric vehicles with an altitude H corresponding to each of the positions of the plurality of electric vehicles, and creates a planned three-dimensional traveling route to obtain the three-dimensional traveling plan. Get directions.

Next, the power consumption prediction unit 135 predicts power consumption when the electric vehicle travels along the planned three-dimensional travel route (step S12). Specifically, the power consumption prediction unit 135 inputs the planned three-dimensional travel route into the power consumption prediction model, thereby obtaining a prediction result of power consumption when the electric vehicle travels the planned three-dimensional travel route.

When the electric vehicle starts traveling along the planned three-dimensional traveling route, the planned traveling route acquisition unit 133 sequentially acquires the current positions of the traveling electric vehicle (step S13). The planned travel route acquisition unit 133 determines whether the electric vehicle has arrived at the destination based on the current position of the electric vehicle (step S14). For example, when the current position of the electric vehicle is within a predetermined range from the destination, which is the end point of the three-dimensional planned travel route, the planned travel route acquisition unit 133 determines that the electric vehicle has arrived at the destination. The predetermined range may be determined appropriately, for example, within the premises of the facility corresponding to the destination. When the electric vehicle has arrived at the destination (Yes in step S14), the power consumption prediction device 1 terminates the power consumption prediction process.

If the electric vehicle has not arrived at the destination (No in step S14), the planned travel route acquisition unit 133 acquires a remaining planned route (step S15). Specifically, the planned travel route acquisition unit 133 acquires the remaining planned route, which is the route from the current position of the electric vehicle to the destination.

The prediction information acquisition unit 134 acquires the prediction information of the remaining planned route acquired by the planned travel route acquisition unit 133 (step S16). For example, the prediction information acquisition unit 134 acquires environmental information in a time period during which the vehicle is expected to travel the remaining scheduled route as prediction information from an external device.

The power consumption prediction unit 135 predicts power consumption when the electric vehicle travels the remaining scheduled route (step S17). Specifically, the power consumption prediction unit 135 inputs the remaining scheduled route and the prediction information of the remaining scheduled route into the power consumption prediction model, so that when the electric vehicle travels the remaining scheduled route that is the prediction information, Get the prediction result of the power consumption of .

The power consumption prediction device 1 repeats steps S13 to S17 from when the electric vehicle starts running until it reaches the destination, and sequentially predicts the power consumption of the remaining scheduled routes. Then, the power consumption prediction device 1 notifies the terminal of the administrator who manages the operation of the electric vehicle and the terminal of the driver who drives the electric vehicle of the predicted power consumption. In this way, the administrator or driver who has confirmed the notification can compare the predicted power consumption according to the constantly changing state and situation with the remaining capacity of the battery installed in the electric vehicle. become. As a result, it becomes easier for the administrator and the driver to determine whether the electric vehicle can travel to the destination along the planned travel route.

(Modification)
In the embodiment, the travel data acquisition unit 131 and the planned travel route acquisition unit 133 acquire the 3D map information from the storage unit 12, but the 3D map information may be acquired from other devices. . The other device is, for example, a server managed by a surveying organization or a server managed by a company that creates maps (hereinafter referred to as a "map information providing server"). The map information providing server stores information indicating a three-dimensional map including altitude, longitude, and latitude as height information. The travel data acquisition unit 131 and the planned travel route acquisition unit 133 acquire a three-dimensional map corresponding to the two-dimensional travel locus and the planned two-dimensional travel route from the map information providing server. For example, the travel data acquisition unit 131 acquires a part of the 3D map that includes the 2D travel locus from among the 3D maps stored by the map information providing server.

[Effects of the power consumption prediction device 1 according to the embodiment]
As described above, when the electric vehicle finishes running, the power consumption prediction device 1 generates a plurality of values that associate the three-dimensional running trajectory along which the electric vehicle traveled and the actual power consumption when the three-dimensional running trajectory was traveled. A power consumption prediction model for an electric vehicle is created using three-dimensional driving data as teacher data. Then, the power consumption prediction device 1 inputs a planned three-dimensional travel route along which the electric vehicle is scheduled to travel to the created power consumption prediction model before or during the travel of the electric vehicle. Predict power consumption when traveling along a planned three-dimensional travel route.

In this way, the power consumption prediction device 1 creates a power consumption prediction model using the three-dimensional travel trajectory of an electric vehicle that has traveled in the past. Can create predictive models. In addition, by inputting a three-dimensional planned route, the power consumption prediction device 1 can predict the power consumption according to changes in road gradients even on a route that an electric vehicle has never traveled. Power prediction accuracy can be improved.

Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of the gist thereof. be. For example, all or part of the device can be functionally or physically distributed and integrated in arbitrary units. In addition, new embodiments resulting from arbitrary combinations of multiple embodiments are also included in the embodiments of the present invention. The effect of the new embodiment caused by the combination has the effect of the original embodiment.

1 power consumption prediction device 11 communication unit 12 storage unit 13 control unit 131 travel data acquisition unit 132 model creation unit 133 scheduled travel route acquisition unit 134 prediction information acquisition unit 135 power consumption prediction unit

Claims (7)

a traveling data acquisition unit that acquires a plurality of pieces of traveling data that associate a traveling locus including height information on which the electric vehicle has traveled in the past with the actual power consumption that the electric vehicle consumed when traveling the traveling locus;
When an arbitrary route is input using the plurality of traveling data acquired by the traveling data acquisition unit as teaching data, the electric vehicle outputs a prediction result of power consumption when the electric vehicle travels on the arbitrary route. a model creation unit that creates a prediction model;
A planned travel route acquisition unit that acquires a planned travel route including height information on which the electric vehicle is scheduled to travel;
By inputting the planned traveling route acquired by the planned traveling route acquisition unit into the power consumption prediction model created by the model creating unit, power consumption when the electric vehicle travels on the planned traveling route is predicted. a power consumption prediction unit that
has
The travel data acquisition unit acquires a two-dimensional travel trajectory, which is a past travel trajectory of the electric vehicle and does not include height information, and three-dimensional map information including height information, and obtains the three-dimensional map information. By acquiring height information corresponding to each of the plurality of points included in the two-dimensional running trajectory from the above, and associating each of the plurality of points with height information corresponding to each of the plurality of points, generating a three-dimensional running trajectory including height information;
Power consumption prediction device. a traveling data acquisition unit that acquires a plurality of pieces of traveling data that associate a traveling locus including height information on which the electric vehicle has traveled in the past with the actual power consumption that the electric vehicle consumed when traveling the traveling locus;
When an arbitrary route is input using the plurality of traveling data acquired by the traveling data acquisition unit as teaching data, the electric vehicle outputs a prediction result of power consumption when the electric vehicle travels on the arbitrary route. a model creation unit that creates a prediction model;
A planned travel route acquisition unit that acquires a planned travel route including height information on which the electric vehicle is scheduled to travel;
By inputting the planned traveling route acquired by the planned traveling route acquisition unit into the power consumption prediction model created by the model creating unit, power consumption when the electric vehicle travels on the planned traveling route is predicted. a power consumption prediction unit that
has
The planned travel route acquisition unit acquires a planned travel route that is a route that the electric vehicle travels and does not include height information, and three-dimensional map information that includes height information. Obtaining height information corresponding to each of a plurality of points included in the two-dimensional planned travel route from map information, and associating each of the plurality of points with height information corresponding to each of the plurality of points. to generate a 3D planned travel route including height information,
Power consumption prediction device. The planned travel route acquisition unit acquires the current position of the electric vehicle that is running, and sequentially acquires a remaining planned route that is a route from the current position to the end point of the planned travel route,
The power consumption prediction unit inputs the remaining planned route to the power consumption prediction model each time the planned driving route acquisition unit acquires the remaining planned route, thereby causing the electric vehicle to travel the remaining planned route. Predict power consumption when
The power consumption prediction device according to claim 1 or 2 . The travel data acquisition unit acquires the travel data in which the travel locus and the actual power consumption are further associated with history information indicating the state of the electric vehicle when the electric vehicle travels along the travel locus,
The model creation unit uses the travel data that associates the travel trajectory, the actual power consumption, and the history information as training data to create an arbitrary route and a predicted state of the electric vehicle when traveling on the arbitrary route. is input, creating the power consumption prediction model that outputs a prediction result of the power consumption when the electric vehicle in the prediction state travels on the arbitrary route;
The power consumption prediction device according to any one of claims 1 to 3 . The planned travel route acquisition unit acquires the current position of the electric vehicle that is running, and sequentially acquires a remaining planned route that is a route from the current position to the end point of the planned travel route,
The power consumption prediction device further includes a prediction information acquisition unit that sequentially acquires a prediction state predicted as a state when the electric vehicle travels the remaining planned route acquired by the planned travel route acquisition unit,
The power consumption prediction unit inputs the remaining scheduled route and the prediction state corresponding to the remaining scheduled route to the power consumption prediction model each time the prediction information acquisition unit acquires the prediction state. , predicting power consumption when the electric vehicle in the predicted state travels the remaining scheduled route;
The power consumption prediction device according to claim 4 . The travel data acquisition unit obtains the travel data in which the travel locus and the actual power consumption are further associated with history information indicating road conditions when the electric vehicle travels on the road corresponding to the travel locus. Acquired,
The model creation unit uses the travel data in which the travel locus, the actual power consumption, and the history information are associated with each other as teaching data, and prediction information indicating an arbitrary route and a predicted situation of a road corresponding to the arbitrary route. is input, creating the power consumption prediction model that outputs a prediction result of the power consumption when the electric vehicle travels the arbitrary route that is the predicted situation;
The power consumption prediction device according to any one of claims 1 to 3 . The planned travel route acquisition unit acquires the current position of the electric vehicle that is running, and sequentially acquires a remaining planned route that is a route from the current position to the end point of the planned travel route,
The power consumption prediction device further includes a prediction information acquisition unit that sequentially acquires predicted road conditions predicted as conditions when the electric vehicle travels the remaining planned route acquired by the planned travel route acquisition unit. ,
The power consumption prediction unit inputs the remaining scheduled route and the prediction state corresponding to the remaining scheduled route to the power consumption prediction model each time the prediction information acquisition unit acquires the prediction state. , predicting power consumption when the electric vehicle travels the planned travel route that is the predicted situation;
The power consumption prediction device according to claim 6 .

Images