Cruising distance prediction method and cruising distance prediction device

Abstract

A cruising distance prediction method for predicting the cruising distance of a vehicle traveling by electricity, wherein a controller (20) mounted on the vehicle acquires electricity efficiency element information for a plurality of elements related to the electricity efficiency of the vehicle, calculates the electricity efficiency for each user by machine learning the electricity efficiency element information, and calculates the cruising distance on the basis of the calculated electricity efficiency.

Description

Driving range prediction method and driving range prediction device

[0001] This invention relates to a method for predicting cruising range and a device for predicting cruising range.

[0002] Conventionally, electric vehicles (hereinafter simply referred to as "vehicles") that run on electricity display the vehicle's remaining driving range based on the remaining battery charge (see, for example, Patent Document 1). The vehicle in Patent Document 1 calculates the remaining driving range (driving range) using actual fuel consumption data based on the driver's driving style, driving environment, and driving time.

[0003] Japanese Patent Publication No. 2011-172407

[0004] However, in Patent Document 1, depending on the estimation accuracy of the State of Charge (SOC) of the high-voltage battery, the actual fuel consumption data may not be accurate to begin with. In this case, there is a problem that the accuracy of the calculation of the driving range will also decrease.

[0005] The present invention relates to a method for predicting cruising range that can predict cruising range with high accuracy, and a cruising range prediction device.

[0006] In a driving range prediction method according to one aspect of this disclosure, a computer acquires energy efficiency element information for each of several elements related to the vehicle's energy efficiency, and calculates the energy efficiency for each user by machine learning this energy efficiency element information. Then, based on the calculated energy efficiency for each user, it calculates the driving range when the user drives the vehicle.

[0007] In this embodiment, instead of using actual measured data of electricity consumption itself, multiple electricity consumption element information related to electricity consumption is acquired, and electricity consumption for each user is calculated by machine learning using this electricity consumption element information. This makes it possible to predict electricity consumption with high accuracy based on electricity consumption element information for each user, without relying on the estimation accuracy of the SOC, and the driving range can be calculated with greater accuracy.

[0008] A diagram showing the schematic configuration of a vehicle equipped with the range prediction device of the first embodiment of this disclosure. A block diagram showing the functional configuration of the controller of the first embodiment. A diagram showing the changes in the power consumption calculated by the power consumption calculation unit when using initial data values, and the changes in the vehicle speed learning value and driving power learning value learned for each trip. A flowchart showing the range prediction method of the first embodiment. A diagram showing the method for acquiring power consumption element information in the second embodiment.

[0009] [First Embodiment] The first embodiment of the present disclosure will be described below. Figure 1 is a schematic diagram of a vehicle equipped with a range prediction device of the first embodiment. The vehicle of this embodiment is an electric vehicle driven by an electric motor, and as shown in Figure 1, the vehicle includes a measurement sensor 11, a position detection sensor 12, a navigation device 13, a battery unit 14, an actuator 15, an air conditioner 16, other auxiliary equipment 17, a controller 20, etc. The vehicle of this embodiment predicts the power consumption according to the driving tendencies of the user using the vehicle, and predicts the vehicle's range (possible driving range) based on the predicted power consumption (power consumption rate) and the remaining power of the battery 141.

[0010] The measurement sensor 11 functions as one of the element measurement units of this disclosure and measures elements related to the vehicle's energy consumption. Specifically, the measurement sensor 11 includes at least a wheel speed sensor 111. The wheel speed sensor 111 functions as an element measurement unit of this disclosure, detects the rotational speed of the vehicle's wheels, and outputs the measured wheel speed as a wheel speed signal to the controller 20. In addition, the measurement sensor 11 may include various sensors that measure the vehicle's driving state and the surrounding conditions of the vehicle. For example, the measurement sensor 11 that measures the driving state may include an acceleration sensor, a position detection sensor, a torque sensor, a yaw rate sensor, etc. It may also include sensors that detect the driver's input, such as an accelerator position sensor and a brake position sensor. In this embodiment, the wheel speed measured by the wheel speed sensor 111 is described as one of the elements related to energy consumption, and the average vehicle speed calculated based on the wheel speed is described as the energy consumption element information of this disclosure. However, energy consumption element information may be calculated using other parameters measured by the acceleration sensor, torque sensor, accelerator position sensor, etc., as described above. The measuring sensor 11 may also be equipped with other sensors for measuring the surrounding conditions. Examples of sensors for measuring the surrounding conditions include image sensors (cameras), LiDAR, laser radar, and temperature sensors for measuring the temperature inside and outside the vehicle.

[0011] The position detection sensor 12 is a sensor that detects the current position of the vehicle. Examples of the position detection sensor 12 include a receiver that receives satellite signals from GNSS (Global Navigation Satellite System) to determine the current position.

[0012] The navigation device 13 searches for a route to the vehicle's destination based on destination information input by the user (passenger) and map information. The map information may be stored in the memory installed in the navigation device 13, stored in a separately provided map database, or obtained from an external device using a communication line such as the internet. The map information is data that represents the shape of roads using nodes and links, and various information related to roads may be associated with these nodes and links. For example, it is preferable to include road-related information such as the type of road (expressway, general road, etc.) and the legal speed limit of the road. The navigation device 13 also displays the map and the searched route on a display (not shown) installed in the vehicle cabin. At this time, the navigation device 13 may receive traffic information from a predetermined traffic management system via wireless communication and display the road congestion status (congestion, congestion, normal, etc.) on the display.

[0013] The battery unit 14 includes a battery 141 and a battery controller 142. The battery 141 is a secondary battery that stores power supplied to various parts of the vehicle. The battery controller 142 is a computer that manages the battery 141. The battery controller 142 includes, for example, a voltmeter and an ammeter connected to the battery 141, measures the state of charge (SOC), and outputs the measured state of the battery 141 to the controller 20.

[0014] The actuator 15 is a drive device that moves the vehicle based on the control of the controller 20, and includes a drive motor 151, an inverter 152 that outputs power supplied from the battery 141 to the drive motor 151, and a motor controller 153, etc. The drive motor 151 supplies the driving force to move the vehicle. The inverter 152 supplies power to the drive motor 151 based on the control of the motor controller 153. The motor controller 153 drives the drive motor 151 by adjusting the voltage at the inverter 152 based on the speed command from the controller 20. The motor controller 153 also measures the power used to move the vehicle. This power is one of the elements related to the vehicle's energy consumption, and the motor controller 153 also functions as an element measuring unit in this disclosure.

[0015] Although not shown in the diagram, the actuator 15 may also include a brake hydraulic actuator and a steering angle actuator. The brake hydraulic actuator controls the brake hydraulic force based on a braking control command from the controller 20. In the case of an electric vehicle that does not have a hydraulic booster, an electric booster may be used. The steering angle actuator controls the steering angle of the steering wheels based on a steering angle control command from the controller 20.

[0016] The air conditioner 16 adjusts the temperature inside the vehicle. The air conditioner 16 adjusts the temperature inside the vehicle based on, for example, the temperature inside the vehicle measured by a temperature sensor, the temperature outside the vehicle, and a target temperature set by the user. Other auxiliary equipment 17 are various electric devices other than the air conditioner 16 installed in the vehicle. Examples of other auxiliary equipment 17 include electric power steering, a coolant pump for the cooling system, lighting devices such as headlights, wipers, audio equipment, etc.

[0017] The controller 20 is a computer that controls the vehicle. The controller 20 is composed of a storage device such as semiconductor memory and an arithmetic circuit such as a CPU (Central Processing Unit). The controller 20 realizes various functions by having the arithmetic circuit read and execute a program stored in the storage device. Figure 2 is a block diagram showing the functional configuration of the controller 20. Specifically, as shown in Figure 2, the controller 20 functions as a power consumption element learning unit 21, an air conditioner power acquisition unit 22, an auxiliary power acquisition unit 23, a battery loss power acquisition unit 24, a power consumption calculation unit 25, a battery remaining capacity calculation unit 26, and a driving range calculation unit 27. Here, we show an example in which the functional configurations of the power consumption element learning unit 21, air conditioner power acquisition unit 22, auxiliary power acquisition unit 23, battery loss power acquisition unit 24, power consumption calculation unit 25, battery remaining capacity calculation unit 26, and driving range calculation unit 27 are realized by the arithmetic circuit of the controller 20 executing a program, but some or all of these may be realized by individual hardware configurations.

[0018] The energy consumption element learning unit 21 acquires energy consumption element information based on the energy consumption elements of the vehicle measured by the element measurement unit and stores it in the storage device. The energy consumption element learning unit 21 also uses machine learning to learn the stored energy consumption element information and calculates a learned value. Here, the energy consumption element learning unit acquires the energy consumption element information and driving time for one run of the vehicle and calculates the learned value. One run of the vehicle refers to the run of the vehicle from when the vehicle's power is switched on until the vehicle travels a predetermined distance or more and the vehicle's power is switched off. Hereafter, one run of the vehicle will be referred to as one trip. In this embodiment, the elements related to energy consumption are the vehicle speed and the power supplied to the drive motor 151 used to run the vehicle, and the average vehicle speed and driving power for one trip are exemplified as energy consumption element information.

[0019] Specifically, the power consumption factor learning unit 21 functions as an average vehicle speed acquisition unit 211, an average vehicle speed learning unit 212, a traveling power acquisition unit 213, and a traveling power learning unit 214. The average vehicle speed acquisition unit 211 acquires an average vehicle speed in one trip. Specifically, the average vehicle speed acquisition unit 211 measures the travel distance of the vehicle in one trip based on the number of wheel rotations detected by the wheel speed sensor 111 and the circumferential length of the wheel, and calculates the average vehicle speed, which is power consumption factor information, based on the travel distance and travel time. The average vehicle speed calculated for each trip is accumulated in a storage device.

[0020] The average vehicle speed learning unit 212 performs weighted moving average processing on the average vehicle speed based on travel time as shown in the following formula (1) to calculate a learned vehicle speed value, and stores the learned vehicle speed value in the storage device. In formula (1), let V be the learned vehicle speed value Ｌ (km / h), and let T be the travel time in the i-th trip ｉ (h), and let V be the average vehicle speed in the i-th trip ｉ (km / h). Although the number of data used in machine learning is not particularly limited, it is preferable to use data related to trips within the most recent two weeks, for example.

[0021]

[0022] The traveling power acquisition unit 213 acquires, as power consumption factor information, the power consumed for traveling of the vehicle in one trip (traveling power) every time the vehicle travels. The traveling power consumed in one trip can be detected, for example, by monitoring the power output from an inverter to a drive motor, and is acquired from the motor controller 153. The traveling power for each trip is accumulated in a storage device. Further, the traveling power learning unit 214 performs weighted moving average processing on traveling power based on travel time as shown in the following formula (2) to calculate a learned traveling power value, and stores the learned traveling power value in the storage device. In formula (2), let E be the learned traveling power value Ｌ (kW), and let E be the traveling power in the i-th trip ｉ (kW). Although the number of data used in machine learning is not particularly limited, it is preferable to use data related to trips within the most recent two weeks, similarly to the learned vehicle speed value, for example.

[0023]

[0024] An air conditioner power acquisition unit 22 measures the amount of power used by an air conditioner 16. An auxiliary machine power acquisition unit 23 measures the amount of power used by other auxiliary machines 17. A battery loss power acquisition unit 24 acquires the internal power loss amount of a battery 141. For example, battery loss map data indicating the relationship among outside air temperature, SOC, and the power loss amount of the battery 141 is stored in a storage device in advance. Accordingly, the battery loss power acquisition unit 24 can calculate the internal power loss amount of the battery 141 corresponding to the outside air temperature measured by a temperature sensor and the SOC detected by a battery controller 142.

[0025] A power consumption calculation unit 25 calculates power consumption based on the vehicle speed learning value, driving power learning value, air conditioner power amount, auxiliary machine power amount, and internal power loss amount of the battery 141 stored in the storage device. The power consumption E ｆ (km / kWh) is calculated by setting the air conditioner power amount as E ａ (kW), the auxiliary machine power amount as E ｅ (kW), and the internal power loss amount of the battery 141 as ΔE ｂ (kW), for example, according to the following formula (3).

[0026]

[0027] Note that, although it is preferable that the number of data used for calculating the vehicle speed learning value and the driving power learning value is calculated based on trips over a predetermined period (e.g., two weeks) or a predetermined number or more of trips, for a newly purchased vehicle or the like, the learning accuracy of the vehicle speed learning value and the driving power learning value may be low. Therefore, when the learning period of the vehicle speed learning value and the driving power learning value is less than a predetermined period, or when the number of data (number of trips) used for learning the vehicle speed learning value and the driving power learning value is less than a predetermined value, the power consumption calculation unit 25 may use initial data values instead of the vehicle speed learning value and the driving power learning value.

[0028] As initial data values, for example, predetermined performance index values ​​can be used. The performance index values ​​can be, for example, catalog fuel consumption obtained by running the vehicle under predetermined conditions through catalog tests. Alternatively, statistical values ​​calculated using big data may be used as initial data values. In this case, the average vehicle speed, driving time, and driving power for each trip of multiple users are collected by a server device that is connected to the vehicle via a communication line such as the Internet. Then, based on the average vehicle speed, driving time, and driving power collected by the server device, vehicle speed statistics and driving power statistics are calculated. The method for calculating vehicle speed statistics and driving power statistics is the same as for the vehicle speed learning values ​​and driving power learning values ​​described above, but differs in that the data source is the vehicles of multiple users, that is, it is calculated using big data.

[0029] Even when using the initial data values ​​described above, as the number of times the user drives the vehicle (number of trips) increases, user-specific learned values ​​are calculated. For example, Figure 3 shows an example of the changes in the power consumption calculated by the power consumption calculation unit 25 when catalog power consumption is used as the initial data value, and the changes in the vehicle speed learned value and driving power learned value learned for each trip. The example shown in Figure 3 is a simulation result for a user who frequently drives on highways, and it can be seen that even if the initial value is far from the user's specific average actual power consumption, it will converge to within ±10% of the average actual power consumption by learning a predetermined number of data points (14 trips in this example) or a predetermined period of time (2 weeks in this example).

[0030] Furthermore, the energy consumption calculation unit 25 of this embodiment sets a rate limit on the calculated energy consumption. This suppresses abrupt changes in the calculated energy consumption.

[0031] The battery remaining capacity calculation unit 26 calculates the remaining charge of the battery 141. The remaining charge of the battery 141 can be calculated from the State of Charge (%) input from the battery controller 142 and the maximum capacity (kWh) of the battery 141.

[0032] The range calculation unit 27 calculates the remaining charge E of the battery 141. ｂ(kWh) and the electricity consumption E calculated by the electricity consumption calculation unit 25 ｆ Based on (km / kWh), the driving range DTE (km) is calculated as DTE = E ｂ ×E ｆ It is calculated by [method].

[0033] [Method for Predicting Driving Range] Next, the method for predicting driving range in the vehicle of this embodiment will be described. Figure 4 is a flowchart showing the method for predicting driving range in this embodiment. First, the driving range calculated when the vehicle's power is turned on will be described. In this embodiment, when the vehicle's power is turned on, the power consumption calculation unit 25 of the controller 20 determines whether the number of past power consumption element information data (average vehicle speed, driving power) stored in the storage device is greater than or equal to a predetermined value (step S1). As mentioned above, instead of determining whether the number of data is greater than or equal to a predetermined value, it may be determined whether the period during which the power consumption element information was acquired is greater than or equal to a predetermined period (for example, two weeks or more).

[0034] If the result in step S1 is NO, the energy consumption element learning unit 21 determines whether or not there is data value for energy consumption element information (step S2). In step S2, if energy consumption element information has never been acquired in the past, or if the energy consumption element information acquired in the past is from more than a predetermined period ago (the data is old), then the result in step S2 is NO.

[0035] If NO is determined in step S2, the energy consumption element learning unit 21 cannot calculate the learned values ​​(vehicle speed learned value, driving power learned value), so it calculates the energy consumption using the initial data values. This involves acquiring other energy amounts other than the power related to the vehicle's operation (power supplied to the drive motor) (step S3). Specifically, the air conditioner power acquisition unit 22, the auxiliary equipment power acquisition unit 23, and the battery loss power acquisition unit 24 acquire the air conditioner power amount, other auxiliary equipment power amount, and the internal power loss amount in the battery 141 as other energy amounts, respectively. Then, the energy consumption calculation unit 25 calculates the vehicle speed learned value V in equation (3). Ｌ and the learned value of the driving power E ＬInstead, the electric power consumption is calculated using initial data values ​​(for example, vehicle speed statistics and driving power statistics set based on the big data described above) (step S4). If catalog electric power consumption (km / kWh) is used as the initial data value, the electric power consumption calculation unit 25 only needs to read the catalog electric power consumption recorded in the storage device, in which case step S3 may be omitted.

[0036] If the determination in step S2 is YES, that is, if energy efficiency element information obtained from multiple trips has been accumulated, but the number of data points used to calculate the vehicle speed learning value and the driving power learning value is less than a predetermined value, the energy efficiency element learning unit 21 calculates the energy efficiency using the acquired energy efficiency element information and the initial data value. In this case, the energy efficiency element learning unit 21 calculates the learning value based on the acquired energy efficiency element information and equations (1) and (2) (step S5). Also, similar to step S3, the air conditioner power amount, other auxiliary equipment power amount, and the amount of internal power loss in the battery 141 are acquired as other power amounts (step S6). Next, the energy efficiency calculation unit 25 calculates the energy efficiency using the learning value calculated in step S5 and the initial data value (step S7). For example, the energy efficiency calculation unit 25 integrates the learning value calculated in step S5 and the initial data value at a ratio based on the number of data points (number of trips) of the energy efficiency element information. The integrated vehicle speed learning value is V Ｌ Initial data value of vehicle speed statistics V ｔ V Ｌ '=βV Ｌ + (1-β)V ｔ The learned value of the integrated driving power is calculated as E Ｌ Initial data value of driving power statistics E ｔ as E Ｌ ' = βE Ｌ + (1-β)E ｔ The calculation is performed as follows. The variable β is a value between 0 and 1, and it is preferable to increase it according to the number of data points (number of trips) of the energy consumption element information. This allows the influence of initial data values ​​to be eliminated as the number of data points increases, and an energy consumption personalized to the user's driving tendencies can be calculated.

[0037] If the answer in step S1 is YES, that is, if a sufficient amount of energy consumption element information has been collected, the energy consumption can be calculated without using initial data values. In this case, the energy consumption element learning unit 21 calculates a learned value based on equations (1) and (2) based on the acquired energy consumption element information (step S8). Also, similar to steps S3 and S6, the amount of electricity consumed by the air conditioner, the amount of electricity consumed by other auxiliary equipment, and the amount of internal power loss in the battery 141 are acquired (step S9). Then, the energy consumption calculation unit 25 calculates the energy consumption using the learned value calculated in step S8, for example, by equation (3) (step S10).

[0038] Once the energy consumption is calculated in step S4, step S7, or step S10, the driving range is calculated using the calculated energy consumption. Specifically, the battery remaining capacity calculation unit 26 obtains SOC information and the maximum capacity (kWh) of the battery 141 from the battery controller 142 and calculates the remaining charge of the battery 141 (step S11). Then, the driving range calculation unit 27 calculates the driving range DTE (km) based on the remaining charge of the battery 141 calculated in step S11 and the energy consumption calculated in step S4, step S7, or step S10 (step S12). The controller 20 then displays the calculated driving range on the in-cabin display. This allows the user to be guided to a driving range calculated based on personalized energy consumption according to the user's past driving trends.

[0039] [Effects of this Embodiment] This embodiment includes a power consumption element measurement unit that measures multiple elements related to the vehicle's power consumption while the vehicle is running, and a controller 20. Specifically, the power consumption element measurement unit includes a wheel speed sensor 111 that measures the rotation speed of the wheels, and a motor controller 153 that measures the power supplied to the drive motor when the vehicle is running. The controller 20 acquires the average vehicle speed per trip and the driving power as power consumption element information from the measured values ​​(wheel rotation speed and motor power), and calculates the power consumption for each user by machine learning this power consumption element information, and calculates the driving range based on the calculated power consumption. In this embodiment, by using power consumption element information based on multiple elements related to power consumption, it is possible to accurately calculate personalized power consumption according to the user's driving tendencies. Furthermore, when using actual measured data of power consumption itself, the accuracy of power consumption calculation and driving range calculation may deteriorate depending on the measurement accuracy of the SOC, but in this embodiment, it is possible to predict power consumption with high accuracy that does not depend on the estimation accuracy of the SOC, and the driving range can also be calculated with greater accuracy.

[0040] In this embodiment, the energy consumption element information includes the average vehicle speed in one trip. The controller 20 learns the average vehicle speed over multiple trips to calculate a vehicle speed learning value, and then uses this vehicle speed learning value to calculate the energy consumption. The vehicle speed in a vehicle is closely related to the power supplied to the drive motor. Therefore, by machine learning the average vehicle speed over multiple trips, the energy consumption corresponding to the user's driving tendencies can be calculated with high accuracy.

[0041] In this embodiment, the controller 20 calculates a vehicle speed learning value by performing a weighted moving average of the average vehicle speed with respect to the driving time. In this embodiment, as shown in equation (1), by performing a weighted moving average of the average vehicle speed with respect to the driving time, the average value of the vehicle speed in the user's driving tendencies is used as the vehicle speed learning value. This makes it possible to calculate the energy consumption corresponding to the user's driving tendencies with greater accuracy.

[0042] In this embodiment, the energy consumption element information includes the power consumption during one trip, and the controller 20 calculates a power consumption learning value based on the power consumption during multiple trips, and calculates the energy consumption based on the power consumption learning value. As described above, there is a close relationship between the power supplied to the drive motor 151 when the vehicle is running and the vehicle speed, and by machine learning the power consumption during multiple trips, the energy consumption corresponding to the user's driving tendencies can be calculated with high accuracy.

[0043] In this embodiment, the controller 20 calculates a learned value of driving power by performing a weighted moving average of the driving power with respect to the driving time. In this embodiment, as shown in equation (2), by performing a weighted moving average of the driving power with respect to the driving time, the average value of the driving power under the user's driving tendencies is used as the learned value of driving power. This makes it possible to calculate the energy consumption corresponding to the user's driving tendencies with greater accuracy.

[0044] In this embodiment, if the number of acquired energy consumption element data is less than or equal to a predetermined value, the controller 20 calculates the energy consumption using a preset initial data value and the energy consumption element information. This allows the controller to calculate the energy consumption even when there is insufficient energy consumption element information, based on catalog energy consumption as initial data or energy consumption element information of other vehicles stored on a predetermined server.

[0045] [Second Embodiment] Next, a second embodiment will be described. In the first embodiment described above, the average vehicle speed and driving power for each trip were acquired as energy efficiency element information, and the results of the moving average processing shown in equations (1) and (2) were used as the vehicle speed learned value and driving power learned value, respectively, to calculate the energy efficiency. In contrast, the second embodiment differs from the first embodiment in that the trip is further subdivided according to the vehicle's driving conditions to learn the energy efficiency element information. In the following description, components that have already been described will be denoted by the same reference numerals, and their explanations will be omitted or simplified.

[0046] The vehicle of the second embodiment has the same configuration as the first embodiment, and as shown in Figure 1, it includes a measuring sensor 11, a position detection sensor 12, a navigation device 13, a battery unit 14, an actuator 15, an air conditioner 16, other auxiliary equipment 17, a controller 20, etc. In addition, the controller 20 of this embodiment functions as a power consumption element learning unit 21, an air conditioner power acquisition unit 22, an auxiliary equipment power acquisition unit 23, a battery loss power acquisition unit 24, a power consumption calculation unit 25, a battery remaining capacity calculation unit 26, and a driving range calculation unit 27, etc., by having the calculation circuit read and execute a program stored in the memory device, similar to the first embodiment.

[0047] In this embodiment, the processing of the energy consumption element learning unit 21, the energy consumption calculation unit 25, and the driving range calculation unit 27 differs from that of the first embodiment. The energy consumption element learning unit 21 in this embodiment functions as an average vehicle speed acquisition unit 211, an average vehicle speed learning unit 212, a driving power acquisition unit 213, and a driving power learning unit 214, similar to the first embodiment, but the average vehicle speed acquisition unit 211 and the driving power acquisition unit 213 acquire energy consumption element information for each situation. Figure 5 is a diagram showing the method of acquiring energy consumption element information in this embodiment. In this embodiment, energy consumption element information (average vehicle speed, driving power) is acquired separately for each day of the week and time of day, and situation identification information that identifies each situation is added and stored in a storage device. Alternatively, the corresponding energy consumption element information may be stored in a database separated for multiple situations, as shown in Figure 5. Accordingly, the average vehicle speed learning unit 212 calculates the vehicle speed learning value for each day of the week and time of day. For example, in the example in Figure 5, the vehicle speed learning value for Monday morning, Monday afternoon, and Monday evening is calculated. The driving power acquisition unit 213 works similarly, acquiring the driving power for each day of the week and time period, and the driving power learning unit 214 calculates the driving power learning value for each day of the week and time period. For example, in the example in Figure 5, the driving power learning value for Monday morning, Monday afternoon, and Monday evening is calculated.

[0048] Figure 5 shows an example where energy consumption element information is acquired for each day of the week and time of day, and a learned value is calculated for each. However, as energy consumption element information for each situation, energy consumption element information may also be acquired for various other situations, such as for each type of road or each level of road congestion, and a combination of these may also be used. For example, when acquiring energy consumption element information for each type of road, the average vehicle speed acquisition unit 211 and the driving power acquisition unit 213 can identify the type of road (general road, expressway, suburban road, city road, etc.) and the legal speed limit of the road based on the current position detected by the position detection sensor 12 and road-related information associated with the road of the vehicle's current position in the map information. Therefore, the average vehicle speed acquisition unit 211 and the driving power acquisition unit 213 can acquire energy consumption element information such as average speed and driving power for each type of road. Therefore, the average vehicle speed learning unit 212 and the driving power learning unit 214 can calculate the vehicle speed learned value and driving power learned value for each type of road from this energy consumption element information for each type of road. The method for calculating each learned value is the same as in the first embodiment, for example, using equations (1) and (2).

[0049] Furthermore, when acquiring energy consumption element information for each road congestion level, the average vehicle speed acquisition unit 211 and the driving power acquisition unit 213 can acquire road congestion levels based on road traffic information received via wireless communication, and can acquire the congestion level of the road the vehicle is traveling on from the current location detected by the position detection sensor 12 and the road of the vehicle's current location in the map information. Therefore, the average vehicle speed acquisition unit 211 and the driving power acquisition unit 213 can acquire energy consumption element information such as average speed and driving power for each road congestion level. As a result, the average vehicle speed learning unit 212 and the driving power learning unit 214 can calculate vehicle speed learning values ​​and driving power learning values ​​for each road congestion level from this energy consumption element information for each road congestion level.

[0050] In this embodiment, the energy consumption calculation unit 25 calculates the energy consumption based on the learned values ​​calculated separately for each of the above-described situations. In other words, the energy consumption calculation unit 25 calculates the energy consumption based on the learned values ​​learned for each day of the week and each time period. The method for calculating the energy consumption is the same as in the first embodiment, for example, using formula (3). The same applies when learned values ​​have been calculated for each type of road and each level of road congestion; the energy consumption calculation unit 25 calculates the energy consumption for each type of road and for each level of road congestion, respectively.

[0051] The cruising range calculation unit 27 predicts the cruising range based on the power consumption according to each situation. In other words, when the vehicle's power is turned on, the cruising range calculation unit 27 identifies the day of the week and time of day from the internal clock built into the controller 20, and calculates the cruising range using the power consumption calculated corresponding to the identified day of the week and time of day. The method for calculating the cruising range is the same as in the first embodiment, for example, the cruising range DTE (km) is calculated as DTE = E ｂ ×E ｆ It is calculated by [method].

[0052] Furthermore, if the electricity consumption for each type of road and the electricity consumption for each level of road congestion are calculated, the cruising range calculation unit 27 may calculate the cruising range based on the target driving route set by the user operating the navigation device 13 before driving the vehicle. For example, if the target driving route includes a highway, the amount of electricity consumed on the highway may be estimated using the distance of the highway section based on the target driving route and the electricity consumption calculated for the highway. The cruising range may then be calculated based on the remaining amount of electricity obtained by subtracting the estimated amount of electricity consumed on the highway from the remaining charge of the battery 141, and the electricity consumption on the general road. Also, if the received traffic information determines that the target driving route will pass through a congested or jammed section, the amount of electricity consumed in the congested and jammed sections may be estimated using the distance of the congested and jammed sections and the electricity consumption calculated for the congested and jammed sections, respectively. The cruising range may then be calculated based on the remaining amount of electricity obtained by subtracting the estimated amount of electricity consumed in the congested and jammed sections from the remaining charge of the battery 141, and the electricity consumption on the normal section.

[0053] The method for predicting the driving range in the vehicle of this embodiment is substantially the same as that of the first embodiment, and the driving range is predicted by the same process as the flowchart shown in Figure 4. In this case, in the calculation of the learned value in steps S5 and S8, the energy consumption element learning unit 21 reads energy consumption element information corresponding to the vehicle's driving conditions from the storage device and calculates the learned value corresponding to the driving conditions. For example, if the vehicle starts driving on Monday morning, the learned value is calculated based on the energy consumption element information corresponding to Monday morning. Also, if a target driving route is set, the learned value for each section is calculated from the energy consumption element information for each type of road included in the target driving route and for each congestion condition based on the received road information.

[0054] Furthermore, in steps S7 and S10, the energy consumption calculation unit 25 calculates the energy consumption using the learned values ​​for each driving condition calculated in steps S5 and S8. For example, if the vehicle starts driving on Monday morning, the energy consumption is calculated from the learned values ​​calculated for Monday morning. Also, if a target driving route is set, and the target driving route includes roads other than general roads, or if there are sections with congestion other than normal sections, the energy consumption is calculated for each of these road types, and for each congested or crowded section.

[0055] Then, in step S12, the cruising range calculation unit 27 calculates the cruising range based on the electricity consumption for each driving condition calculated in steps S7 and S10. For example, if the vehicle starts driving on Monday morning, the cruising range is calculated based on the electricity consumption calculated for Monday morning. Also, if the target driving route includes roads other than general roads, or if there are sections with congestion other than normal sections, the amount of electricity consumed on roads other than general roads and the amount of electricity consumed on roads other than normal sections are subtracted from the remaining charge of the battery 141, and the cruising range is calculated assuming that the vehicle is driving on normal sections of general roads with the remaining amount of electricity.

[0056] In this embodiment, the controller 20 acquires and uses machine learning to obtain energy consumption element information for each day and time (by day of the week or time of day), calculates the energy consumption for each day and time, and calculates the driving range for each day and time. This makes it possible to calculate the driving range appropriate to the day and time the user drives the vehicle, even if the vehicle's driving tendencies change from day to day depending on the user's habits. In other words, depending on the user's lifestyle, the user's driving tendencies may change from day to day, for example, by traveling on highways in the morning on weekdays, traveling on general roads in the afternoon on weekdays, and not using the vehicle on weekends. Even in such cases, the controller 20 can calculate the optimal energy consumption and driving range corresponding to the user's driving tendencies by identifying the day and time when the vehicle's power is switched on.

[0057] In this embodiment, the controller 20 may acquire and machine-learn energy consumption element information for each type of road, calculate the energy consumption for each type of road, and calculate the driving range for each type of road. This allows the system to calculate the driving range according to the type of road the user is driving on. Furthermore, when calculating the energy consumption and driving range for each day and time based on energy consumption element information for each day and time, it is necessary that the same driving tendencies are repeated at regular intervals due to the user's lifestyle, etc. In contrast, by learning the energy consumption element information for each type of road, it is possible to calculate appropriate energy consumption and driving range even when the user's driving tendencies are not based on their lifestyle. For example, when a user goes on a trip, their driving tendencies may deviate from their usual routine, but even in such cases, the system can guide them to an appropriate driving range along their target route.

[0058] In this embodiment, the controller 20 may acquire and machine-learn energy consumption element information for each road congestion level, calculate the energy consumption for each road congestion level, and calculate the driving range for each road congestion level. This allows the user to calculate a driving range that is appropriate to the congestion level of the road they are traveling on. Furthermore, by learning the energy consumption element information for each road congestion level, it is possible to calculate appropriate energy consumption and driving range even when unpredictable traffic jams or congestion occur due to road construction, accidents, etc.

[0059] [Modifications] The present invention is not limited to the embodiments described above, and also includes the following modifications to the extent that the objectives of the present invention can be achieved. For example, in the first embodiment described above, when calculating the power consumption using the initial data value and the learned value, the corrected vehicle speed learned value is V Ｌ Initial data value of vehicle speed statistics V ｔ V Ｌ '=βV Ｌ + (1-β)V ｔ The corrected driving power learning value is calculated as E Ｌ Initial data value of driving power statistics E ｔ as E Ｌ ' = βE Ｌ + (1-β)E ｔ The calculation was performed as described above, but is not limited to this. For example, the energy consumption element learning unit 21 may use the initial data values ​​(vehicle speed statistics, driving power statistics) as one of the average vehicle speed, driving power, and driving time obtained in one trip, and calculate the vehicle speed learning value and driving power learning value using equations (1) and (2), and the energy consumption calculation unit 25 may use the vehicle speed learning value and driving power learning value to calculate the energy consumption.

[0060] Furthermore, if the initial data value is the catalog fuel consumption, the fuel consumption calculation unit 25 may calculate the fuel consumption based on the fuel consumption elements using the learned value calculated from the fuel consumption element information stored in the memory device (a number of data points less than a predetermined number), and then calculate the fuel consumption by integrating the fuel consumption based on the fuel consumption elements with the catalog fuel consumption. In this case, the integrated fuel consumption is E １ E ２ , catalog electricity consumption E ３ As E ３ = βE １ + (1-β)E ２ It may also be calculated by the following method. The variable β is the same as above, and by changing it according to, for example, the number of data points (number of trips) of the energy consumption element information, personalized energy consumption can be calculated.

[0061] In the first embodiment described above, average vehicle speed, driving time, and driving power were given as examples of energy efficiency element information, but the embodiment is not limited to these. For example, the motor torque of the drive motor may be measured as an element related to energy efficiency, and the average motor torque may be calculated as energy efficiency element information. In the first embodiment, the energy efficiency related to the driving of the vehicle was calculated based on the vehicle speed learning value and driving power learning value calculated as energy efficiency element information, and the power consumption of the air conditioner 16 and the power consumption of other auxiliary equipment 17 were added to that energy efficiency to obtain the final energy efficiency. The power consumption of the air conditioner 16 and the power consumption of other auxiliary equipment 17 may be used as elements related to energy efficiency, and the average power consumption of the air conditioner 16 and the average power consumption of other auxiliary equipment 17 may be calculated by machine learning and added to the energy efficiency related to the driving of the vehicle based on the vehicle speed learning value and driving power learning value.

[0062] In the second embodiment described above, we explained that we calculate learned values ​​for each date and time by machine learning each of the energy consumption element information for each date and time, calculate learned values ​​for each road type by machine learning each of the energy consumption element information for each road type, and calculate learned values ​​for each road congestion level by machine learning each of the energy consumption element information for each road congestion level. However, learned values ​​may also be calculated for combinations of these. For example, we may calculate learned values ​​for when driving on a highway and through a congested section on a Monday morning, and learned values ​​for when driving on a highway on a Monday morning but not through a congested section, and then calculate the energy consumption.

[0063] 11...Measurement sensor, 12...Position detection sensor, 13...Navigation device, 14...Battery unit, 15...Actuator, 16...Air conditioner, 17...Other auxiliary equipment, 20...Controller, 21...Energy consumption element learning unit, 22...Air conditioner power acquisition unit, 23...Auxiliary equipment power acquisition unit, 24...Battery loss power acquisition unit, 25...Energy consumption calculation unit, 26...Battery remaining capacity calculation unit, 27...Driving range calculation unit, 111...Wheel speed sensor, 141...Battery, 142...Battery controller, 151...Drive motor, 152...Inverter, 153...Motor controller, 211...Average vehicle speed acquisition unit, 212...Average vehicle speed learning unit, 213...Driving power acquisition unit, 214...Driving power learning unit.

Claims

1. A method for predicting the driving range of an electric vehicle using a computer, wherein the computer acquires energy efficiency element information for a plurality of elements related to the vehicle's energy efficiency, calculates the energy efficiency for each user by machine learning the energy efficiency element information, and calculates the driving range based on the calculated energy efficiency.

2. The energy consumption element information includes the average vehicle speed in at least one trip, and the computer calculates a vehicle speed learning value based on the average vehicle speed over multiple trips, and calculates the energy consumption using the vehicle speed learning value. The method for predicting driving range according to claim 1.

3. The method for predicting driving range according to claim 2, wherein the energy consumption element information further includes the driving time in one trip, and the computer calculates the vehicle speed learned value by performing a weighted moving average process on the average vehicle speed with respect to the driving time.

4. The energy consumption element information includes the power consumed during at least one run, and the computer calculates a user-specific power consumption learning value based on the power consumed during multiple runs for each user, and calculates the energy consumption based on the power consumption learning value. The range prediction method according to any one of claims 1 to 3.

5. The method for predicting driving range according to claim 4, wherein the power consumption element information further includes the driving time in one trip, and the computer calculates the driving power learning value by performing a weighted moving average process on the driving power with respect to the driving time.

6. The method for predicting driving range according to any one of claims 1 to 5, wherein the computer acquires the power consumption element information for each date and time the vehicle is driven, performs machine learning to calculate the power consumption for each date and time, and calculates the driving range for each date and time.

7. The method for predicting driving range according to any one of claims 1 to 6, wherein the computer acquires the fuel consumption element information for each type of road traveled by the vehicle, performs machine learning to calculate the fuel consumption for each type of road, and calculates the driving range for each type of road.

8. The method for predicting driving range according to any one of claims 1 to 7, wherein the computer acquires the electricity consumption element information for each traffic congestion condition on the road the vehicle has traveled on, performs machine learning to calculate the electricity consumption for each traffic congestion condition on the road, and calculates the driving range for each traffic congestion condition.

9. The method for predicting driving range according to any one of claims 1 to 8, wherein the computer calculates the fuel consumption using a preset initial data value and the fuel consumption element information when the number of data points of the acquired fuel consumption element information is less than or equal to a predetermined value.

10. A range prediction device for predicting the range of an electric vehicle, comprising: an element measuring unit that measures a plurality of elements related to the vehicle's energy consumption while the vehicle is running; and a controller, wherein the controller acquires energy consumption element information for the plurality of elements related to the vehicle's energy consumption based on the elements measured by the element measuring unit, calculates the energy consumption for each user by machine learning the energy consumption element information, and calculates the range based on the calculated energy consumption.

Images