Energy prediction device
By dividing the driving path into multiple divided paths and constructing a targeted consumption prediction model, the problem of low energy consumption prediction accuracy in the mixed existing paths of various traffic states in the prior art is solved, and higher prediction accuracy and accuracy.
Patent Information
- Application Number
- CN202410566724.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-05-11
- Filing Date
- 2024-05-09
- Publication Date
- 2024-11-12
AI Technical Summary
The prior art is difficult to improve the accuracy of vehicle energy consumption prediction in paths where various traffic states exist in a mixed manner.
By dividing the driving path into multiple divided paths, and a consumption prediction model is constructed for each divided path, and precise prediction is made using vehicle characteristics, driver information, environmental information and road status information.
The accuracy of vehicle energy consumption prediction in paths with mixed existence in various traffic states is improved, and the accuracy and reliability of predictions are enhanced.
Smart Images

Figure CN118940222A_ABST
Abstract
Description
Citation of related applications This application is based on Japanese Patent Application No. 2023-078311 filed on May 11, 2023, and the contents thereof are incorporated herein by reference. Technical Field
[0001] The present disclosure relates to an energy prediction device. Background Art
[0002] A technology for predicting the energy required for a vehicle to travel on an arbitrary path is known. In the invention described in JP2020-27432A described below, driving data including multiple feature quantities of a moving body is acquired based on information related to the moving state of the moving body. A prediction unit calculates a predicted value of the energy consumption of the moving body based on a first classification rule that corresponds multiple first conditions related to at least one of the multiple feature quantities to multiple first prediction models of energy consumption and the driving data.
[0004] When a vehicle is traveling on a path, various traffic conditions are mixed in the path, and changes in traffic conditions have different effects on energy consumption. For example, there are paths or path parts that are easily blocked due to increased traffic volume on rainy days, and on the other hand, there are paths or path parts where traffic volume does not change much even on rainy days. In the invention described in JP2020-27432A, since one prediction model is used to predict energy consumption for the entire path, it is difficult to improve the prediction accuracy in paths where various traffic conditions are mixed. Summary of the invention
[0005] An object of the present disclosure is to provide an energy prediction device that can improve the prediction accuracy of energy consumption in a vehicle even on a route where various traffic conditions are mixed.
[0006] The present invention discloses an energy prediction device, comprising: a path setting unit, which divides a driving path of a vehicle and sets divided paths; and a model construction unit, which constructs a consumption prediction model for predicting the energy consumption of the vehicle for each divided path.
[0007] According to the present disclosure, it is possible to provide an energy prediction device that can improve the prediction accuracy of energy consumption in a vehicle even on a route where various traffic conditions are mixed. BRIEF DESCRIPTION OF THE DRAWINGS
[0008] Figure 1 This is a block diagram for explaining the energy prediction device according to this embodiment. Figure 2 It is used to illustrate Figure 1 A flowchart of the information processing flow of the energy prediction device shown. Figure 3A to Figure 3C This is a diagram for explaining the setting of the segmentation path. Figure 4 This is a diagram for explaining the setting of the segmentation path. Figure 5 This is a diagram for explaining the setting of the segmentation path. Figure 6 This is a diagram for explaining the setting of the segmentation path. Figure 7 It is a diagram for explaining energy prediction. Figure 8A to Figure 8C It is a diagram for explaining energy prediction. Fig. 9 It is a diagram for explaining energy prediction. Fig.10 It is a diagram for explaining energy prediction. DETAILED DESCRIPTION
[0009] Hereinafter, the present embodiment will be described with reference to the accompanying drawings. In order to facilitate understanding of the description, the same components are denoted by the same reference numerals as much as possible in each figure, and repeated description is omitted.
[0010] Reference Figure 1 , the energy prediction device 10 of this embodiment is described. The energy prediction device 10 is configured to be able to send and receive information with the vehicle 30 via the network NW. The energy prediction device 10 is configured to be able to send and receive information with the vehicle characteristic storage unit 201, the driving route storage unit 202, the driver information storage unit 203, the environment information storage unit 204, and the state information storage unit 205. The energy prediction device 10 is configured as a computer including, as hardware, a computing unit such as a CPU (Central Processing Unit), a storage unit such as a RAM (Random Access Memory) or a ROM (Read Only Memory), and an interface unit for transmitting and receiving data.
[0012] The energy prediction device 10 includes a path setting unit 101 , a model building unit 102 , an information acquiring unit 103 , an energy prediction unit 104 , a notification unit 105 , and a storage unit 110 .
[0013] The route setting unit 101 is a part that divides the travel route of the vehicle and sets the divided routes. The setting of the divided routes will be described in detail later.
[0014] The model construction unit 102 is a part that constructs a consumption prediction model for predicting the energy consumption of the vehicle for each segmented route. The construction of the consumption prediction model will be described in detail later.
[0015] The information acquisition unit 103 is a part that acquires state information related to energy consumption when the vehicle travels on the travel route. The state information will be described in detail later.
[0016] The energy prediction unit 104 uses the consumption prediction model constructed for each segmented path to predict the energy consumption for each segmented path, and predicts the energy consumption portion in the travel path. As another embodiment, the energy prediction unit 104 also generates a vehicle speed pattern for each segmented path based on an index value related to the vehicle speed pattern, predicts the energy consumption based on the vehicle speed pattern for each segmented path, and predicts the energy consumption portion in the travel path. The prediction of energy consumption will be described in detail later.
[0017] The notification unit 105 is a part that notifies a predetermined notification destination of the energy consumption predicted by the energy prediction unit 104. The predetermined notification destination can be set arbitrarily, and the notification may be made to the vehicle 30 or to another notification destination.
[0018] The storage unit 110 is a part that stores the division paths set by the path setting unit 101. The storage unit 110 is a part that stores the consumption prediction model built by the model building unit 102. The storage unit 110 is a part that stores the state information acquired by the information acquisition unit 103.
[0019] The vehicle characteristic storage unit 201 is a part that stores characteristic information of various vehicles including the vehicle 30. The characteristic information includes all information necessary for predicting the energy consumption of the vehicle.
[0020] The travel route storage unit 202 is a part that stores route information related to the travel routes traveled or planned to be traveled by various vehicles including the vehicle 30. The route information is information on the travel route, and includes at least one of position information on each point, information on the slope, information on the width, information on the speed limit, information on the curvature of the curve, information on the position of the signal, information on the position of the bus stop, and information on the position of the intersection.
[0021] The driver information storage unit 203 is a part that stores driver information related to drivers who drive various vehicles including the vehicle 30. The driver information includes information indicating driver attributes and driver characteristics corresponding to the attributes.
[0022] The environmental information storage unit 204 is a part for storing environmental information related to the environment in which various vehicles including the vehicle 30 are traveling or scheduled to travel. The environmental information includes at least one of weather information and date and time information. The weather information includes at least one of a probability of precipitation and an amount of precipitation. The date and time information includes at least one of a time, a day of the week, a holiday, and a season.
[0023] The state information storage unit 205 is a part for storing state information related to energy consumption when the vehicle is traveling on the driving path. The state information includes at least one of driving state information, energy state information, and road state information. The driving state information is information for determining the mode in which the vehicle is traveling on the driving path. For example, the driving state information includes information indicating at least one of the stopping frequency, the maximum speed, the speed at steady state, the average speed, the acceleration, and the deceleration.
[0024] The energy status information is information for determining the energy required when the vehicle travels on the travel route. For example, the energy status information includes information indicating at least one of a fuel consumption rate, a power consumption rate, a SOC (State of Charge), and an inverter power.
[0025] The road state information is information for determining the road state that affects the energy required when the vehicle travels in the state of the travel path. For example, the road state information includes information indicating at least one of information related to the slope, information related to the location of the bus stop, information related to the location of the intersection, information related to the width, information related to the speed limit, information related to the location of the traffic light, information related to the curvature of the curve, and information related to the degree of congestion.
[0026] The energy prediction device 10 connects the vehicle 30 as a prediction target to the network NW, and the travel data of the vehicle 30 is transmitted to the energy prediction device 10 via the network NW and stored in the storage unit 110 .
[0027] The vehicle 30 includes a state detection unit 301, a vehicle speed detection unit 302, a communication unit 303, a battery ECU 304, and an inverter 305. The state detection unit 301 obtains the charge state SOC of the battery used for the travel of the vehicle 30 from the battery ECU. The state detection unit 301 obtains the current value from the inverter 305. The charge state SOC and the current value are sent to the energy prediction device 10 by the communication unit 303. The state detection unit 301 also detects the position information of the vehicle 30. The position information is sent to the energy prediction device 10 by the communication unit 303. The vehicle speed detection unit 302 detects the vehicle speed of the vehicle 30. The vehicle speed is sent to the energy prediction device 10 by the communication unit 303. In addition, in the present embodiment, the vehicle 30 is set as an electric vehicle, but the power source of the vehicle set as the prediction object by the energy prediction device 10 is not limited to an electric motor using a battery as a power generation source, and includes well-known gasoline engines, diesel engines, hybrid power composed of an electric motor and an internal combustion engine, and an electric motor using a fuel cell as a power generation source. All power sources can be set as the object.
[0028] Next, refer to Figure 2 , the information processing flow of the energy prediction device 10 is described. In step S001, the route setting unit 101 performs processing to divide the driving route. Figure 3A to Figure 3C , the driving route segmentation process is explained.
[0029] like Figure 3A As shown, the path setting unit 101 pre-divides the driving path and sets eight pre-divided paths SA1, SA2, SA3, SA4, SA5, SA6, SA7, and SA8. For the convenience of illustration and description, the pre-divided paths SA1, SA2, SA3, SA4, SA5, SA6, SA7, and SA8 are all of the same length, but they may also include pre-divided paths of different lengths. How to set the length of the pre-divided paths is arbitrary. In addition, the pre-divided paths SA1, SA2, SA3, SA4, SA5, SA6, SA7, and SA8 are eight pre-divided paths as an example, and as long as multiple or more than three divided paths are used as the object, the number can be appropriately determined.
[0030] The path setting unit 101 determines whether adjacent pre-divided paths satisfy a connection condition. Figure 3AIn the example, the path setting unit 101 determines whether the pre-divided path SA1 and the pre-divided path SA2 satisfy the combination condition. The path setting unit 101 determines whether the pre-divided path SA2 and the pre-divided path SA3 satisfy the combination condition. The path setting unit 101 determines whether the pre-divided path SA3 and the pre-divided path SA4 satisfy the combination condition. The path setting unit 101 determines whether the pre-divided path SA4 and the pre-divided path SA5 satisfy the combination condition. The path setting unit 101 determines whether the pre-divided path SA5 and the pre-divided path SA6 satisfy the combination condition. The path setting unit 101 determines whether the pre-divided path SA6 and the pre-divided path SA7 satisfy the combination condition. The path setting unit 101 determines whether the pre-divided path SA7 and the pre-divided path SA8 satisfy the combination condition. In addition, the combination condition will be described in detail later.
[0031] exist Figure 3A In the example of , the pre-divided path SA1 and the pre-divided path SA2 meet the combination condition. The pre-divided path SA2 and the pre-divided path SA3 do not meet the combination condition. The pre-divided path SA3 and the pre-divided path SA4 meet the combination condition. The pre-divided path SA4 and the pre-divided path SA5 meet the combination condition. The pre-divided path SA5 and the pre-divided path SA6 do not meet the combination condition. The pre-divided path SA6 and the pre-divided path SA7 do not meet the combination condition. The pre-divided path SA7 and the pre-divided path SA8 meet the combination condition.
[0032] In the above-described satisfying relationship of the combining conditions, the path setting unit 101 combines the pre-divided paths as follows: Figure 3B Specifically, the path setting unit 101 combines the pre-divided path SA1 and the pre-divided path SA2, and sets the divided path SB1. The path setting unit 101 combines the pre-divided path SA3, the pre-divided path SA4, and the pre-divided path SA5, and sets the divided path SB2. The path setting unit 101 does not combine the pre-divided path SA6 with other pre-divided paths, but sets the pre-divided path SA6 as the divided path SB3. The path setting unit 101 combines the pre-divided path SA7 and the pre-divided path SA8, and sets the divided path SB4.
[0033] The process of combining the pre-divided paths and setting the divided paths by the path setting unit 101 may also be performed by combining the pre-divided paths based on the combining condition. Figure 3A The pre-divided paths SA1, SA2, SA3, SA4, SA5, SA6, SA7, SA8 are shown and are set as Figure 3B The division paths SB1, SB2, SB3, and SB4 shown are completed.
[0034] The path setting unit 101 may also Figure 3BThe divided paths SB1, SB2, SB3, and SB4 shown are considered as pre-divided paths and further combined. Figure 3B , the division path SB1 and the division path SB2 regarded as the pre-division path meet the combination condition. The division path SB2 and the division path SB3 regarded as the pre-division path do not meet the combination condition. The division path SB3 and the division path SB4 regarded as the pre-division path meet the combination condition.
[0035] In the above-described satisfying relationship of the combining condition, the path setting unit 101 combines the divided paths SB1, SB2, SB3, and SB4 regarded as the pre-divided paths, as shown in FIG. Figure 3C Specifically, the path setting unit 101 combines the division path SB1 and the division path SB2 as the pre-division path, and sets the division path SC1. The path setting unit 101 combines the division path SB3 and the division path SB4 as the pre-division path, and sets the division path SC2.
[0036] Next, the combination condition is described. The combination condition is determined based on state information related to energy consumption when the vehicle is traveling on the driving path. The state information includes at least one of driving state information, energy state information, and road state information. The combination condition is determined based on at least one of the driving state information, energy state information, and road state information.
[0037] The route setting unit 101 divides the driving route into pre-divided routes, and performs processing of combining the pre-divided routes and setting the divided routes according to whether the adjacent pre-divided routes satisfy the combining condition. Since the combining condition is determined based on the state information related to the energy consumption of the vehicle when traveling on the driving route, whether the adjacent pre-divided routes can be combined is determined based on whether the state information of the adjacent pre-divided routes has a similarity that allows them to be combined.
[0038] For example, refer to Figure 4 , the case where the combination condition is judged based on the number of stops per unit distance in the same driving is explained. Figure 4 The number of stops per unit distance in each section is plotted when the vehicle travels in section 1 and section 2 in the same trip. Figure 4 By finding the correlation coefficient, the path setting unit 101 can determine whether the connection condition is satisfied. Alternatively, the connection condition may be determined using the coefficient of a regression curve or a straight line instead of the correlation coefficient.
[0039] For example, the combination condition may be determined using the correlation coefficient of the characteristic quantity of the driving state. For example, a threshold value may be set for a plurality of correlation coefficients, and the combination condition is satisfied when the correlation coefficient exceeds the threshold value. The plurality of correlation coefficients may also be weighted arbitrarily, and the combination condition may be determined based on the calculated value.
[0040] Reference Figure 5 , another example of judging the combined conditions is described. Figure 5 The example shown is an example of using cluster analysis (Japanese: cluster analysis) to determine the combination conditions. Figure 5 As shown, for driving 1, 2, and 3, the number of stops per unit distance is recorded in sections 1, 2, 3, 4, and 5, respectively. The route setting unit 101 determines whether section 1 and section 2 belong to the same cluster, for example, by a known cluster analysis method. If the route setting unit 101 determines that section 1 and section 2 belong to the same cluster, the combination condition is satisfied, and section 1 and section 2 are combined. If the route setting unit 101 determines that section 1 and section 2 do not belong to the same cluster, the combination condition is not satisfied, and section 1 and section 2 are not combined. The route setting unit 101 performs the same processing for other sections.
[0041] You can also not use Figure 5 Instead of clustering the single data shown, clustering is performed using a plurality of representative indices of each section obtained through one or more driving. Figure 6 In the example shown, a plurality of representative indicators such as "number of stops", "average vehicle speed", and "average road slope" are used. Figure 5 Similarly to the above description, for example, a known cluster analysis method is used to determine whether interval 1 and interval 2 belong to the same cluster. If the path setting unit 101 determines that interval 1 and interval 2 belong to the same cluster, the combination condition is satisfied, and interval 1 and interval 2 are combined. If the path setting unit 101 determines that interval 1 and interval 2 do not belong to the same cluster, the combination condition is not satisfied, and interval 1 and interval 2 are not combined. The path setting unit 101 may also use the results of multiple clusters to determine the combination condition.
[0042] Refer again Figure 2 , the information processing flow of the energy prediction device 10 is described. After the driving route segmentation process of step S001 is completed, the process proceeds to step S002. In step S002, the model construction unit 102 constructs a consumption prediction model.
[0043] The model building unit 102 builds a consumption prediction model for each segmented path. Figure 3BIn the example shown, the model building unit 102 builds consumption prediction models for each of the segmented paths SB1, SB2, SB3, and SB4. Figure 3C In the example shown, consumption prediction models are constructed for the segmented paths SC1 and SC2, respectively. The model construction unit 102 can construct different consumption prediction models for each segmented path, or can construct the same consumption prediction model. The model construction unit 102 constructs an appropriate consumption prediction model according to the conditions of each segmented path. The model construction unit 102 stores the constructed consumption prediction model in the storage unit 110.
[0044] An example of the consumption prediction model constructed by the model construction unit 102 is a consumption prediction model that directly predicts energy consumption. The consumption prediction model uses at least one of weather information, road state information, date and time information, and driver information acquired by the information acquisition unit 103 as an explanatory variable.
[0045] The weather information includes at least one of the probability of precipitation, the amount of precipitation, and the temperature. The road state information includes at least one of information related to the slope, information related to the location of the bus stop, information related to the location of the intersection, information related to the width, information related to the speed limit, information related to the location of the traffic light, information related to the curvature of the curve, and information related to the degree of congestion. The date and time information includes at least one of the time, day of the week, holiday, and season. The driver information includes information indicating the driver's attributes and the characteristics of the driver corresponding to the attributes.
[0046] The target variable of the consumption prediction model is a variable related to energy consumption. As a variable related to energy consumption, a quantity related to energy consumption is used. Specifically, as a variable related to energy consumption, for example, fuel consumption, power consumption, change in SOC, change in fuel consumption, and the cumulative value of inverter current are used.
[0047] As a modeling method for the consumption prediction model, any regression method such as linear regression, multiple regression, nonlinear regression, generalized linear model, support vector regression, Gaussian process regression, ensemble method, decision tree, neural network, etc. can be used. Fig. 9 An example of using a decision tree is given. Fig. 9 In the example shown, the target variable is set to power consumption, and the feature quantities used are set to time zones and the presence or absence of precipitation.
[0048] The learning method for the consumption prediction model can formulate the influence of each explanatory variable by constructing a linear polynomial such as formula (f01) using multiple regression analysis. In formula (f01), Y is a variable related to energy consumption, and βn represents the influence of each explanatory variable Xn. Y=β0+β1X1+β2X3+β3X3+β4X4+···(f01)
[0049] As a learning method for the consumption prediction model, other machine learning methods may also be used. As a learning method for the consumption prediction model, methods other than machine learning methods may also be used. For example, based on the information acquired by the information acquisition unit 103, any formula method such as using a constant value may be used.
[0050] As another example of the consumption prediction model constructed by the model construction unit 102, there is a consumption prediction model that predicts the vehicle speed waveform. In this case, the energy consumption can be calculated based on the physical model that takes into account the vehicle weight, energy efficiency of the drive system, driving resistance, etc. stored in the vehicle characteristics storage unit 201 and the predicted vehicle speed waveform.
[0051] When the model building unit 102 predicts the vehicle speed waveform, at least one of the weather information, road state information, date and time information, and driver information acquired by the information acquisition unit 103 is used as an explanatory variable.
[0052] The weather information includes at least one of the probability of precipitation, the amount of precipitation, and the temperature. The road state information includes information indicating at least one of information related to the slope, information related to the location of the bus stop, information related to the location of the intersection, information related to the width, information related to the speed limit, information related to the location of the traffic light, information related to the curvature of the curve, and information related to the degree of congestion. The date and time information includes at least one of the time, day of the week, holiday, and season. The driver information includes information indicating the driver's attributes and the characteristics of the driver corresponding to the attributes.
[0053] The target variable of the consumption prediction model may also be all indicators related to vehicle speed, such as the number of stops, average vehicle speed, and steady-state vehicle speed (the speed from the completion of acceleration to the start of deceleration). For example, the model building unit 102 uses different prediction models to estimate the number of stops and the steady-state vehicle speed. Figure 7 As shown in FIG. 1 , the model building unit 102 generates a trapezoidal vehicle speed pattern reflecting the estimated number of stops and the vehicle speed at steady state. The model building unit 102 sets acceleration and deceleration based on driver information and various vehicle characteristic information. The model building unit 102 arbitrarily sets the intervals between stop positions so as to control the driving distance within one section.
[0054] As a modeling method for the consumption prediction model, any regression method such as linear regression, multiple regression, nonlinear regression, generalized linear model, support vector regression, Gaussian process regression, ensemble method, decision tree, neural network, etc. can be used. Fig.10 An example of using a decision tree is shown. Fig.10In the example shown, the target variable is set to the average vehicle speed, and the feature quantities used are set to the time zone and the driver information.
[0055] The model building unit 102 may prepare a vehicle speed pattern corresponding to the average vehicle speed in advance, and generate a vehicle speed pattern by selecting a vehicle speed pattern suitable for the average vehicle speed. For example, if the average vehicle speed is 10 km / h to 15 km / h, the model building unit 102 selects Fig. 8A For example, if the average vehicle speed is 15 km / h to 25 km / h, the model building unit 102 selects Figure 8B For example, if the average vehicle speed is 25 km / h to 35 km / h, the model building unit 102 selects Figure 8C The model building unit 102 may determine the vehicle speed pattern based on a plurality of vehicle speed index values such as the number of stops and the average vehicle speed, for example.
[0056] The model building unit 102 may prepare a plurality of vehicle speed patterns and directly estimate which vehicle speed pattern to adopt by using a classification model based on an explanatory variable as a feature quantity. When the model building unit 102 adopts a classification model, the explanatory variable uses at least one of weather information, road state information, date and time information, and driver information acquired by the information acquisition unit 103.
[0057] The weather information includes at least one of the probability of precipitation, the amount of precipitation, and the temperature. The road state information includes information indicating at least one of information related to the slope, information related to the location of the bus stop, information related to the location of the intersection, information related to the width, information related to the speed limit, information related to the location of the traffic light, information related to the curvature of the curve, and information related to the degree of congestion. The date and time information includes at least one of the time, day of the week, holiday, and season. The driver information includes information indicating the driver's attributes and the characteristics of the driver corresponding to the attributes.
[0058] The target variable of the consumption prediction model is the classification of vehicle speed modes such as suburban mode, urban mode, and congestion mode. As the modeling method of the consumption prediction model, support vector machine, logistic regression, k-nearest neighbor method, neural network, simple Bayes classification, discriminant analysis, decision tree and other methods are used.
[0059] After the processing of step S002 is completed, the processing proceeds to step S003. In step S003, the energy prediction unit 104 determines whether to directly predict the energy consumption. In the case of directly predicting the energy consumption (step S003: Yes), the processing proceeds to step S004. In the case of not directly predicting the energy consumption (step S003: No), the processing proceeds to step S005.
[0060] In step S004, the energy prediction unit 104 predicts energy. The energy prediction unit 104 uses the consumption prediction model set for each segmented path (refer to formula f01) to predict the energy consumption for each segmented path. The energy prediction unit 104 can use any mathematical formula or mapping to convert the target variable into energy consumption. The energy prediction unit 104 sums up the energy consumption predicted for each segmented path and predicts the energy consumption of the entire driving path.
[0061] In step S005, the energy prediction unit 104 estimates the vehicle speed pattern. The energy prediction unit 104 estimates the vehicle speed pattern using the vehicle speed pattern generated by the model construction unit 102 in step S002. The energy prediction unit 104 uses the vehicle speed pattern generated by the model construction unit 102 unless there is any special additional processing.
[0062] After the process of step S005 is completed, the process proceeds to step S006. In step S006, the energy prediction unit 104 predicts the energy consumption for each divided route using the vehicle speed pattern V. The energy prediction unit 104 predicts the total energy consumption using, for example, the following equation (f02). Total energy consumption = driving energy + air conditioning energy + other energy (f02)
[0063] The running energy can be calculated by converting the running resistance obtained by the following formula (f03) into energy. Driving resistance = acceleration resistance + air resistance + rolling resistance + slope resistance (f03)
[0064] The acceleration resistance, air resistance, slope resistance, and rolling resistance can be obtained as follows. Acceleration resistance: Ma Air resistance: 0.5ρλAV 2 Slope resistance: Mgsinθ Rolling resistance: μMg M: Vehicle weight A: Frontal projection area λ: air resistance coefficient μ: rolling resistance coefficient θ: road slope V: Vehicle speed a: acceleration ρ: air density g: acceleration due to gravity
[0065] For vehicles with regenerative functions such as electric vehicles and hybrid vehicles, the regenerative energy can be determined in a manner related to the magnitude of the negative running resistance caused by deceleration. For other energy consumptions such as air conditioning energy consumption or auxiliary equipment, a predetermined value can be used, or a value corresponding to the weather or season such as temperature can be used.
[0066]
appendix
[0067]
Appendix 1
[0068] According to Appendix 1, since the consumption prediction model is constructed for each divided route after the travel route is divided, it is possible to construct a consumption prediction model more suitable for the divided route.
[0069]
Appendix 2
[0070] According to Appendix 2, since the segmented paths are set based on the similarity of the state information, for example, the segmented paths can be subdivided and set in sections where the state information is not similar and various conditions are mixed in the driving path, thereby improving the accuracy of the consumption prediction model. In addition, for example, if the sections where the state information is similar in the driving path continue to extend, the segmented paths can be set longer, which can make the consumption prediction model universal and reduce the calculation load.
[0071]
Appendix 3
[0072] According to Appendix 3, since the combination condition is determined based on the approximation of the pre-divided paths and the combination process is performed, it is easy to set the divided paths reflecting the approximation. Since it is sufficient to combine the pre-divided paths that satisfy the combination condition, the divided paths combined with the pre-divided paths can be regarded as pre-divided paths again to determine the combination condition.
[0073]
Appendix 4
[0074] According to Appendix 4, since the divided routes are set based on the approximation of at least one of the driving state information, the energy state information, and the road state information, the divided routes can be set while reflecting various aspects of the driving route.
[0075]
Appendix 5
[0076]
Appendix 6
[0077]
Appendix 7
[0078] The control unit and method thereof described in the present disclosure may also be implemented by a special-purpose computer, which is provided by constituting a processor and a memory, and the above-mentioned processor is programmed to perform one or more functions embodied by a computer program. Alternatively, the control unit and method thereof described in the present disclosure may be implemented by a special-purpose computer, which is provided by constituting a processor composed of one or more special-purpose hardware logic circuits. Alternatively, the control unit and method thereof described in the present disclosure may be implemented by one or more special-purpose computers, which are constituted by a combination of a processor and a memory programmed to perform one or more functions and a processor constituted by one or more hardware logic circuits. In addition, the computer program may also be stored in a computer-readable non-temporary tangible storage medium as an instruction executed by a computer.
[0079] Above, the present embodiment has been described with reference to specific examples. However, the present disclosure is not limited to these specific examples. Even if a person skilled in the art makes appropriate design changes to these specific examples, as long as the features of the present disclosure are included, they are included in the scope of the present disclosure. The various elements and their configurations, conditions, shapes, etc. included in the above-mentioned specific examples are not limited to the illustrated cases and can be appropriately changed. As long as no technical contradictions are generated, the various elements included in the above-mentioned specific examples can be appropriately combined and changed.
Claims
1. An energy prediction device, comprising: A path setting unit, the path setting unit divides the driving path of the vehicle and sets the divided paths; as well as A model construction unit is configured to construct a consumption prediction model for predicting energy consumption of the vehicle for each of the divisional routes.
2. The energy prediction device according to claim 1, characterized in that: It also includes an information acquisition unit, which acquires state information related to energy consumption of the vehicle when traveling on the driving path, The route setting unit divides the travel route based on the similarity of the state information and sets divided routes.
3. The energy prediction device according to claim 2, characterized in that: The route setting unit pre-divides the travel route to set a plurality of pre-divided routes, determines a combination condition based on the similarity of the state information in each of the pre-divided routes, and sets the divided routes by combining the pre-divided routes that satisfy the combination condition.
4. The energy prediction device according to claim 2, characterized in that: The state information includes at least one of driving state information for determining a driving state of the vehicle on the driving path, energy state information for determining energy required for the vehicle to travel on the driving path, and road state information for determining a road state that affects energy required for the vehicle to travel under the state of the driving path. The route setting unit sets the divided routes based on the approximation of at least one of the driving state information, the energy state information, and the road state information.
5. The energy prediction device according to claim 1, characterized in that: The model building unit builds the consumption prediction model as a learning model based on explanatory variables and target variables, The explanatory variable is at least one of weather information, road conditions, time information and driver information, The target variable is at least one of an index value related to a vehicle speed pattern and an index value related to consumed energy.
6. The energy prediction device according to claim 1, characterized in that: The system further includes an energy prediction unit that predicts energy consumption for each of the divided routes using the consumption prediction model constructed for each of the divided routes, and predicts energy consumption in the travel route.
7. The energy prediction device according to claim 5, characterized in that: An energy prediction unit is included. When the target variable is an index value related to the vehicle speed pattern, the energy prediction unit predicts the energy consumption in the vehicle speed pattern generated for each of the divided routes based on the index value related to the vehicle speed pattern, and predicts the energy consumption of the driving route.
8. An energy prediction method, the energy prediction method predicts the energy required for a vehicle to travel on a driving path, the energy prediction method comprising: a path setting step, in which the driving path is divided and the divided paths are set; as well as A model building step, in which a consumption prediction model for predicting the energy consumption of the vehicle is built for each of the divided routes.
9. A computer-readable non-transitory storage medium, The storage medium stores an energy prediction program. The energy prediction program is used to make the computer function as a path setting unit and a model building unit. The path setting unit divides the driving path of the vehicle and sets the divided paths. The model construction unit constructs a consumption prediction model for predicting energy consumption of a vehicle for each of the divisional routes.
