Electricity prediction method, device, equipment, storage medium and program product
By comprehensively considering the environmental information of the route to be traveled, the user driving characteristics and vehicle attributes, the use of the power prediction model improves the accuracy of power prediction, and solves the problems of high cost and low accuracy of power prediction in the prior art.
Patent Information
- Application Number
- CN202110673989.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-06-17
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2041-06-17
AI Technical Summary
The prior art has high experimental cost in power prediction and cannot predict the current user's driving style. When estimating the model based on historical statistical energy consumption, it only relies on historical data to affect the prediction accuracy.
By using the power prediction model to make comprehensive prediction based on the environmental information, user driving characteristics, vehicle attribute information and road information of the route to be traveled, the vehicle power information is predicted in combination with environmental information, user driving characteristics, vehicle information and road information.
It reduces the cost of power forecasting, improves the accuracy of power forecasting, helps users plan their itineraries reasonably, and avoids the dilemma of not being able to reach their destination due to battery life problems.
Smart Images

Figure CN113344278B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of computer technology, in particular to the field of intelligent transportation technology, and specifically to a power prediction method, device, equipment, storage medium and program product. Background Art
[0002] Pure electric vehicles output power based on electrical energy. Factors such as ambient temperature, driving speed, road conditions, driving style, and vehicle status during driving will affect power consumption.
[0003] The power prediction of pure electric vehicles is related to the energy replenishment of the vehicle's energy storage system, the formulation of user travel plans, and the maintenance of the vehicle. Therefore, it is crucial to accurately predict the power of pure electric vehicles during driving. Summary of the invention
[0004] The present disclosure provides a power prediction method, apparatus, device, storage medium and program product.
[0005] According to one aspect of the present disclosure, a method for power prediction is provided, the method comprising:
[0006] According to the current user's upcoming travel route, predict the travel environment information of the upcoming travel section in the upcoming travel route;
[0007] Determine the current user's driving characteristics;
[0008] The vehicle power information on the route to be traveled is determined according to the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled.
[0009] According to another aspect of the present disclosure, a power prediction device is provided, the device comprising:
[0010] A travel environment prediction module is used to predict the travel environment information of the travel section in the travel route according to the travel route of the current user;
[0011] A driving characteristics determination module, used to determine the driving characteristics of the current user;
[0012] The power information determination module is used to determine the vehicle power information on the section to be traveled based on the travel environment information of the section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the section to be traveled.
[0013] According to another aspect of the present disclosure, there is provided an electronic device, comprising:
[0014] at least one processor; and
[0015] a memory communicatively connected to the at least one processor; wherein,
[0016] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the power prediction method described in any one of the embodiments of the present disclosure.
[0017] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the power prediction method described in any one of the embodiments of the present disclosure.
[0018] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the power prediction method described in any one of the embodiments of the present disclosure is implemented.
[0019] According to the technology disclosed in the present invention, it is possible to jointly predict vehicle power information based on environmental information, user driving characteristics, road information and vehicle information.
[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.
[0022] Figure 1 is a schematic diagram of a power prediction method according to an embodiment of the present disclosure;
[0023] Figure 2 is a schematic diagram of a power prediction method according to an embodiment of the present disclosure;
[0024] Figure 3a is a schematic diagram of a power prediction method according to an embodiment of the present disclosure;
[0025] Figure 3b is a schematic diagram of a power prediction model structure according to an embodiment of the present disclosure;
[0026] Figure 3c is a schematic diagram of a sample expansion method according to an embodiment of the present disclosure;
[0027] Figure 4 is a schematic diagram of a power prediction device according to an embodiment of the present disclosure;
[0028] Figure 5It is a block diagram of an electronic device used to implement the power prediction method of the embodiment of the present disclosure. DETAILED DESCRIPTION
[0029] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.
[0030] Figure 1 is a schematic diagram of a power prediction method in an embodiment of the present disclosure. The technical solution of the embodiment of the present disclosure is applicable to the case where vehicle power information is predicted through environmental information, user driving characteristics, vehicle information and road information. The method can be executed by a power prediction device, which can be implemented by software and / or hardware, and can generally be integrated in an electronic device, such as a terminal device. The method of the embodiment of the present disclosure specifically includes the following:
[0031] S110 . Predicting travel environment information of a travel section in the travel route according to the travel route of the current user.
[0032] Among them, the route to be traveled is the route that the current user is about to drive. The route to be traveled can be the route that the user is about to drive obtained by the map application according to the departure location and target location input by the current user through path planning. For example, the route to be traveled is route AB with two endpoints respectively being the departure location A and the target location B.
[0033] The road section to be traveled can be understood as the road section obtained by dividing the route to be traveled according to the set granularity. In map applications, a route is often divided into multiple fine-grained route segments Link. Link is the smallest road unit with exactly the same attributes in the road network. Each Link contains rich attribute information, including coordinates, starting and ending angles, travel direction, road width, road length, road grade, number of lanes, number of traffic lights, speed limit, bump information, slope and other static attribute information, and can also include road conditions, traffic flow, average speed and estimated travel time and other dynamic attribute information. In the disclosed embodiment, a section of the road section to be traveled can correspond to the Link contained in the section to be traveled, and can also correspond to multiple Links in the section to be traveled.
[0034] The environmental information for the trip to be made is used to represent the environmental information of the section to be traveled at the time when the user arrives. The environmental information for the trip to be made is environmental information predicted based on the current user's travel route. By way of example, the environmental information for the trip to be made may include information such as the temperature, air pressure, wind speed, and wind direction of the area where the section to be traveled is located when the user arrives at the section to be traveled.
[0035] In the disclosed embodiment, the pure electric vehicle outputs power based on electric energy during driving. Environmental information such as temperature and wind speed on the route traveled by the vehicle is one of the important factors affecting the energy consumption of the pure electric vehicle. In order to accurately predict the amount of electricity of the vehicle during driving, after obtaining the departure location and target location input by the user, a route for the current user to travel can be planned according to the departure location, target location and road network data, and then the environmental information of the section to travel contained in the route to travel is predicted based on the route to travel of the current user. Specifically, the estimated arrival time of the user at the section to travel is first determined, and then the environmental information of the area where the section to travel is located at the estimated arrival time is obtained as the environmental information to travel. The information on the amount of electricity of the user driving the vehicle on the section to travel can be combined with the environmental information to travel, so as to improve the accuracy of the predicted amount of electricity. The environmental information to travel can include the temperature, wind speed and wind direction information of the section to travel at the estimated arrival time.
[0036] Exemplarily, based on the current user's travel route, road information of the section to be traveled in the route to be traveled is obtained from the map service, and the road information includes the estimated arrival time of the section to be traveled. Specifically, the estimated arrival time of the section to be traveled can be obtained from the ETA (Estimated Time of Arrival) service. Furthermore, the temperature, wind speed and direction information of the area where the section to be traveled is located at the estimated arrival time is obtained from the meteorological big data as the environmental information for the travel.
[0037] S120: Determine the driving characteristics of the current user.
[0038] Among them, the driving characteristics are used to describe the current user's vehicle driving style. The driving characteristics are characteristic data that can characterize the user's driving status. For example, the driving characteristics include information such as the user's driving speed, acceleration, number of sudden accelerations, and waiting time during the user's historical driving process.
[0039] In the disclosed embodiment, in addition to environmental information, the user's driving style also has a certain impact on the energy consumption of pure electric vehicles. For example, users who are accustomed to sudden acceleration and deceleration consume more power for the same mileage than users who are accustomed to smooth driving. In order to more accurately predict the vehicle power information on the road section to be driven, the historical trajectory data of the current user in the historical driving process can be obtained, wherein the historical trajectory data includes the current user's position at each time point in the historical driving process and the driving state at that time (for example, driving speed), and then the user's driving speed, acceleration, and frequency of sudden acceleration and deceleration during the historical travel process can be calculated based on the historical trajectory data.
[0040] S130, determining the vehicle power information on the route to be traveled according to the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled.
[0041] Among them, vehicle attribute information is used to describe the static and / or dynamic attribute information of the vehicle. Exemplarily, the vehicle attribute information includes static attribute information such as vehicle model, vehicle series, battery capacity, length, width, height, maximum speed, vehicle weight, etc. that is fixed when the vehicle leaves the factory, and also includes dynamic attribute information such as driving mode, energy recovery mode, air conditioning gear, air conditioning power, etc. during vehicle driving. Road information is used to characterize the information of the road dimension of the current user's upcoming travel section. Specifically, road information may include static road network data and dynamic road condition data. For example, static road network data may include road width, length, number of lanes, road grade, speed grade, number of traffic lights, number of left and right turns, number of U-turns, etc. Dynamic road condition data may include real-time road conditions, estimated arrival time, length of unobstructed or congested roads, and traffic volume, etc.
[0042] In the disclosed embodiment, in addition to environmental information and user driving characteristics, vehicle attribute information and road information are also important factors affecting vehicle power consumption. Therefore, in the power prediction process, the environment information of the road section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the road section to be traveled can be comprehensively considered. Exemplarily, the environment information of the road section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the road section to be traveled can be input into the pre-trained power prediction model to obtain the predicted power output by the power prediction model. Another exemplary method is to perform power prediction according to the pre-set power prediction rules, comprehensively consider the environment information of the road section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the road section to be traveled. Specifically, the vehicle power information on the road section to be traveled is predicted through the pre-set correlation between the environment information to be traveled and power consumption, the correlation between the driving characteristics of the user and power consumption, the correlation between the road information and power consumption, and the correlation between each vehicle attribute information and power consumption.
[0043] It is worth noting that the vehicle's dynamic attribute information needs to be uploaded in real time during driving. Ordinary vehicles do not have the conditions to upload dynamic attribute information in real time. Therefore, when predicting the power of ordinary vehicles, only the vehicle's static attribute information can be considered. For vehicles with the ability to upload data in real time, both the vehicle's static attribute information and dynamic attribute information can be considered at the same time.
[0044] The technical solution of the disclosed embodiment predicts the travel environment information of the section to be traveled in the travel route according to the current user's travel route, and determines the driving characteristics of the current user, and finally determines the vehicle power information on the travel route according to the travel environment information of the section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the section to be traveled. This overcomes the problems in the prior art of using an energy consumption model constructed based on the principle of automobile dynamics to predict power, such as high experimental cost and inability to predict the current user's driving style, and the problem of using a historical statistical energy consumption estimation model to predict current, which only relies on historical data to affect the prediction accuracy. By predicting vehicle power information through environmental information, user driving characteristics, vehicle information and road information, the power prediction cost can be reduced and the power prediction accuracy can be improved.
[0045] Figure 2 This is a schematic diagram of a power prediction method in an embodiment of the present disclosure, which is further refined on the basis of the above embodiment, and provides specific steps for predicting the travel environment information of the travel section in the travel route according to the current user's travel route, and specific steps for determining the driving characteristics of the current user. Figure 2A power prediction method provided by an embodiment of the present disclosure is described, including the following:
[0046] S210: predicting an estimated arrival time of a section of the route to be traveled according to the route to be traveled of the current user.
[0047] In the disclosed embodiment, in order to obtain the travel environment information of the road section to be traveled, it is necessary to first predict the time it takes for the user to travel to the road section to be traveled. The road information of the road section to be traveled can be obtained in the map application according to the current user's road section to be traveled. Specifically, the road information can include static road network data and dynamic road condition data. For example, the static road network data can include the width, length, number of lanes, road grade, speed grade, number of traffic lights, number of left and right turns, number of U-turns, etc. of the road. The dynamic road condition data can include real-time road conditions, estimated arrival time, length of unobstructed or congested roads, and traffic volume, etc. After obtaining the road information, the estimated arrival time in the dynamic road condition data can be obtained as the estimated arrival time of the current user traveling to the road section to be traveled.
[0048] For example, after obtaining the current user's travel route, the route can be divided into multiple travel sections according to the road network data in the map application, and then the estimated arrival time is obtained from the traffic data of the travel section to determine the travel environment information.
[0049] S220: Determine the travel environment information of the road section to be traveled according to the estimated arrival time and the association between the candidate time and the candidate environment information of the road section to be traveled.
[0050] In the disclosed embodiment, after the estimated arrival time of the section to be traveled is obtained, the environmental information of the section to be traveled is determined based on the correlation between the candidate times and the candidate environmental information of the section to be traveled. Specifically, the environmental information of the section to be traveled within a set time in the future can be requested from the meteorological big data server, and then the environmental information of the section to be traveled at the estimated arrival time can be obtained as the environmental information to be traveled, so as to perform power prediction based on the environmental information to be traveled, thereby improving the accuracy of power prediction.
[0051] For example, if the estimated arrival time of the road section to be traveled is 10 minutes, the temperature, wind speed and wind direction information of the road section to be traveled 10 minutes later can be obtained from the meteorological big data as the environmental information of the road section to be traveled.
[0052] Optionally, the environmental information to be traveled includes at least one of the air pressure, wind speed and wind direction of the road section to be traveled, and the temperature of the road section to be traveled;
[0053] The current vehicle attribute information includes at least one of the vehicle model, vehicle series, battery capacity, length, width, height, maximum driving speed, vehicle weight, real-time driving mode, real-time energy recovery mode and real-time air-conditioning power.
[0054] In this optional embodiment, specific content of the environmental information to be traveled and specific content of the current vehicle attribute information are provided, wherein the environmental information to be traveled includes at least one of the air pressure, wind speed and wind direction of the section to be traveled, and the temperature of the section to be traveled, and the current vehicle attribute information includes static attribute information such as vehicle model, vehicle series, battery capacity, length, width, height, maximum driving speed, vehicle weight, and dynamic attribute information such as real-time driving mode, real-time energy recovery mode and real-time air-conditioning power related to the dynamic settings of the user during driving. Considering the environmental information to be traveled and the vehicle attribute information in the process of power prediction can improve the accuracy of power prediction, facilitate users to reasonably plan their itineraries, and avoid the dilemma of users being unable to reach their destination due to battery life issues.
[0055] Among them, dynamic attribute information refers to attribute information that can be set independently by the user during driving and can be adjusted in real time. For example, the real-time driving mode in the dynamic attribute information may include comfort mode and economy mode, the energy recovery mode may be divided into three gears: high, medium and low, and the real-time air-conditioning power is related to the air-conditioning mode and gear adjusted by the user. For example, the air-conditioning mode includes cold air and hot air, and the air-conditioning gear may be divided into three gears: high, medium and low. In view of the different vehicle software settings of different car manufacturers, the specific contents contained in the above-mentioned driving mode, energy recovery mode and real-time air-conditioning power are only used for illustration. In actual applications, dynamic attribute information can be obtained based on the factory settings of different vehicles and the real-time settings of the vehicle by the user during driving.
[0056] S230: Determine the driving characteristics of the current user according to the historical travel trajectory of the current user.
[0057] In the disclosed embodiments, it is considered that the user's driving style also has a certain impact on the power consumption of pure electric vehicles. For example, users who are accustomed to sudden acceleration and deceleration consume more power when driving. Therefore, the historical travel trajectory of the current user in the historical driving process can be obtained, and the driving characteristics of the current user can be extracted based on the historical travel trajectory. Among them, the historical travel trajectory contains the position of the current user at each time point in the historical driving process and the driving state at that time (for example, driving speed), and then the user's driving speed, acceleration, and frequency of sudden acceleration and deceleration in the historical travel process can be calculated based on the historical trajectory data. The consideration of the user's driving characteristics in the power prediction process further improves the accuracy of power prediction.
[0058] S240, determining the vehicle power information on the route to be traveled according to the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled.
[0059] The technical solution of the disclosed embodiment predicts the estimated arrival time of the travel section in the travel route according to the current user's travel route, and determines the travel environment information of the travel section according to the estimated arrival time and the correlation between the candidate time and the candidate environment information of the travel section. Furthermore, the driving characteristics of the current user are determined according to the current user's historical travel trajectory. Finally, the vehicle power information on the travel route is determined according to the travel environment information of the travel section, the driving characteristics of the current user and the current vehicle attribute information. The vehicle power prediction can be performed by comprehensively considering the environmental information, vehicle attribute information, road information and user driving characteristics during the driving process, thereby improving the accuracy of power prediction.
[0060] Figure 3a This is a schematic diagram of a power prediction method in an embodiment of the present disclosure, which is further refined on the basis of the above embodiment, and provides specific steps for determining the vehicle power information on the route to be traveled according to the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled. Figure 3a A power prediction method provided by an embodiment of the present disclosure is described, including the following:
[0061] S310: predicting travel environment information of a section of the travel route according to the current user's travel route.
[0062] S320: Determine the driving characteristics of the current user.
[0063] S330, input the travel environment information of the travel section, the driving characteristics of the current user, the current vehicle attribute information and the road information of the travel section into the power prediction model, and obtain the vehicle power information on the travel route output by the power prediction model.
[0064] In an embodiment of the present disclosure, a power prediction method is provided, specifically, the travel environment information of the upcoming travel section, the driving characteristics of the current user, the current vehicle attribute information and the road information of the upcoming travel section are input into a power prediction model, and then the vehicle power information on the upcoming travel route output by the power prediction model is obtained. Power prediction is achieved by comprehensively considering the travel environment information, the driving characteristics of the current user, the current vehicle attribute information and the road information of the upcoming travel section. This is beneficial for users to make reasonable itinerary planning based on accurate power prediction results, and alleviate the mileage anxiety problem during long-distance travel in pure electric vehicles.
[0065] Optionally, before determining the vehicle power information on the route to be traveled according to the travel environment information of the route to be traveled, the current user's driving characteristics, the current vehicle attribute information and the road information of the route to be traveled, the method further includes:
[0066] According to the historical travel trajectory of the historical user, the actual environmental information of the historical travel section in the historical travel trajectory is determined;
[0067] According to the power information uploaded by historical vehicles, determine the actual power usage information of historical travel sections;
[0068] The actual environmental information of historical travel sections, the driving characteristics of historical users, the historical vehicle attribute information and the road information of historical travel sections are used as sample data, and the actual power usage information is used as the label corresponding to the sample data to train the power prediction model.
[0069] Considering that the data that the power prediction model needs to process include multiple dimensions, including the travel environment information, the current user's driving characteristics, the current vehicle attribute information, and the road information of the road section to be traveled, each dimension contains different types of data. The power prediction model adopts the Wide+Deep+Recurrent structure model, such as Figure 3bAs shown in Figure 1, the Wide+Deep+Recurrent structural model specifically includes the Wide module, the Deep module and the Recurrent module. The Wide module is essentially a generalized linear model, which is conducive to enhancing the memory capacity of the model. This module is used to learn the direct relationship between the input features and the power information, such as the length of the travel route, the direct relationship between the ambient temperature and the power. The input of the Wide module is dense continuous features and sparse discrete features. The Deep module is a deep neural network structure, which is conducive to enhancing the generalization ability of the model. This module vectorizes the discrete features through embedded coding, and then splices the vectorized discrete features and continuous features into the multi-layer network to learn the generalized features that do not directly appear in the input features, such as the abstract relationship between driving mode and power consumption. The input of the Deep module is a low-dimensional dense feature representation of dense continuous features and discrete features. The Wide module and the Deep module target the global information of the travel route, such as the length of the travel route, the number of traffic lights, the average temperature, the average speed level, the average number of lanes, the average estimated arrival data, the vehicle model identification, the driving mode, the air conditioning gear, etc. The Recurrent module is a recurrent neural network structure, including fully connected layers (FC) and LSTM (Long short-term memory) models. This module learns the relationship between Link-level features and power consumption. For example, based on the relationship between the temperature sequence, wind speed sequence, speed level sequence, lane number sequence, and road condition sequence of each section in the route and power consumption, the input of the Recurrent module is Link-level sequence features, or sequence features that are coarser-grained than Link.
[0070] Among them, continuous features can be road grades in road information and information such as temperature and wind speed in environmental information, and discrete features can be real-time road conditions in road information. For example, real-time road conditions can include "smooth", "slow moving", "congested", "extremely congested" and other states.
[0071] In this optional embodiment, a training process of an electricity training model is provided. First, the user's historical travel trajectory can be obtained according to the user's identification information, and the actual environmental information of the historical travel section in the historical travel trajectory can be determined according to the historical travel trajectory, wherein the historical travel trajectory includes the user's location and current status at each point in time, and the actual environmental information of the historical travel section can be determined according to the time when the user travels to the historical travel section. It should be noted that the user's historical travel trajectory information is obtained through various public, legal and compliant methods, for example, it can be obtained from a public data set, or obtained from the user after authorization by the user; further, the actual electricity usage information of the historical travel section is determined according to the electricity information uploaded by historical vehicles, wherein the electricity information It is collected by the vehicle's battery management system. The power collection frequency can be set to the same as the collection frequency of the trajectory points in the historical travel section. Specifically, the high-frequency signal is reduced to the same frequency as the low-frequency signal to achieve the matching of the trajectory points and the power information, so as to obtain the actual power usage information of the historical travel section; finally, the actual environmental information of the historical travel section, the driving characteristics of the historical users, the historical vehicle attribute information and the road information of the historical travel section are used as sample data, and the actual power usage information is used as the label corresponding to the sample data to train the power prediction model, so that the power prediction model can be used in the subsequent travel process, comprehensively processing the travel environment information, the current user's driving characteristics, the current vehicle attribute information and the road information of the travel section to be traveled, and finally outputting accurate power usage information.
[0072] Among them, after obtaining the actual environment information and the actual power usage information, in order to obtain the input matching the power prediction model of the Wide+Deep+Recurrent structure, the obtained information can be feature processed. For example, for the Wide module in the power prediction model, its input features are dense continuous features and sparse discrete features, among which the continuous features need to be normalized, discretized and missing value processed, and the discrete features are encoded, such as OneHot encoding and Label encoding; the sparse features in the Deep module do not need to be manually encoded, and low-dimensional dense feature representations can be directly learned through the embedding layer, and the dense continuous features also need to be normalized and other related processing; the input of the Recurrent module is a sequence feature, and for example, link-level equal-length sequence features can be input, such as temperature sequence, speed level sequence, lane number sequence, road condition sequence, and other variable-length sequence features can also be input, such as the user's driving feature sequence in the last 30 days. The target value of the power prediction model fitting is the vehicle power information.
[0073] Optionally, according to the historical travel trajectory of the historical user, the actual environment information of the historical travel section in the historical travel trajectory is determined, including:
[0074] According to the historical travel segments of historical users, determine the actual arrival time of the historical travel segments;
[0075] Determine the actual environmental information of the historical travel section based on the actual arrival time and the collection time of the candidate environmental information;
[0076] According to the power information uploaded by historical vehicles, the actual power usage information of the historical travel section is determined, including:
[0077] According to the historical travel segments of historical users, the actual arrival time of the historical travel segments is determined;
[0078] According to the actual arrival time and the collection time of the candidate power usage information, the actual power usage information of the historical travel section is determined.
[0079] In this optional embodiment, a specific method is provided for determining the actual environmental information of the historical travel sections in the historical travel trajectory according to the historical travel trajectory of the historical user. First, according to the historical travel sections of the historical user, the road information of the historical travel sections is obtained in the map data, and the road information includes the actual arrival time of the historical travel sections. Then, according to the actual arrival time and the collection time of the candidate environmental information, the actual environmental information matching the historical travel sections is determined, so as to use the actual environmental information as the model input to train the power prediction model, thereby improving the accuracy of the power usage information output by the model. This optional embodiment also provides a specific method for determining the actual power usage information of the historical travel sections according to the power information uploaded by the historical vehicles. First, according to the historical travel sections of the historical users, the actual arrival time of the historical travel sections is determined. Then, according to the actual arrival time and the collection time of the candidate power usage information, the historical arrival time is determined. The actual power usage information of the travel section can be obtained by matching the actual arrival time of the historical travel section with the collection time of the actual power usage information. In addition, when the collection frequency of the actual power usage information is low, in order to obtain the power usage information on the historical travel section, when the power collection frequency is different from the collection frequency of the trajectory points of the user's historical travel trajectory, the signal with a higher collection frequency can be first down-converted to a signal frequency with a lower collection frequency to achieve matching of the trajectory points on the historical travel trajectory and the power usage information, and then the historical travel trajectory can be matched with the road network information in the map data to obtain the corresponding relationship between the road network information and the actual power usage information, and finally the matching of the historical travel section and the power usage information can be achieved. Subsequently, the actual power usage information can be used as a sample label to train the power prediction model to improve the accuracy of the power usage information output by the model.
[0080] Optionally, after determining the actual power usage information of the historical travel section according to the power information uploaded by the historical vehicles, the method further includes:
[0081] The window clipping method is used to split and reorganize the historical travel trajectories to obtain at least two historical travel sections to expand the sample data.
[0082] In this optional embodiment, in order to obtain enough sample data and optimize the prediction effect of the model, after determining the actual power usage information of the historical travel section based on the power information uploaded by the historical vehicles, the window clipping method is used to process the historical travel trajectory to obtain at least two historical travel sections to expand the sample data. Specifically, window clipping is to clip a complete historical travel trajectory to obtain several trajectory segments. Since there is a corresponding power signal at the time corresponding to each trajectory point, there is a corresponding power value for the starting and ending points of each trajectory segment. Figure 3c As shown in the figure, a historical travel trajectory starts from S and ends at E. A, B, and C are certain clipping points on the historical travel trajectory. The selection of these clipping points needs to have sufficient distance and sampling time to ensure that the collected power usage information has undergone minimal changes in accuracy. For example, the accuracy of power collection setting of a certain vehicle system is 0.1 kWh, so the clipping points cannot be too close, and at least 0.1 kWh of electricity is required to be consumed between adjacent clipping points. When constructing samples based on the window clipping method, in addition to taking the starting and ending points SE of the complete trajectory as a sample data, SA, SB, SC, AB, AC, AE, BC, BE and other trajectory segments can also be used as new starting and ending points for trajectory truncation, thereby obtaining sample data such as actual power usage information, road information, and actual environmental information corresponding to the trajectory segment.
[0083] Since each sample contains both global route-level features and local link-level features, the route-level features describe the relationship between the statistical information of the route corresponding to the start and end points and the power consumption, for example, the impact of route length, number of traffic lights, average road grade and other features on power consumption; link-level features characterize the details of the road sections on the route, supplementing the impact of detailed features on power consumption. Global information and local information can be well integrated through window cropping. The smaller the window, the closer the global features are to the local features, and the larger the window, the less likely the global features are to be affected by details. The dynamic adjustment of the window size not only expands the samples, but also enriches the feature data through the integration of global and local information, which helps the accuracy and scalability of the power prediction model.
[0084] The technical solution of the disclosed embodiment predicts the travel environment information of the section to be traveled in the travel route according to the current user's travel route, and determines the driving characteristics of the current user. Finally, the travel environment information of the section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the section to be traveled are input into a power prediction model, and the vehicle power information on the route to be traveled output by the power prediction model is obtained. The vehicle power prediction can be performed by comprehensively considering the environmental information, vehicle attribute information, road information and user driving characteristics during the driving process, thereby further improving the accuracy of power prediction.
[0085] Figure 4 4 is a schematic diagram of the structure of a power prediction device in an embodiment of the present disclosure. The power prediction device 400 includes: a travel environment prediction module 410, a driving feature determination module 420 and a power information determination module 430.
[0086] The travel environment prediction module 410 is used to predict the travel environment information of the travel section in the travel route according to the travel route of the current user;
[0087] A driving characteristics determination module 420, for determining the driving characteristics of the current user;
[0088] The power information determination module 430 is used to determine the vehicle power information on the route to be traveled based on the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled.
[0089] The technical solution of the disclosed embodiment predicts the travel environment information of the travel section in the travel route according to the current user's travel route, determines the driving characteristics of the current user, and finally determines the vehicle power information on the travel route according to the travel environment information of the travel section, the driving characteristics of the current user, the current vehicle attribute information and the road information of the travel section. This overcomes the problems in the prior art of using an energy consumption model constructed based on the principle of automobile dynamics to predict power, such as high experimental cost and inability to predict the current user's driving style, and the problem of using a historical statistical energy consumption estimation model to predict current, which only relies on historical data to affect the prediction accuracy. By predicting vehicle power information through environmental information, user driving characteristics, road information and vehicle information, the power prediction cost can be reduced and the power prediction accuracy can be improved.
[0090] Optionally, the travel environment prediction module 410 includes:
[0091] An estimated arrival time prediction unit, used to predict an estimated arrival time of a section of a route to be traveled in accordance with a route to be traveled by a current user;
[0092] The travel environment prediction unit is used to determine the travel environment information of the road section to be traveled according to the estimated arrival time and the association relationship between the candidate time and the candidate environment information of the road section to be traveled.
[0093] Optionally, the driving characteristics determination module 420 includes:
[0094] The driving characteristic determination unit is used to determine the driving characteristics of the current user according to the historical travel trajectory of the current user.
[0095] Optionally, the travel environment information includes at least one of the air pressure, wind speed and wind direction of the road section to be traveled, and the temperature of the road section to be traveled;
[0096] The current vehicle attribute information includes at least one of the vehicle model, vehicle series, battery capacity, length, width, height, maximum driving speed, vehicle weight, real-time driving mode, real-time energy recovery mode and real-time air-conditioning power.
[0097] Optionally, the power information determination module 430 is specifically configured to:
[0098] The travel environment information of the travel section, the driving characteristics of the current user, the current vehicle attribute information and the road information of the travel section are input into the power prediction model to obtain the vehicle power information on the travel route output by the power prediction model.
[0099] Optionally, the power prediction device 400 further includes:
[0100] An actual environment information determination module is used to determine the actual environment information of the historical travel section in the historical travel trajectory according to the historical travel trajectory of the historical user before determining the vehicle power information on the route to be traveled according to the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled;
[0101] The actual power information determination module is used to determine the actual power usage information of the historical travel section based on the power information uploaded by historical vehicles;
[0102] The power prediction model training module is used to train the power prediction model by taking the actual environmental information of historical travel sections, the driving characteristics of historical users, the historical vehicle attribute information and the road information of historical travel sections as sample data, and taking the actual power usage information as the label corresponding to the sample data.
[0103] Optionally, the actual environment information determination module includes:
[0104] A first actual arrival time determination unit, configured to determine an actual arrival time of a historical travel section according to a historical travel section of a historical user;
[0105] An actual environment information determination unit, configured to determine the actual environment information of the historical travel section according to the actual arrival time and the collection time of the candidate environment information;
[0106] The actual power information determination module includes:
[0107] A second actual arrival time determination unit, configured to determine an actual arrival time of a historical travel section according to a historical travel section of a historical user;
[0108] The actual power usage information determining unit is used to determine the actual power usage information of the historical travel section according to the actual arrival time and the collection time of the candidate power usage information.
[0109] Optionally, the power prediction device 400 further includes:
[0110] The sample data expansion module is used to determine the actual power usage information of the historical travel section based on the power information uploaded by the historical vehicles, and then use the window clipping method to split and reorganize the historical travel trajectory to obtain at least two historical travel sections to expand the sample data.
[0111] The power prediction device provided in the embodiments of the present disclosure can execute the power prediction method provided in any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.
[0112] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.
[0113] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.
[0114] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.
[0115] like Figure 5 As shown, the device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.
[0116] A number of components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0117] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as the power prediction method. For example, in some embodiments, the power prediction method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the power prediction method described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform the power prediction method in any other appropriate manner (e.g., by means of firmware).
[0118] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0119] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0120] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0121] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).
[0122] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0123] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.
[0124] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0125] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. A method for predicting power consumption, comprising: According to the current user's upcoming travel route, predict the travel environment information of the upcoming travel section in the upcoming travel route; Determine the current user's driving characteristics; Inputting the travel environment information of the road section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the road section to be traveled into the power prediction model, and obtaining the vehicle power information on the route to be traveled output by the power prediction model; Among them, the power prediction model includes a Wide module, a Deep module and a Recurrent module; the Wide module is a generalized linear model, which is used to learn the direct relationship between input features and power information, and the input of the Wide module is dense continuous features and sparse discrete features; the Deep module is a deep neural network structure, which vectorizes discrete features through embedded coding, and splices the vectorized discrete features and continuous features into multiple hidden layer networks of the Deep module to learn generalized features that do not directly appear in the input features. The input of the Deep module is a low-dimensional dense feature representation of dense continuous features and discrete features; the Recurrent module is a recurrent neural network structure, which is used to learn the relationship between the Link-level features of the route segment and power consumption, and the input of the Recurrent module is Link-level sequence features; the outputs of the Wide module, Deep module and Recurrent module are respectively input into the fully connected layer, and the prediction results of the power prediction model are output through the connection layer.
2. The method according to claim 1, wherein: According to the current user's upcoming travel route, predicting the upcoming travel environment information of the upcoming travel section in the upcoming travel route, including: Predicting the estimated arrival time of the travel section in the travel route according to the current user's travel route; The travel environment information of the road section to be traveled is determined according to the estimated arrival time and the association relationship between the candidate time and the candidate environment information of the road section to be traveled.
3. The method according to claim 1, wherein: Determine the current user's driving characteristics, including: Determine the driving characteristics of the current user based on the historical travel trajectory of the current user.
4. The method according to claim 1, wherein: The travel environment information includes at least one of the air pressure, wind speed and wind direction of the road section to be traveled, and the temperature of the road section to be traveled; The current vehicle attribute information includes at least one of the vehicle model, vehicle series, battery capacity, length, width, height, maximum driving speed, vehicle weight, real-time driving mode, real-time energy recovery mode and real-time air-conditioning power.
5. The method according to claim 1, before inputting the travel environment information of the to-be-traveled section, the driving characteristics of the current user, the current vehicle attribute information and the road information of the to-be-traveled section into the power prediction model, and obtaining the vehicle power information on the to-be-traveled route output by the power prediction model, further comprises: According to the historical travel trajectory of the historical user, the actual environmental information of the historical travel section in the historical travel trajectory is determined; According to the power information uploaded by historical vehicles, determine the actual power usage information of historical travel sections; The actual environmental information of historical travel sections, the driving characteristics of historical users, the historical vehicle attribute information and the road information of historical travel sections are used as sample data, and the actual power usage information is used as the label corresponding to the sample data to train the power prediction model.
6. The method according to claim 5, wherein: According to the historical travel trajectory of the historical user, the actual environmental information of the historical travel section in the historical travel trajectory is determined, including: Determine the actual arrival time of the historical travel section according to the historical travel section of the historical user; Determine the actual environmental information of the historical travel section according to the actual arrival time and the collection time of the candidate environmental information; According to the power information uploaded by historical vehicles, the actual power usage information of the historical travel section is determined, including: Determine the actual arrival time of the historical travel section according to the historical travel section of the historical user; The actual power usage information of the historical travel section is determined according to the actual arrival time and the collection time of the candidate power usage information.
7. The method according to claim 5, after determining the actual power usage information of the historical travel section according to the power information uploaded by the historical vehicles, further comprises: The window clipping method is adopted to split and reorganize the historical travel trajectory to obtain at least two historical travel sections to expand the sample data.
8. A power prediction device, comprising: A travel environment prediction module is used to predict the travel environment information of the travel section in the travel route according to the travel route of the current user; A driving characteristics determination module, used to determine the driving characteristics of the current user; A power information determination module is used to input the travel environment information of the road section to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the road section to be traveled into the power prediction model, and obtain the vehicle power information on the route to be traveled output by the power prediction model; Among them, the power prediction model includes a Wide module, a Deep module and a Recurrent module; the Wide module is a generalized linear model, which is used to learn the direct relationship between input features and power information, and the input of the Wide module is dense continuous features and sparse discrete features; the Deep module is a deep neural network structure, which vectorizes discrete features through embedded coding, and splices the vectorized discrete features and continuous features into multiple hidden layer networks of the Deep module to learn generalized features that do not directly appear in the input features. The input of the Deep module is a low-dimensional dense feature representation of dense continuous features and discrete features; the Recurrent module is a recurrent neural network structure, which is used to learn the relationship between the Link-level features of the route segment and power consumption, and the input of the Recurrent module is the Link-level sequence features; the outputs of the Wide module, the Deep module and the Recurrent module are respectively input into the fully connected layer, and the prediction results of the power prediction model are output through the fully connected layer.
9. The device according to claim 8, wherein: The travel environment prediction module includes: An estimated arrival time prediction unit, used to predict an estimated arrival time of a section of a route to be traveled in accordance with a route to be traveled by a current user; The travel environment prediction unit is used to determine the travel environment information of the road section to be traveled according to the estimated arrival time and the association relationship between the candidate time and the candidate environment information of the road section to be traveled.
10. The device according to claim 8, wherein: The driving characteristics determination module comprises: The driving characteristic determination unit is used to determine the driving characteristics of the current user according to the historical travel trajectory of the current user.
11. The device according to claim 8, wherein: The travel environment information includes at least one of the air pressure, wind speed and wind direction of the road section to be traveled, and the temperature of the road section to be traveled; The current vehicle attribute information includes at least one of the vehicle model, vehicle series, battery capacity, length, width, height, maximum driving speed, vehicle weight, real-time driving mode, real-time energy recovery mode and real-time air-conditioning power.
12. The apparatus according to claim 8, further comprising: An actual environment information determination module is used to determine the actual environment information of the historical travel section in the historical travel trajectory according to the historical travel trajectory of the historical user before determining the vehicle power information on the route to be traveled according to the travel environment information of the route to be traveled, the driving characteristics of the current user, the current vehicle attribute information and the road information of the route to be traveled; The actual power information determination module is used to determine the actual power usage information of the historical travel section based on the power information uploaded by historical vehicles; The power prediction model training module is used to train the power prediction model by taking the actual environmental information of historical travel sections, the driving characteristics of historical users, the historical vehicle attribute information and the road information of historical travel sections as sample data, and taking the actual power usage information as the label corresponding to the sample data.
13. The device according to claim 12, wherein: The actual environment information determination module includes: A first actual arrival time determination unit, configured to determine an actual arrival time of a historical travel section according to a historical travel section of a historical user; An actual environment information determination unit, configured to determine the actual environment information of the historical travel section according to the actual arrival time and the collection time of the candidate environment information; The actual power information determination module includes: A second actual arrival time determination unit, configured to determine an actual arrival time of a historical travel section according to a historical travel section of a historical user; The actual power usage information determining unit is used to determine the actual power usage information of the historical travel section according to the actual arrival time and the collection time of the candidate power usage information.
14. The apparatus according to claim 12, further comprising: The sample data expansion module is used to determine the actual power usage information of the historical travel section based on the power information uploaded by the historical vehicles, and then use the window clipping method to split and reorganize the historical travel trajectory to obtain at least two historical travel sections to expand the sample data.
15. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the power prediction method according to any one of claims 1 to 7.
16. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable the computer to execute the power prediction method according to any one of claims 1-7.
17. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the power prediction method according to any one of claims 1 to 7.
Citation Information
Patent Citations
Method and system for predicting remaining driving mileage of electric automobile
CN104442825A
Method and device for predicating energy consumption in vehicle travel
CN110675619A
