CN110909907A - Method and device for predicting fuel consumption of truck and storage medium - Google Patents
Method and device for predicting fuel consumption of truck and storage medium Download PDFInfo
- Publication number
- CN110909907A CN110909907A CN201811085613.4A CN201811085613A CN110909907A CN 110909907 A CN110909907 A CN 110909907A CN 201811085613 A CN201811085613 A CN 201811085613A CN 110909907 A CN110909907 A CN 110909907A
- Authority
- CN
- China
- Prior art keywords
- information
- prediction
- sample
- road
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000000446 fuel Substances 0.000 title claims abstract description 101
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 28
- 238000010801 machine learning Methods 0.000 claims abstract description 18
- 238000013507 mapping Methods 0.000 claims abstract description 18
- 230000010365 information processing Effects 0.000 claims abstract description 7
- 238000003066 decision tree Methods 0.000 claims description 93
- 238000004590 computer program Methods 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 2
- 239000003921 oil Substances 0.000 description 69
- 239000003981 vehicle Substances 0.000 description 23
- 238000007637 random forest analysis Methods 0.000 description 14
- 238000012549 training Methods 0.000 description 14
- 238000012360 testing method Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 11
- 230000001133 acceleration Effects 0.000 description 7
- 238000001914 filtration Methods 0.000 description 6
- 239000000463 material Substances 0.000 description 6
- 238000005070 sampling Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000012417 linear regression Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 239000000346 nonvolatile oil Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Abstract
The invention provides a method and a device for predicting the fuel consumption of a truck and a storage medium. The method comprises the following steps: obtaining forecast information according to TBox data collected by a remote information processing box TBox of the truck, wherein the forecast information comprises one or more of the following: road information, vehicle information, driving habit information; determining oil consumption information corresponding to the prediction information according to the prediction information and a prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information. The method and the device improve the accuracy of the predicted oil consumption information.
Description
Technical Field
The invention relates to the technical field of fuel consumption prediction, in particular to a method and a device for predicting fuel consumption of a truck and a storage medium.
Background
Under the requirements of energy conservation and emission reduction, a fuel consumption prediction mode for predicting the fuel consumption of a road is provided at present.
In the prior art, fixed fuel consumption coefficients are designed for different road information respectively, for example, a fixed fuel consumption coefficient corresponding to a type 1 road is 1, a fixed fuel consumption coefficient corresponding to a type 2 road is 1.01, a fixed fuel consumption coefficient corresponding to a straight road is 1, a fixed fuel consumption coefficient corresponding to a curved road is 1.1, and the like. When the oil consumption of a road needs to be predicted, specifically, according to a plurality of pieces of road information of the road and a plurality of fixed oil consumption coefficients corresponding to the plurality of pieces of road information, an oil consumption prediction result of the road is obtained through a calculation formula.
However, the conventional technology has the problem of low accuracy of the fuel consumption prediction result.
Disclosure of Invention
The invention provides a method and a device for predicting oil consumption of a truck and a storage medium, which are used for solving the problem of low accuracy of an oil consumption prediction result in the prior art.
In a first aspect, the present invention provides a fuel consumption prediction method, including:
obtaining forecast information according to TBox data collected by a remote information processing box TBox of the truck, wherein the forecast information comprises one or more of the following: road information, vehicle information, driving habit information;
determining oil consumption information corresponding to the prediction information according to the prediction information and a prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
Optionally, the prediction model includes at least one decision tree, where each decision tree in the at least one decision tree is used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information based on different decision trees are completely different or not completely the same;
the determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model comprises the following steps:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
Optionally, the TBox data includes position information of a current track point;
the obtaining the prediction information according to the TBox data comprises:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information according to the road sections and the corresponding relation between the road sections and the road information.
Optionally, the road information includes at least one of the following: road grade, road curvature, road material;
the vehicle information includes at least one of: speed, acceleration, load.
Optionally, the TBox data includes the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding accelerator pedal opening degree;
the obtaining the prediction information according to the TBox data comprises:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
Optionally, after determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further includes: and marking the corresponding relation between the road information and the oil consumption information in the map data.
Optionally, a plurality of paths exist between the departure place and the destination of the truck, and the prediction information includes prediction information of each of the plurality of paths;
after determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further comprises the following steps:
and determining a path with the minimum oil consumption between the departure place and the destination according to the oil consumption information corresponding to the prediction information of each road.
In a second aspect, the present invention provides a fuel consumption prediction device for a truck, including:
the information module is used for obtaining prediction information according to TBox data acquired by a remote information processing box TBox of the truck, and the prediction information comprises one or more of the following: road information, vehicle information, driving habit information;
the oil consumption determining module is used for determining oil consumption information corresponding to the prediction information according to the prediction information and the prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
Optionally, the prediction model includes at least one decision tree, where each decision tree in the at least one decision tree is used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information based on different decision trees are completely different or not completely the same;
the oil consumption determining module is specifically configured to:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
Optionally, the TBox data includes position information of a current track point;
the information module is specifically configured to:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information corresponding to the road section according to the road section and the corresponding relation between the different road sections and the road information.
Optionally, the TBox data includes the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding accelerator pedal opening degree;
the information module is specifically configured to:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
Optionally, the apparatus further comprises: a labeling module and/or a path determining module;
the marking module is used for marking the corresponding relation between the road information and the oil consumption information in map data;
the route determining module is configured to determine, according to fuel consumption information corresponding to prediction information of each of a plurality of roads, a route with a minimum fuel consumption between a departure location of the truck and a destination of the truck, where the plurality of routes exist between the departure location and the destination, and the prediction information includes prediction information of each of the plurality of routes.
In a third aspect, the present invention provides a fuel consumption prediction apparatus, comprising: a processor and a memory for storing computer instructions; the processor executes the computer instructions to perform the method of any of the first aspects described above.
In a fourth aspect, the present invention provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method of any of the first aspects described above.
The method, the device and the storage medium for predicting the fuel consumption of the truck obtain the prediction information according to the TBox data acquired by the TBox of the truck, determining fuel consumption information corresponding to the prediction information according to the prediction information and the prediction model, the prediction model is generated by adopting a machine learning algorithm to bring sample information, the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information, because the prediction model can embody the mapping relation between the sample prediction information indicated by the sample information and the sample oil consumption information and has the capability of predicting the oil consumption information with higher accuracy according to the prediction information, therefore, according to the prediction model and the prediction information, the determined oil consumption information corresponding to the prediction information has higher accuracy, and the accuracy of the predicted oil consumption information is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a flowchart of a first embodiment of a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 2 is a flowchart of a second method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 3 is a first schematic diagram illustrating a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 4A is a schematic diagram of a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 4B is a third schematic diagram of a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 5 is a fourth schematic diagram of a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 6 is a fifth schematic diagram of a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 7 is a sixth schematic diagram of a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 8 is a flowchart of a third method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 9 is a seventh schematic diagram illustrating a method for predicting fuel consumption of a truck according to an embodiment of the present invention;
fig. 10 is a schematic structural diagram of an embodiment of a fuel consumption prediction apparatus for a truck according to an embodiment of the present invention;
fig. 11 is a schematic structural diagram of an embodiment of a fuel consumption prediction apparatus for a truck according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a flowchart of a first embodiment of a method for predicting fuel consumption of a truck according to an embodiment of the present invention, where the method for predicting fuel consumption of a truck according to the embodiment may be executed by a navigation device or a server. As shown in fig. 1, the method of this embodiment may include:
In this step, the prediction information includes one or more of the following: road information, vehicle information, driving habit information. The road information may be any one or more of all information that can be used to represent road surface characteristics, and optionally, the road information may include a road gradient, a road curvature, a road material, and/or the like. The vehicle information may specifically be any one or more of all information that can be used to represent the vehicle characteristics, and optionally, may include speed, acceleration, load, and/or the like. The driving habit information may specifically be any one or more of all information that can be used to represent driving habit characteristics, and optionally, the driving habit information may include aggressive or robust information.
The TBox is an intelligent vehicle-mounted terminal in the car networking system, and directly communicates with a Controller Area Network (CAN) bus of a vehicle to obtain data (i.e., TBox data) of a car body state, a car condition and the like, and the data CAN be uploaded to a background of a car remote Service Provider (TSP). Furthermore, the TBox data covers a large number of types of vehicle-related data, and the prediction information can be obtained according to the TBox data.
And step 102, determining oil consumption information corresponding to the prediction information according to the prediction information and the prediction model.
In this step, the prediction model is a model generated by substituting sample information into a machine learning algorithm, and the sample information indicates a mapping relationship between sample prediction information and sample oil consumption information. The prediction model may specifically be any model that can be generated by a machine learning algorithm, and may specifically be, for example, a random forest model, a neural network model, or the like. The prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information. Therefore, the prediction model can embody the mapping relation between the sample prediction information indicated by the sample information and the sample oil consumption information, and has the capability of predicting the oil consumption information with higher accuracy according to the prediction information. Therefore, according to the prediction model and the prediction information, the determined fuel consumption information corresponding to the prediction information has high accuracy. Specifically, when the prediction information is used as the input of the prediction model, the output of the prediction model is the fuel consumption information corresponding to the prediction information. The fuel consumption information may specifically be a fuel consumption grade or a fuel consumption per hundred kilometers.
It should be noted that the present invention mainly analyzes the fuel consumption of a truck, considering that the load of the truck is generally large, the road has a large influence on the fuel consumption of the truck, and the road has a small influence on the fuel consumption of a car.
In the embodiment, the prediction information is obtained according to TBox data acquired by a TBox of a truck, the fuel consumption information corresponding to the prediction information is determined according to the prediction information and the prediction model, the prediction model is a model generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample fuel consumption information. In addition, as the TBox data contains more types of vehicle related data, the fuel consumption information can be determined according to the TBox data, and the accuracy of the fuel consumption information can be further improved.
Fig. 2 is a flowchart of a second embodiment of a method for predicting fuel consumption of a truck according to an embodiment of the present invention, and the method for predicting fuel consumption of a truck according to the embodiment of the present invention mainly describes a specific implementation manner when executed by a server on the basis of the embodiment shown in fig. 1. As shown in fig. 2, the method of this embodiment may include:
In this step, the sample information indicates a mapping relationship between the sample prediction information and the sample oil consumption information. The machine learning algorithm may be, for example, a neural network algorithm, a random forest algorithm, or the like. Optionally, step 201 may further include: and obtaining the sample information according to the TBox data. It should be noted that the TBox data is different from the TBox data in step 101, where the TBox data is used to obtain sample information, and the TBox data in step 101 is used to obtain prediction information.
It is to be understood that the sample prediction information is consistent with the prediction information, and optionally may include one or more of the following: sample road information, sample vehicle information, and sample driving habit information. And a mapping relation exists between the sample prediction information and the sample oil consumption information.
Optionally, the TBox data may include respective trace point information of a plurality of trace points, and the sample information may be determined according to the trace point information of each trace point. Usually, only the track point information of the vehicle at a track point is recorded in the TBox data, for example, one track point may be recorded every 30 seconds in the TBox data. And the mapping relation between the sample prediction information and the sample oil consumption information can be embodied through the track points. Namely, the sample prediction information determined according to one track point has a mapping relation with the sample oil consumption information determined according to the track point.
Optionally, the track point information may include position information, and the road information of each track point may be determined according to the track point information of each track point. Further optionally, the road section to which each track point belongs may be determined according to position information in the track point information of each track point, and the road information corresponding to the road section to which each track point belongs may be determined according to the road section to which each track point belongs and the pre-stored correspondence between different road sections and the road information. For example, a Hidden Markov Model (HMM) may be used to determine the road to which each trajectory belongs. Optionally, one trace point may correspond to one sample information, and the number of the sample information may be less than or equal to the number of the trace points.
Optionally, the track point information may further include instant information. The instantaneous information may include, for example, mileage, instantaneous fuel consumption, instantaneous speed, instantaneous accelerator pedal opening, instantaneous altitude, etc. In order to avoid the problem caused by inaccurate instant information, optionally, other information of a track point besides the road information may be obtained according to instant information of the track point and other track points (i.e. the following adjacent points) having a time difference with the track point within a preset time length range.
Optionally, the distance difference, the oil consumption difference, the average speed, the speed difference, the average accelerator pedal opening and other information of a track point can be obtained according to the instantaneous information of the track point and the adjacent point of the track point within the preset duration range of the time difference between the track point and the track point. Optionally, the mileage difference and the fuel consumption difference can be used for generating sample fuel consumption information; the average speed may be a speed of the sample vehicle information in the sample prediction information, and the speed difference and the time difference may be used to generate an acceleration of the vehicle information in the sample prediction information.
Alternatively, the elevation difference may be calculated according to the elevation of one track point and an adjacent point having a time difference with the track point within a preset range, and the road gradient of the road information in the sample prediction information may be generated according to the elevation difference and the time difference.
Assuming that the points that are earlier than the sampling point a0 and have a time difference with the sampling point a0 within a preset time range are a1-A5, the points that are later than the sampling point a0 and have a time difference with the sampling point a0 within a preset time range are A6-a10, and the times of the points a1 to a10 are shorter, the mileage difference may be specifically the mileage of a10 minus the mileage of a1, the fuel consumption difference may be specifically the fuel consumption of a10 minus the fuel consumption of a1, the speed difference may be specifically the speed of a10 minus the speed of a1, the average speed may be specifically the sum of the speeds a0 to a10 divided by 11, the average accelerator pedal opening may be specifically the sum of the accelerator pedal openings a0 to a10 divided by 11, and the altitude difference may be specifically the altitude of a10 minus the altitude of a 1.
Optionally, least squares linear regression may be performed on the mileage difference and the oil consumption difference to generate the sample oil consumption information. Specifically, as shown in fig. 3, the horizontal axis may be the mileage difference, and the vertical axis may be the fuel consumption difference, and then the coefficient of the linear regression may be the sample fuel consumption information. It should be noted that the specific manner of generating the acceleration according to the speed difference and the time difference and generating the gradient according to the altitude difference and the time difference may be similar to the sample oil consumption information, and will not be described herein again.
In order to avoid the influence of the start and stop of the vehicle on the fuel consumption, optionally, when the number of the adjacent points within the preset time range with the time difference of one track point is less than or equal to the preset number, the track point can be filtered, that is, the sample information is not determined based on the track point.
Further optionally, in order to ensure the validity and accuracy of the sample information obtained according to the TBox data, before obtaining the sample information according to the TBox data, the method may further include: and filtering the track points in the TBox data according to preset filtering conditions. Optionally, the preset filtering condition may include at least one of the following: reduced track points for truck range compared to previous track points; trace points with reduced oil consumption compared to previous trace points; the positioning precision (the higher the positioning precision, the more accurate the positioning can be represented) is greater than the track point with the preset precision; track points with speed less than or equal to a preset speed; and the time interval is less than or equal to the track point of the first time length or greater than or equal to the track point of the second time length. Assuming that the preset speed is 0 and the preset precision is 10, and the track formed by the track points in the TBox data before filtering is as shown in fig. 4A, after the track points in the TBox data are filtered according to the preset filtering condition, the obtained track formed by the track points in the filtered TBox data is as shown in fig. 4B. As can be seen from fig. 4A and 4B, the track formed by the track points in the filtered TBox data is clearer than that before filtering, and the purpose of denoising is achieved.
The present invention is not limited to the manner of acquiring the load and the sample driving habit information in the sample vehicle information. Optionally, the load in the sample vehicle information may also be obtained through TBox data. Optionally, the track point information may further include: the stepping times of the accelerator pedal can be further determined according to the stepping times of the accelerator pedal at each track point and the average opening degree of the accelerator pedal; alternatively, the sample driving habit information may be input by the user.
After obtaining the sample information, a machine learning algorithm may be employed to bring in the sample information to generate the predictive model. Alternatively, the machine learning algorithm may be a random forest algorithm. The random forest algorithm is a supervised learning algorithm and is an integration of a plurality of decision trees. For each of the decision trees, a bootstrap aggregation method (bagging) training may be adopted, specifically, a classifier may be constructed by selecting training sample information that is randomly put back, and a final prediction model may be obtained by merging the decision trees. Specifically, the sample information used for generating the prediction model may be divided into training sample information and test sample information, where the training sample information may be used for training the prediction model, and the test sample information may be used for testing the prediction model. It should be noted that the training sample information and the testing sample information may not be fixed, for example, in a round of training and testing process, sample information 1-sample information 103 may be training sample information, and sample information 104-sample information 120 may be testing sample information; in another round of training and testing, sample information 1-sample information 90, sample information 104-sample information 110 may be training sample information, sample information 91-sample information 103, and sample information 111-sample information 120 may be testing sample information.
The training process for the random forest algorithm model (i.e., the prediction model) may be as follows:
Wherein, a group of samples corresponds to a decision tree, ntree represents the number of decision trees forming a random forest.
It can be seen that the predictive model includes at least one decision tree.
And 2, applying a decision tree calculation method to each group of samples in the ntree group of samples to obtain ntree decision trees.
Specifically, m types of information may be randomly extracted without being replaced from n types of information, where n is equal to the number of types of information included in the same sample prediction information, and m is an integer greater than 0 and less than or equal to n. The gradient, curvature, material, driving habit and the like can be used as sample prediction information respectively. As shown in fig. 5, for each of the m types of information extracted from a set of samples: firstly, traversing the information value of the class information of the group of samples, cutting the group of samples into two parts, and calculating the splitting error; and then, judging whether the error is smaller than a threshold th, if so, judging that the current segmentation is the optimal segmentation and updating the th to be the error, otherwise, returning to the step of traversing the information value for execution. After traversing the information value of the information, dividing the sample information smaller than th into a left node and the rest into a right node according to th.
It should be noted that the random forest is a forest in which ntree decision trees are collected. When the result is predicted from the random forest, as shown in fig. 6, each decision tree in the ntree decision trees can obtain a predicted result, and it is assumed that the product of the predicted result obtained by the decision tree i and the normalized accuracy after the accuracy normalization of each decision tree is Pi(c | f), i equals 1, 2, 3 … … ntree, and the prediction results of the random forest can be obtained by summing the products of the normalized accuracy and the prediction results of each of the ntree decision trees. That is, the prediction result P (c | f) of the random forest may be equal to
Wherein c can be as shownThe fuel consumption information may be represented by f, which may represent input information for determining the fuel consumption information. It can be seen that the random forest algorithm can obtain more accurate and stable predictions by building and merging multiple decision trees together.
It can be seen that the prediction model comprises at least one decision tree, each decision tree being used for determining a decision tree prediction result based on at least one type of sample prediction information. Further, in order to further balance the deviation generated by each decision tree, the types of sample prediction information based on different decision trees are completely different or not completely the same.
After the random forest algorithm model is obtained through training, the accuracy of the model can be evaluated through testing sample information. Specifically, the accuracy of the model may be measured by analyzing a mean-square error (MSE) between an estimated value obtained based on the model (i.e., estimated fuel consumption information) and an actual value in the test sample information (i.e., sample fuel consumption information). For example, other information except for the sample fuel consumption information in the sample information 103 may be used as an input of the model, and the obtained output is an estimated value, and is subjected to mean square error calculation with the sample fuel consumption information in the sample information 103. The higher the accuracy of the model is, the smaller the mean square error between the estimated value obtained based on the model and the actual value in the test sample information is. When the accuracy of the model reaches a certain degree, the curve corresponding to the estimated value of each test sample and the curve corresponding to the actual value of each test sample may approximately coincide, for example, as shown in fig. 7.
When the random forest algorithm is adopted to bring sample information into the prediction model, the random forest algorithm balances the deviation generated by each decision tree, so that the problem of over-fitting or under-fitting can be avoided.
And step 202, obtaining prediction information according to TBox data acquired by the TBox of the truck.
In this step, similar to the sample prediction information, the prediction information may include one or more of the following: road information, vehicle information, driving habit information.
The specific manner of obtaining the prediction information according to the acquired TBox data is not limited in the present invention.
Optionally, the road information in the prediction information may be determined according to position information in trace point information of a trace point recorded last.
Optionally, the driving habit information can be determined according to the treading times and the opening degree of an accelerator pedal in the trace point information of each trace point of the trace point information within the preset time range.
And 203, determining oil consumption information corresponding to the prediction information according to the prediction information and the prediction model.
It should be noted that step 203 is similar to step 102, and is not described herein again.
Optionally, in order to enrich the map data, after determining the oil consumption information corresponding to the prediction information, the map data may further be marked with a corresponding relationship between the road information in the prediction information and the oil consumption information. For example, the fuel consumption information corresponding to the road information may be displayed on a road corresponding to the road information.
In the embodiment, the sample information is brought in by adopting a machine learning algorithm to generate the prediction model, and the fuel consumption information corresponding to the prediction information is determined according to the prediction model and the prediction information obtained according to the TBox data.
Fig. 8 is a flowchart of a third embodiment of a method for predicting fuel consumption of a truck according to an embodiment of the present invention, where the method for predicting fuel consumption of a truck according to the embodiment of the present invention is mainly described in a specific implementation manner when executed by a navigation device based on the embodiment shown in fig. 1. As shown in fig. 8, the method of this embodiment may include:
In this step, optionally, a prediction file including the prediction model sent by the server may be received, and the prediction model is obtained according to the prediction file. Further optionally, it may be determined whether the prediction file exists locally. When the prediction file does not exist, it may end; when the prediction file exists, step 802 can be performed as follows.
And step 802, obtaining respective prediction information of a plurality of paths between the departure place and the destination according to the Tbox data of the truck.
In this step, as shown in fig. 9, it is assumed that there are 3 routes between the departure point and the destination, which are route 1 to route 3. Alternatively, a path may be composed of one road or a plurality of roads. Optionally, the length of one road may be the entire length corresponding to one road name, or may also be a partial length in the entire length corresponding to one road name.
It is understood that the difference between the prediction information of each of the plurality of paths is mainly the difference between the road information.
And 803, determining fuel consumption information corresponding to the prediction information of the paths according to the prediction information and the prediction model of the paths.
In this step, the respective prediction information of the multiple paths may be input into the prediction model, so as to obtain the fuel consumption information corresponding to the respective prediction information of the multiple paths.
Corresponding to step 201, optionally, the prediction model includes at least one decision tree, where each decision tree in the at least one decision tree is used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information based on different decision trees are completely different or not completely the same. Correspondingly, determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, and the method comprises the following steps: obtaining decision tree prediction results of the decision trees according to the prediction information; and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
And step 804, determining the path with the minimum oil consumption between the departure place and the destination according to the oil consumption information corresponding to the prediction information of each of the plurality of paths.
In this step, it is assumed that the fuel consumption information is specifically fuel consumption per kilometer, and in fig. 9, the fuel consumption per kilometer of the path 1 is 7.0L, the fuel consumption per kilometer of the path 2 is 10.0L, the fuel consumption per kilometer of the path 3 is 7.5L, the path length of the path 1 is 10 kilometers, the path length of the path 2 is 6 kilometers, and the path length of the path 3 is 9 kilometers, and therefore, since 7.0 × 10 is greater than 7.5 × 9 and greater than 10.0 × 6, it can be determined that the path with the minimum fuel consumption between the departure place and the destination may be the path 2.
Similarly, in order to enrich the map data, after determining the fuel consumption information corresponding to the prediction information, the map data may also be marked with the corresponding relationship between the road information in the prediction information and the fuel consumption information. For example, the fuel consumption information corresponding to the road information is displayed on the road corresponding to the road information.
In this embodiment, prediction information of each of a plurality of paths between a departure place and a destination is obtained according to Tbox data of a truck by receiving a prediction model sent by a server, fuel consumption information corresponding to the prediction information of each of the plurality of paths is determined according to the prediction information of each of the plurality of paths and the prediction model, a path with the minimum fuel consumption between the departure place and the destination is determined according to the fuel consumption information corresponding to the prediction information of each of the plurality of paths, and a path with the minimum fuel consumption can be selected for a user.
Fig. 10 is a schematic structural diagram of an embodiment of a fuel consumption prediction apparatus for a truck according to an embodiment of the present invention, where the apparatus provided in this embodiment may be applied to the above method embodiment, the apparatus may be a navigation apparatus, or may be an apparatus, and when the apparatus is an apparatus, the apparatus may specifically be a server or a navigation apparatus. As shown in fig. 10, the fuel consumption prediction apparatus for a truck according to the present embodiment may include: an information module 1001 and a fuel consumption determination module 1002.
The information module 1001 is configured to obtain prediction information according to TBox data acquired by a telematics box TBox of a truck, where the prediction information includes one or more of the following: road information, vehicle information, driving habit information;
the oil consumption determining module 1002 is configured to determine oil consumption information corresponding to the prediction information according to the prediction information and the prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
Optionally, the prediction model includes at least one decision tree, where each decision tree in the at least one decision tree is used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information based on different decision trees are completely different or not completely the same;
the oil consumption determining module 1002 is specifically configured to:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
Optionally, the TBox data includes position information of a current track point;
the information module 1001 is specifically configured to:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information corresponding to the road section according to the road section and the corresponding relation between the different road sections and the road information.
Optionally, the road information includes at least one of the following: road grade, road curvature, road material.
Optionally, the vehicle information includes at least one of: speed, acceleration, load.
Optionally, the TBox data includes the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding accelerator pedal opening degree;
the information module 1001 is specifically configured to:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
Optionally, the device for predicting fuel consumption of a truck provided in this embodiment may further include: and the marking module 1003 is configured to mark the corresponding relationship between the road information and the oil consumption information in the map data.
Optionally, a plurality of paths exist between the departure place and the destination of the truck, and the prediction information includes prediction information of each of the plurality of paths;
the fuel consumption prediction apparatus for a truck according to the present embodiment may further include: and a path determining module 1004, configured to determine, according to the fuel consumption information corresponding to the prediction information of each of the multiple roads, a path with the minimum fuel consumption between the departure place and the destination.
The apparatus of this embodiment may be configured to implement the technical solutions of the embodiments shown in fig. 1 to fig. 3, and the implementation principles and technical effects are similar, which are not described herein again.
Fig. 11 is a schematic structural diagram of a second embodiment of a fuel consumption prediction apparatus for a truck according to an embodiment of the present invention, as shown in fig. 11, the fuel consumption prediction apparatus may include: a processor 1101 and a memory 1102 for storing computer instructions.
Wherein the processor 1101 executes the computer instructions to perform the following method:
obtaining forecast information according to TBox data collected by a remote information processing box TBox of the truck, wherein the forecast information comprises one or more of the following: road information, vehicle information, driving habit information;
determining oil consumption information corresponding to the prediction information according to the prediction information and a prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
Optionally, the prediction model includes at least one decision tree, where each decision tree in the at least one decision tree is used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information based on different decision trees are completely different or not completely the same;
the determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model comprises the following steps:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
Optionally, the TBox data includes position information of a current track point;
the obtaining the prediction information according to the TBox data comprises:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information corresponding to the road section according to the road section and the corresponding relation between the different road sections and the road information.
Optionally, the road information includes at least one of the following: road grade, road curvature, road material;
the vehicle information includes at least one of: speed, acceleration, load.
Optionally, the TBox data includes the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding accelerator pedal opening degree;
the obtaining the prediction information according to the TBox data comprises:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
Optionally, after determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further includes: and marking the corresponding relation between the road information and the oil consumption information in the map data.
Optionally, a plurality of paths exist between the departure place and the destination of the truck, and the prediction information includes prediction information of each of the plurality of paths;
after determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further comprises the following steps:
and determining a path with the minimum oil consumption between the departure place and the destination according to the oil consumption information corresponding to the prediction information of each road.
Embodiments of the present invention also provide a non-transitory computer-readable storage medium, wherein instructions of the storage medium, when executed by a processor of a fuel consumption prediction device of a truck, enable the fuel consumption prediction device of the truck to perform a fuel consumption prediction method of the truck, the method including:
obtaining forecast information according to TBox data collected by a remote information processing box TBox of the truck, wherein the forecast information comprises one or more of the following: road information, vehicle information, driving habit information;
determining oil consumption information corresponding to the prediction information according to the prediction information and a prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
Optionally, the prediction model includes at least one decision tree, where each decision tree in the at least one decision tree is used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information based on different decision trees are completely different or not completely the same;
the determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model comprises the following steps:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
Optionally, the TBox data includes position information of a current track point;
the obtaining the prediction information according to the TBox data comprises:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information corresponding to the road section according to the road section and the corresponding relation between the different road sections and the road information.
Optionally, the road information includes at least one of the following: road grade, road curvature, road material;
the vehicle information includes at least one of: speed, acceleration, load.
Optionally, the TBox data includes the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding accelerator pedal opening degree;
the obtaining the prediction information according to the TBox data comprises:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
Optionally, after determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further includes: and marking the corresponding relation between the road information and the oil consumption information in the map data.
Optionally, a plurality of paths exist between the departure place and the destination of the truck, and the prediction information includes prediction information of each of the plurality of paths;
after determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further comprises the following steps:
and determining a path with the minimum oil consumption between the departure place and the destination according to the oil consumption information corresponding to the prediction information of each road.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (11)
1. A method for predicting fuel consumption of a truck, comprising:
obtaining forecast information according to TBox data collected by a remote information processing box TBox of the truck, wherein the forecast information comprises one or more of the following: road information, vehicle information, driving habit information;
determining oil consumption information corresponding to the prediction information according to the prediction information and a prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
2. A method according to claim 1, wherein the prediction model comprises at least one decision tree, each decision tree of the at least one decision tree being used to determine a decision tree prediction result based on at least one type of prediction information, and the types of prediction information on which different decision trees are based are completely different or not completely the same;
the determining the oil consumption information corresponding to the prediction information according to the prediction information and the prediction model comprises the following steps:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
3. The method of claim 1, wherein the TBox data includes location information of a current track point;
the obtaining the prediction information according to the TBox data comprises:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information corresponding to the road section according to the road section and the corresponding relation between the different road sections and the road information.
4. The method according to claim 1, wherein the TBox data comprises the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding opening degree of the accelerator pedal;
the obtaining the prediction information according to the TBox data comprises:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
5. The method according to any one of claims 1 to 4, wherein after determining the fuel consumption information corresponding to the prediction information according to the prediction information and the prediction model, the method further comprises:
marking the corresponding relation between the road information and the oil consumption information in map data;
and/or the presence of a gas in the gas,
and determining a path with the minimum fuel consumption between the departure place of the truck and the destination of the truck according to the fuel consumption information corresponding to the prediction information of each of the roads, wherein the plurality of paths exist between the departure place and the destination, and the prediction information comprises the prediction information of each of the plurality of paths.
6. A fuel consumption prediction device for a truck, comprising:
the information module is used for obtaining prediction information according to TBox data acquired by a remote information processing box TBox of the truck, and the prediction information comprises one or more of the following: road information, vehicle information, driving habit information;
the oil consumption determining module is used for determining oil consumption information corresponding to the prediction information according to the prediction information and the prediction model; the prediction model is generated by adopting a machine learning algorithm to bring sample information, and the sample information indicates the mapping relation between the sample prediction information and the sample oil consumption information.
7. The apparatus according to claim 6, wherein the prediction model comprises at least one decision tree, each decision tree of the at least one decision tree is used for determining a decision tree prediction result based on at least one type of prediction information, and types of prediction information based on which different decision trees are based are completely different or not completely the same;
the oil consumption determining module is specifically configured to:
obtaining decision tree prediction results of the decision trees according to the prediction information;
and obtaining the oil consumption information corresponding to the prediction information according to the decision tree prediction result of each decision tree.
8. The apparatus of claim 6, wherein the TBox data comprises location information of a current track point;
the information module is specifically configured to:
determining a road section to which the current track point belongs according to the position information of the current track point;
and determining the road information corresponding to the road section according to the road section and the corresponding relation between the different road sections and the road information.
9. The device according to claim 6, wherein the TBox data comprises the stepping times of the accelerator pedal at each track point within the latest preset time range and the corresponding opening degree of the accelerator pedal;
the information module is specifically configured to:
and determining the driving habit information according to the stepping times of the accelerator pedal at each track point and the corresponding opening degree of the accelerator pedal.
10. The apparatus according to any one of claims 6-9, further comprising: a labeling module and/or a path determining module;
the marking module is used for marking the corresponding relation between the road information and the oil consumption information in map data;
the route determining module is configured to determine, according to fuel consumption information corresponding to prediction information of each of a plurality of roads, a route with a minimum fuel consumption between a departure location of the truck and a destination of the truck, where the plurality of routes exist between the departure location and the destination, and the prediction information includes prediction information of each of the plurality of routes.
11. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the method of any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811085613.4A CN110909907A (en) | 2018-09-18 | 2018-09-18 | Method and device for predicting fuel consumption of truck and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811085613.4A CN110909907A (en) | 2018-09-18 | 2018-09-18 | Method and device for predicting fuel consumption of truck and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110909907A true CN110909907A (en) | 2020-03-24 |
Family
ID=69813492
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811085613.4A Pending CN110909907A (en) | 2018-09-18 | 2018-09-18 | Method and device for predicting fuel consumption of truck and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110909907A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111554094A (en) * | 2020-05-20 | 2020-08-18 | 四川万网鑫成信息科技有限公司 | Journey prediction model based on vehicle journey, driving behavior and traffic information |
| CN112149909A (en) * | 2020-09-28 | 2020-12-29 | 神华中海航运有限公司 | Ship oil consumption prediction method and device, computer equipment and storage medium |
| CN113112061A (en) * | 2021-04-06 | 2021-07-13 | 深圳市汉德网络科技有限公司 | Method and device for predicting vehicle oil consumption |
| CN114005295A (en) * | 2020-07-28 | 2022-02-01 | 比亚迪股份有限公司 | Method, device, equipment and medium for predicting vehicle energy consumption information |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106767874A (en) * | 2015-11-19 | 2017-05-31 | 通用汽车环球科技运作有限责任公司 | The method and device with cost estimate is predicted for the fuel consumption by the quorum-sensing system in Vehicular navigation system |
| CN107330217A (en) * | 2017-07-13 | 2017-11-07 | 北京市交通信息中心 | A kind of middle sight oil consumption Forecasting Methodology based on RBFNN |
-
2018
- 2018-09-18 CN CN201811085613.4A patent/CN110909907A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106767874A (en) * | 2015-11-19 | 2017-05-31 | 通用汽车环球科技运作有限责任公司 | The method and device with cost estimate is predicted for the fuel consumption by the quorum-sensing system in Vehicular navigation system |
| CN107330217A (en) * | 2017-07-13 | 2017-11-07 | 北京市交通信息中心 | A kind of middle sight oil consumption Forecasting Methodology based on RBFNN |
Non-Patent Citations (1)
| Title |
|---|
| 牟小辉等: "基于随机森林算法的内河船舶油耗预测模型", 《交通信息与安全》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111554094A (en) * | 2020-05-20 | 2020-08-18 | 四川万网鑫成信息科技有限公司 | Journey prediction model based on vehicle journey, driving behavior and traffic information |
| CN114005295A (en) * | 2020-07-28 | 2022-02-01 | 比亚迪股份有限公司 | Method, device, equipment and medium for predicting vehicle energy consumption information |
| CN112149909A (en) * | 2020-09-28 | 2020-12-29 | 神华中海航运有限公司 | Ship oil consumption prediction method and device, computer equipment and storage medium |
| CN113112061A (en) * | 2021-04-06 | 2021-07-13 | 深圳市汉德网络科技有限公司 | Method and device for predicting vehicle oil consumption |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109520744B (en) | Driving performance testing method and device for automatic driving vehicle | |
| CN104331953B (en) | A kind of motor vehicle behavior data identification based on technology of Internet of things and management method | |
| CN109754594B (en) | Road condition information acquisition method and equipment, storage medium and terminal thereof | |
| JP5856387B2 (en) | Vehicle data analysis method and vehicle data analysis system | |
| CN110909907A (en) | Method and device for predicting fuel consumption of truck and storage medium | |
| CN110288096B (en) | Prediction model training method, prediction model training device, prediction model prediction method, prediction model prediction device, electronic equipment and storage medium | |
| EP2255347B9 (en) | Travel pattern information obtaining device, travel pattern information obtaining method, travel pattern information obtaining program and computer readable medium | |
| JP7228151B2 (en) | Traffic congestion prediction system, traffic congestion prediction method, learning device, prediction device, program, and learned model | |
| US20130261966A1 (en) | System and method for determining a vehicle route | |
| CN108444486B (en) | Navigation route sorting method and device | |
| CN111688713B (en) | Driving behavior analysis method and device | |
| CN112785841B (en) | Method and device for judging congestion state of truck driving route | |
| CN112598192B (en) | Method and device for predicting vehicle entering logistics park, storage medium and terminal | |
| CN106940929B (en) | Traffic data prediction method and device | |
| JP2012059058A (en) | Risk estimation device and program | |
| CN111613060A (en) | Data processing method and equipment | |
| US9747801B2 (en) | Method and device for determining surroundings | |
| CN108665723B (en) | Information acquisition method and device | |
| WO2021144923A1 (en) | Energy consumption estimation program, energy consumption estimation method, and energy consumption estimation device | |
| CN110696828B (en) | Forward target selection method and device and vehicle-mounted equipment | |
| CN113283460A (en) | Machine learning system and method for operating the system for determining a time series | |
| CN111160594B (en) | Method and device for estimating arrival time and storage medium | |
| CN109919293B (en) | Dangerous driving judgment method and device | |
| CN113284337B (en) | OD matrix calculation method and device based on vehicle track multidimensional data | |
| EP3859614A1 (en) | Device, method and machine learning system for determining a velocity for a vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200324 |