CN117400785A - Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle - Google Patents
Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle Download PDFInfo
- Publication number
- CN117400785A CN117400785A CN202210795441.XA CN202210795441A CN117400785A CN 117400785 A CN117400785 A CN 117400785A CN 202210795441 A CN202210795441 A CN 202210795441A CN 117400785 A CN117400785 A CN 117400785A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- data
- sample
- historical
- pure electric
- 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
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000010801 machine learning Methods 0.000 claims abstract description 51
- 238000004364 calculation method Methods 0.000 claims abstract description 16
- 238000012549 training Methods 0.000 claims description 36
- 238000005265 energy consumption Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 18
- 230000001133 acceleration Effects 0.000 claims description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 10
- 230000036541 health Effects 0.000 claims description 10
- 238000002790 cross-validation Methods 0.000 claims description 9
- 230000007613 environmental effect Effects 0.000 claims description 5
- 238000010200 validation analysis Methods 0.000 claims description 4
- 208000019901 Anxiety disease Diseases 0.000 abstract description 7
- 230000036506 anxiety Effects 0.000 abstract description 7
- 238000012545 processing Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 7
- 238000010276 construction Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000006399 behavior Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000446 fuel Substances 0.000 description 2
- 208000012868 Overgrowth Diseases 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
- B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries
- B60L58/12—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries responding to state of charge [SoC]
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L2240/00—Control parameters of input or output; Target parameters
- B60L2240/40—Drive Train control parameters
- B60L2240/54—Drive Train control parameters related to batteries
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Power Engineering (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The application relates to a prediction method and device of pure electric endurance mileage, electronic equipment and a vehicle, and relates to the technical field of data processing, wherein the method comprises the following steps: acquiring driving data, vehicle data and driving environment information of a vehicle to be predicted; extracting characteristic data from driving data, vehicle data and driving environment information, inputting the characteristic data into a preset machine learning model for calculation, and obtaining pure electric residual mileage values corresponding to the same model of the vehicle to be predicted according to the characteristic data; the characteristic data are used for representing the state of a battery of the vehicle and the dynamic performance of the vehicle; and determining the pure electric residual mileage value output by the preset machine learning model as the pure electric continuous mileage predicted value of the vehicle to be predicted. By applying the technical scheme, the prediction precision of the pure electric endurance mileage of the new energy automobile can be improved, the trust degree of a user on the prediction result is improved, the mileage anxiety is reduced, and the driving experience of the user can be improved.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a prediction method and device of pure electric endurance mileage, electronic equipment and a vehicle.
Background
The new energy automobile is an automobile which adopts unconventional automobile fuel as a power source (or adopts conventional automobile fuel and a novel automobile-mounted power device) and integrates the advanced technology in the aspects of power control and driving of the automobile, and the formed technical principle is advanced, and the automobile has a new technology and a new structure.
Currently, a new energy automobile obtains a predicted value Of a pure electric range through table lookup calculation according to a State Of Charge (SOC) and a temperature Of a battery. However, the prediction accuracy of this method is low, which results in low confidence of the user in the prediction result, and is easy to cause mileage anxiety, so as to affect the driving experience of the user.
Disclosure of Invention
In view of this, the present application provides a method, an apparatus, an electronic device, and a vehicle for predicting a pure electric range, which are mainly aimed at improving the technical problems that the prediction accuracy of the existing pure electric range prediction method is low, which results in that the confidence of the user to the prediction result is low, the range anxiety is easy to be caused, and the driving experience of the user is further affected.
In a first aspect, the present application provides a method for predicting a pure electric endurance mileage, including:
acquiring driving data, vehicle data and driving environment information of a vehicle to be predicted;
extracting characteristic data from the driving data, the vehicle data and the driving environment information, inputting the characteristic data into a preset machine learning model, and calculating to obtain pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data; the characteristic data are used for representing the state of a vehicle battery and the dynamic performance of the vehicle, and the preset machine learning model is obtained by training according to a historical driving data sample of a sample vehicle data set and a historical environment data sample corresponding to the historical driving data sample;
and determining the pure electric remaining mileage value output by the preset machine learning model as a pure electric endurance mileage predicted value of the vehicle to be predicted.
In a second aspect, the present application provides a prediction apparatus for a pure electric endurance mileage, including:
an acquisition module configured to acquire travel data of a vehicle to be predicted, vehicle data, and travel environment information;
the input module is configured to extract characteristic data from the driving data, the vehicle data and the driving environment information, input the characteristic data into a preset machine learning model for calculation, and obtain pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data; the characteristic data are used for representing the state of a vehicle battery and the dynamic performance of the vehicle, and the preset machine learning model is obtained by training according to a historical driving data sample of a sample vehicle data set and a historical environment data sample corresponding to the historical driving data sample;
and the determining module is configured to determine the pure electric residual mileage value output by the preset machine learning model as the pure electric continuous mileage predicted value of the vehicle to be predicted.
In a third aspect, the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method for predicting a range of pure electric power according to the first aspect.
In a fourth aspect, the present application provides an electronic device, including a storage medium, a processor, and a computer program stored on the storage medium and capable of running on the processor, where the processor implements the method for predicting a pure electric range according to the first aspect when executing the computer program.
In a fifth aspect, the present application provides a vehicle comprising: the electronic device of the fourth aspect.
By means of the technical scheme, the method, the device, the electronic equipment and the vehicle for predicting the pure electric endurance mileage are capable of accurately predicting the pure electric endurance mileage predicted value of the vehicle through a machine learning model according to the driving data, the vehicle data and the driving environment information of the vehicle to be predicted compared with the prior art. Specifically, feature data for representing the battery state and the vehicle power performance of the vehicle can be extracted from running data, vehicle data and running environment information of the vehicle to be predicted and input into a preset machine learning model for calculation, so that a pure electric residual mileage value corresponding to the same vehicle type of the vehicle to be predicted is obtained according to the feature data, the preset machine learning model is obtained by training according to a historical running data sample of a sample vehicle dataset and a historical environment data sample corresponding to the historical running data sample, and further a pure electric residual mileage value predicted value of the vehicle to be predicted can be determined according to a pure electric residual mileage value result output by the preset machine learning model. By applying the technical scheme, the prediction precision of the pure electric endurance mileage of the new energy automobile can be improved, the trust degree of a user on the prediction result is improved, the mileage anxiety is reduced, and the driving experience of the user can be improved.
The foregoing description is only an overview of the technical solutions of the present application, and may be implemented according to the content of the specification in order to make the technical means of the present application more clearly understood, and in order to make the above-mentioned and other objects, features and advantages of the present application more clearly understood, the following detailed description of the present application will be given.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the description of the embodiments or the prior art will be briefly described below, and it will be obvious to those skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Fig. 1 is a flow chart illustrating a method for predicting a pure electric range according to an embodiment of the present application;
fig. 2 is a schematic diagram showing a display effect of a predicted value of a vehicle remaining mileage according to an embodiment of the present application;
fig. 3 is a flow chart illustrating another method for predicting a pure electric range according to an embodiment of the present application;
FIG. 4 is a schematic diagram of an example of use of a machine learning model provided in an embodiment of the present application;
FIG. 5 shows a schematic diagram of an example of data feature extraction provided by an embodiment of the present application;
FIG. 6 shows a schematic diagram of an example of label computation provided by an embodiment of the present application;
fig. 7 shows a schematic structural diagram of a prediction apparatus for pure electric endurance mileage according to an embodiment of the present application.
Detailed Description
In order that the above objects, features and advantages of the present application may be more clearly understood, a further description of the aspects of the present application will be provided below. It should be noted that, in the case of no conflict, the embodiments of the present application and the features in the embodiments may be combined with each other.
In order to improve the technical problems that the prediction accuracy of the existing pure electric endurance mileage prediction mode is low, the confidence level of a user on a prediction result is low, mileage anxiety is easy to cause, and the driving experience of the user is further affected. The embodiment provides a method for predicting pure electric endurance mileage, as shown in fig. 1, the method includes:
step 101, acquiring driving data, vehicle data and driving environment information of a vehicle to be predicted.
The historical driving data may be related data of driving behavior of the user, and may specifically include: and the travel per kilometer energy consumption, the current average value, the vehicle speed average value, the acceleration and other data of the current travel of the vehicle to be predicted. The energy consumption of the journey per kilometer can be calculated through voltage, current and time signals recorded by the controller in the vehicle. The average value of the current can be calculated by the current and time signals recorded by the controller in the vehicle. The average value of the vehicle speed can be calculated by the vehicle speed in the speedometer and the time signal recorded by the controller. The acceleration may be calculated by a vehicle acceleration sensor.
The vehicle data may be state data of the vehicle, and may specifically include: data such as State Of Health (SOH), maximum temperature Of the battery pack, minimum temperature Of the battery pack, state Of Charge (SOC), current voltage Of the battery, and the like. The battery SOH is calculated from the physical manifestation of the voltage and current inside the battery and the electric quantity of the battery. The maximum temperature of the battery pack and the minimum temperature of the battery pack can be obtained by the temperature sensor. The battery SOC is calculated from the physical manifestation of the voltage and current inside the battery. The current voltage of the battery can be calculated by a voltage sensor.
The driving environment information may be environment information outside the vehicle, and may specifically include: information such as the vehicle outdoor temperature. The vehicle exterior temperature may be calculated by a temperature sensor.
And 102, extracting characteristic data from driving data, vehicle data and driving environment information of the vehicle to be predicted, inputting the characteristic data into a preset machine learning model for calculation, and acquiring pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data.
The feature data may be used to characterize a battery state of the vehicle and a dynamic performance of the vehicle, the preset machine learning model may be obtained by performing model training on sample data of a large number of sample vehicles of the same vehicle type as the vehicle to be predicted in advance, the preset machine learning model may be specifically obtained by training according to a historical driving data sample of the sample vehicle data set and a historical environment data sample corresponding to the historical driving data sample, and each vehicle type may have a corresponding preset machine learning model. In a specific application, a preset machine learning model corresponding to the model of the vehicle to be predicted can be found for calculation. For example, by performing calculation through the model, the pure electric residual mileage value corresponding to the same model sample vehicle at the time most similar to the current condition of the vehicle to be predicted (the running data, the vehicle data and the characteristic data used for representing the vehicle battery state and the vehicle dynamic performance in the running environment information) can be found and used as the output result of the model. Compared with the traditional prediction mode of looking up a table according to the SOC and the temperature of the battery, the method provided by the embodiment has more comprehensive considered factors, can utilize a preset machine learning model to calculate according to the running data, the vehicle data and the characteristic data used for representing the state of the battery and the dynamic performance of the vehicle in the running environment information, further combines the historical running data sample of the vehicle data set sample, the historical environment data sample corresponding to the historical running data sample and the like to analyze, and accurately predicts the pure electric range.
And 103, determining the pure electric remaining mileage value output by the preset machine learning model as a pure electric endurance mileage predicted value of the vehicle to be predicted.
For example, after the corresponding pure electric remaining mileage value is obtained by using the preset machine learning model, the predicted pure electric remaining mileage value can be displayed on an instrument panel of a vehicle driving position as the predicted pure electric remaining mileage value of the vehicle to be predicted, as shown in fig. 2, so that the driver can be accurately helped to know the current pure electric remaining mileage of the vehicle, so as to plan the vehicle driving route. If the driver can choose to go on to the destination or choose to go to the nearby charging pile for charging according to the predicted value of the pure electric remaining mileage of the vehicle.
Compared with the prior art, the method and the device can accurately predict the predicted value of the pure electric range of the vehicle through a machine learning model according to the driving data, the vehicle data and the driving environment information of the vehicle to be predicted. Specifically, feature data for representing the battery state and the vehicle power performance of the vehicle can be extracted from running data, vehicle data and running environment information of the vehicle to be predicted and input into a preset machine learning model for calculation, so that a pure electric residual mileage value corresponding to the same vehicle type of the vehicle to be predicted is obtained according to the feature data, the preset machine learning model is obtained by training according to a historical running data sample of a sample vehicle dataset and a historical environment data sample corresponding to the historical running data sample, and further a pure electric residual mileage value predicted value of the vehicle to be predicted can be determined according to a pure electric residual mileage value result output by the preset machine learning model. By applying the technical scheme of the embodiment, the prediction precision of the pure electric endurance mileage of the new energy automobile can be improved, the trust degree of a user on the prediction result is improved, the mileage anxiety is reduced, and the driving experience of the user can be improved.
Further, as a refinement and extension of the foregoing embodiment, in order to fully describe a specific implementation procedure of the method of the present embodiment, the present embodiment provides a specific method as shown in fig. 3, where the method includes:
step 201, obtaining a historical driving data sample, a vehicle data sample and a historical environment data sample of a sample vehicle of a target vehicle type in each journey, and obtaining a historical remaining range value corresponding to each journey of the sample vehicle.
When a preset machine learning model corresponding to a target vehicle type needs to be built, historical driving data, vehicle data and historical driving environment information of a sample vehicle of the target vehicle type in each journey can be obtained, and sample data such as historical remaining endurance mileage values corresponding to the sample vehicle in each journey can be obtained.
Step 202, constructing a data set based on a historical driving data sample, a vehicle data sample and a historical environment data sample of a sample vehicle in each journey and a historical remaining range value corresponding to each journey.
In this embodiment, the construction process of the machine learning model can be divided into several processes of data feature selection, label calculation, data set division, model modeling and training, model prediction, and the like. Wherein, as shown in fig. 4, the feature is the input of the model, and the label is the output of the model.
For the process of data feature selection, the process of determining sample feature data for a data set in step 202 may correspond. Accordingly, step 202 may specifically include: extracting energy consumption per kilometer, historical current average value, historical vehicle speed average value and acceleration from a historical driving data sample; and extracting the battery health degree, the maximum temperature of the battery pack, the minimum temperature of the battery pack, the battery charge state and the battery voltage from the vehicle data sample; extracting the outdoor temperature of the vehicle from the historical environmental data sample; and then taking the extracted travel per kilometer energy consumption, battery charge state, battery voltage, historical current average value, historical vehicle speed average value, acceleration, battery health degree, battery pack maximum temperature, battery pack minimum temperature and vehicle outdoor temperature of each travel as sample characteristic data of a data set, and taking the residual continuous mileage value corresponding to each travel as sample label information corresponding to the sample characteristic data, wherein the travel per kilometer energy consumption, the historical current average value and the historical vehicle speed average value can be non-dynamic data, and the acceleration, the battery health degree, the battery pack maximum temperature, the battery pack minimum temperature, the battery charge state, the battery voltage, the vehicle outdoor temperature and the like are dynamic instantaneous data, and periodically reporting the data on a vehicle side, and then carrying out model training through the massive data samples to obtain a machine learning model capable of accurately predicting the residual mileage, and executing the process shown in step 203.
For example, feature selection and processing are performed based on the historical travel data, the environmental information, and the vehicle data of the vehicle model under study, the selected features being shown in fig. 5. The method comprises the steps of obtaining a history travel distance, dividing the power consumption of a history travel distance by the power consumption of the history travel distance, and obtaining a history current and a history vehicle speed according to the history travel distance, wherein other characteristics are signals acquired from a CAN except for the history_energy_per_km, the BMS_RESSCurMean and the ESP_vehicle_mean, and the ESP_vehicle_mean are average values of the history current and the history vehicle speed respectively.
For the process of tag calculation, the process of determining sample tag information for the dataset in step 202 may correspond. Correspondingly, the obtaining the historical remaining range value corresponding to each trip of the sample vehicle specifically may include: and calculating the remaining range value of the sample vehicle at the tm moment through a formula I.
Wherein the sample vehicle starts from time t1 to time tn using the pure electric mode in each trip, t1<tm<tn，Label tm For the remaining range value of the sample vehicle at tm, S is the actual travel distance of the sample vehicle in the time interval from tm to tn, SOE tn The battery remaining energy of the sample vehicle at the time tn is the average energy consumption per kilometer of the sample vehicle in the time interval from the time t1 to the time tn.
For example, a Label (Label), i.e., the pure electric remaining mileage. Because the situation that the user runs out the electricity completely hardly occurs in the current historical data, the training data cannot acquire accurate pure electric residual mileage, and therefore the embodiment adopts the mode to replace the calculation of a real Label. As shown in fig. 6, if the user starts from time t1 to time tn using the pure electric mode, and t1< tm < tn, the pure electric remaining range value of the sample vehicle at time tm may be shown in formula one.
And when the journey at the tn moment is ended, the battery electric quantity is larger than 0 in most scenes, and for the pure electric residual range corresponding to the battery electric quantity at the moment, the pure electric residual range value of the sample vehicle at the tn moment can be obtained by dividing the available battery electric quantity at the moment by the average energy consumption per kilometer of the whole journey, and the pure electric residual range value can be specifically obtained as shown in a formula II.
Wherein Label is tn For the pure electric remaining range value of the sample vehicle at tn, SOE tn The battery remaining energy of the sample vehicle at the time tn is the average energy consumption per kilometer of the sample vehicle in the time interval from the time t1 to the time tn.
And 203, training to obtain a preset machine learning model through the constructed data set.
For the process of dividing the data set, the process of model training in step 203 may be corresponding, and optionally, step 203 may specifically include: the data set is partitioned into training and Validation sets using Cross Validation (CV) and subsequent Cross Validation. For example, a cross validation method is used to perform training set/validation set partitioning and subsequent cross validation, i.e., the data set is partitioned into k mutually exclusive subsets of similar size, k-1 subsets are used for training each time, 1 subset is used for validation, training is performed k times, and finally, the average value of k training results is returned.
For the process of model modeling and training, the process of model training in step 203 may be corresponding, and optionally, step 203 may specifically further include: firstly, utilizing an Xgboost integrated learning algorithm, and carrying out parameter adjustment on model parameters based on a data set so as to select an optimal parameter combination; model training is then completed after the optimal parameter combination is determined.
The machine learning model is various, the Xgboost integrated learning algorithm is selected by the algorithm, and the Xgboost integrated learning algorithm is an optimized distributed gradient lifting tree algorithm, is the fastest and best open source boosting tree tool kit at present, and is more than 10 times faster than a common tool kit. Xgboost is an integrated algorithm that sums the modeling results of all weak estimators by building multiple weak estimators on the data to obtain better regression or classification performance than a single model. And gradually accumulating a plurality of weak estimators through a plurality of iterations by constructing the weak estimators one by one to form a plurality of tree model integrated strong estimators. For regression trees (predicting pure endurance problem is regression problem), the value on each leaf node is the average of all samples on this leaf node.
Each leaf node has a prediction score, also known as a leaf weight. The leaf weight is the regression value of all the samples on the leaf node on the tree, and is expressed by fk (xi) or w, wherein fk represents the kth decision tree, and xi represents the feature vector corresponding to the sample i. When there are multiple trees, the regression result of the integrated model is the sum of the predictive scores of all the trees.
The core algorithm idea of Xgboost: a. continuously adding trees, and continuously performing feature splitting to grow a tree, wherein each time a tree is added, a new function fk (x) is learned, and the residual error of the last prediction is simulated; b. when training is completed to obtain k trees, predicting the score of one sample, namely, according to the characteristics of the sample, falling a corresponding leaf node in each tree, wherein each leaf node corresponds to one score; c. finally, only the score corresponding to each tree needs to be added up to be the predicted value of the sample.
The structure of the different trees is enumerated continuously, then a scoring function is utilized to find a tree of the optimal structure, then the tree is added to the model, and the operation is repeated continuously. This search uses a greedy algorithm. Selecting a feature split, calculating a loss function minimum value, then selecting a feature split, obtaining a loss function minimum value, and obtaining small sapling by enumerating and finding a tree with the best effect and giving the split. After the feature data and the tag data are prepared, the Xgboost model is determined to be selected, and then the Xgboost model needs to be constructed and trained.
The model construction is realized by using the library of the Xgboost, and the specific process is as follows: a. importing the necessary data packet (including); b. constructing a training set and a testing set into a format which can be used in Xgboost; c. and training a model. The model training process is a process of selecting an optimal parameter combination by tuning parameters of a model based on a training data set, and the grid search algorithm is a method for optimizing the performance of the model by traversing a given parameter combination.
Wherein the model parameters may include: objective, eta, min _child_ weight, gamma, max _delta_ step, subsample, lambda.
The objective (default reg: linear), this parameter defines the loss function that needs to be minimized (two classification problems are typically used "binary: logistic" and "rank: pariwise"). The most common values are (customizable): binary-logistic two-class logistic regression returns the probability of prediction (not class). MultiSoftmax returns predicted categories (not probabilities) using the multiple classifier of Softmax. In this case, one more parameter needs to be set: num_class (number of categories). MultiSoftprob is the same as the MultiSoftmax parameter, but returns the probability that each data belongs to the respective category.
eta [ default 0.3]: a learning rate; for updating the weights of leaf nodes, multiplying by the coefficients, avoiding step size overgrowth. The larger the parameter value, the more likely it is that convergence is not possible. Setting the learning rate eta smaller, the small learning rate can make the later learning more careful.
min_child_weight [ default 1]: the sum of minimum weights of leaf nodes; this parameter is used to avoid overfitting and when this value is large, it can avoid model learning to a local special sample. The smaller the parameter, the more likely the overfitting occurs; however, if this value is too high, a under-fit may result. This parameter needs to be adjusted using cross-validation (CV).
gamma [ default 0]: when a node splits, it will split only if the value of the post-split loss function drops. Gamma specifies the minimum loss function drop value required for node splitting. The larger the value of this parameter, the more conservative the algorithm. Between 0.1 and 0.2. This parameter is then also to be adjusted.
max_delta_step [ default 0]: this parameter acts in the update step to limit the maximum step size of the weight change per tree.
subsamples [ default 1]: a proportion of random sample samples per tree is generated. Lower values make the algorithm more conservative, preventing over-fitting, and values that are too small may result in under-fitting.
lambda [ default 1]: this parameter is used to control the regularized part of the Xgboost. More uses can be mined in reducing overfitting.
The general order of Xgboost tuning can be referenced: and determining a larger learning rate of 0.1, num_boost_round tuning, max_depth and min_weight parameter tuning, gamma parameter tuning, regularization parameter tuning and reducing the learning rate. After the model parameters are determined, model training is completed, and then the model is saved, namely, a preset machine learning model is obtained.
The trained model may then be invoked to predict and evaluate performance of the new data for the range (Label), the accuracy of the range prediction being evaluated by mean absolute error (MeanAbsolute Error, MAE).
And 204, acquiring driving data, vehicle data and driving environment information of the vehicle to be predicted when receiving an instruction for predicting the pure electric range of the vehicle to be predicted.
And 205, extracting characteristic data from the driving data, the vehicle data and the driving environment information of the vehicle to be predicted, inputting the characteristic data into a preset machine learning model for calculation, and acquiring pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data.
Optionally, step 205 may specifically include: extracting the energy consumption per kilometer of the current journey of the vehicle to be predicted, the battery charge state, the battery voltage, the current average value, the vehicle speed average value, the acceleration, the battery health degree, the maximum battery pack temperature, the minimum battery pack temperature and the outdoor temperature, inputting the energy consumption per kilometer into a preset machine learning model for calculation, and obtaining the pure electric residual mileage value corresponding to a sample vehicle of the same vehicle type as the vehicle to be predicted under the condition of similar history according to the characteristic data.
And 206, determining the pure electric remaining mileage value output by the preset machine learning model as a pure electric endurance mileage predicted value of the vehicle to be predicted.
The embodiment provides a pure electric endurance mileage prediction method based on machine learning, which can be used for predicting the endurance mileage by means of a machine learning model in combination with driving behavior history data, environment data and the like, namely, high-precision prediction of the pure electric endurance mileage is performed in combination with vehicle real-time information, environment information, user driving behavior and the like. The prediction accuracy of the cruising mileage is improved, the behavior and the like of the user can be timely fed back to the calculated prediction result of the cruising mileage, the trust degree of the user to the prediction result is improved, the driving experience of the user is reduced, and the driving experience of the user is improved.
Further, as a specific implementation of the method shown in fig. 1 and fig. 3, the embodiment provides a prediction apparatus for pure electric endurance mileage, as shown in fig. 7, where the apparatus includes: an acquisition module 31, an input module 32, a determination module 33.
An acquisition module 31 configured to acquire travel data of a vehicle to be predicted, vehicle data, and travel environment information;
the input module 32 is configured to extract characteristic data from the driving data, the vehicle data and the driving environment information, input the characteristic data into a preset machine learning model for calculation, and obtain pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data; the characteristic data are used for representing the state of a vehicle battery and the dynamic performance of the vehicle, and the preset machine learning model is obtained by training according to a historical driving data sample of a sample vehicle data set and a historical environment data sample corresponding to the historical driving data sample;
the determining module 33 is configured to determine the pure electric remaining mileage value output by the preset machine learning model as a pure electric endurance mileage predicted value of the vehicle to be predicted.
In a specific application scenario, the device further includes: a model building module;
the model construction module is configured to acquire a historical driving data sample, a vehicle data sample and a historical environment data sample of a sample vehicle of a target vehicle type in each journey, and acquire a historical remaining endurance mileage value corresponding to the sample vehicle in each journey; constructing a data set based on the historical driving data sample, the vehicle data sample and the historical environment data sample of each trip of the sample vehicle and the historical remaining range value corresponding to each trip; and training to obtain the preset machine learning model through the data set.
In a specific application scenario, the model construction module is specifically configured to extract stroke per kilometer energy consumption, a historical current average value, a historical vehicle speed average value and acceleration from the historical driving data sample; and extracting battery health, maximum temperature of a battery pack, minimum temperature of the battery pack, battery state of charge and battery voltage from the vehicle data sample; extracting the outdoor temperature of the vehicle from the historical environmental data sample; and taking the extracted energy consumption per kilometer of each travel, the battery charge state, the battery voltage, the historical current average value, the historical vehicle speed average value, the acceleration, the battery health, the maximum battery pack temperature, the minimum battery pack temperature and the outdoor vehicle temperature of each travel as sample characteristic data of the data set, and taking a historical remaining range value corresponding to each travel as sample label information corresponding to the sample characteristic data.
In a specific application scenario, the input module 32 is specifically configured to extract the energy consumption per kilometer of the current journey of the vehicle to be predicted, the battery state of charge, the battery voltage, the current average value, the vehicle speed average value, the acceleration, the battery health, the maximum temperature of the battery pack, the minimum temperature of the battery pack and the outdoor temperature, and input the extracted energy consumption per kilometer of the current journey of the vehicle to be predicted into a preset machine learning model for calculation.
In a specific application scenario, the model building module is specifically further configured to pass through the formulaCalculating a pure electric remaining range value of the sample vehicle at tm, wherein the sample vehicle starts from t1 to tn from a pure electric mode in each journey, t1<tm<tn，Label tm For the pure electric remaining range value of the sample vehicle at the moment tm, S is the actual travel distance of the sample vehicle in the time interval from the moment tm to the moment tn, and SOE tn And e is the average energy consumption per kilometer of the sample vehicle in a time interval from the time t1 to the time tn.
In a specific application scenario, the model building module is specifically further configured to divide the data set into a training set and a verification set by using a cross-verification method, and then perform cross-verification.
In a specific application scene, the model construction module is specifically configured to utilize an Xgboost integrated learning algorithm to tune model parameters based on the data set so as to select an optimal parameter combination; and finishing model training after the optimal parameter combination is determined.
It should be noted that, other corresponding descriptions of each functional unit related to the prediction apparatus for pure electric endurance mileage provided in the present embodiment may refer to corresponding descriptions in fig. 1 and fig. 3, and are not described herein again.
Based on the above-described methods shown in fig. 1 and 3, correspondingly, the present embodiment further provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the above-described methods shown in fig. 1 and 3.
Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.), and includes several instructions for causing a computer device (may be a personal computer, a server, or a network device, etc.) to perform the method of each implementation scenario of the present application.
Based on the methods shown in fig. 1 and fig. 3 and the virtual device embodiment shown in fig. 7, in order to achieve the above objects, the embodiment of the present application further provides an electronic device, which may be configured on an end side of a vehicle (such as a new energy automobile), where the device includes a storage medium and a processor; a storage medium storing a computer program; a processor for executing a computer program to implement the method as described above and shown in fig. 1 and 3.
Optionally, the entity device may further include a user interface, a network interface, a camera, a Radio Frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, and so on. The user interface may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface), etc.
It will be appreciated by those skilled in the art that the above-described physical device structure provided in this embodiment is not limited to this physical device, and may include more or fewer components, or may combine certain components, or may be a different arrangement of components.
The storage medium may also include an operating system, a network communication module. The operating system is a program that manages the physical device hardware and software resources described above, supporting the execution of information handling programs and other software and/or programs. The network communication module is used for realizing communication among all components in the storage medium and communication with other hardware and software in the information processing entity equipment.
Based on the above electronic device, the embodiment of the application further provides a vehicle, which may specifically include: the electronic equipment. The vehicle can be a new energy automobile and the like.
From the above description of the embodiments, it will be apparent to those skilled in the art that the present application may be implemented by means of software plus necessary general hardware platforms, or may be implemented by hardware. By applying the scheme of the embodiment, the prediction precision of the pure electric endurance mileage of the new energy automobile can be improved, the trust degree of a user on the prediction result is improved, the mileage anxiety is reduced, and the driving experience of the user can be improved.
It should be noted that in this document, relational terms such as "first" and "second" and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing is merely a specific embodiment of the application to enable one skilled in the art to understand or practice the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown and described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. The prediction method of the pure electric endurance mileage is characterized by comprising the following steps of:
acquiring driving data, vehicle data and driving environment information of a vehicle to be predicted;
extracting characteristic data from the driving data, the vehicle data and the driving environment information, inputting the characteristic data into a preset machine learning model, and calculating to obtain pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data; the characteristic data are used for representing the state of a vehicle battery and the dynamic performance of the vehicle, and the preset machine learning model is obtained by training according to a historical driving data sample of a sample vehicle data set and a historical environment data sample corresponding to the historical driving data sample;
and determining the pure electric remaining mileage value output by the preset machine learning model as a pure electric endurance mileage predicted value of the vehicle to be predicted.
2. The method of claim 1, wherein the process of constructing the pre-set machine learning model comprises:
acquiring a historical driving data sample, a vehicle data sample and a historical environment data sample of a sample vehicle of a target vehicle type in each journey, and acquiring a historical remaining range value corresponding to each journey of the sample vehicle;
constructing a data set based on the historical driving data sample, the vehicle data sample and the historical environment data sample of each trip of the sample vehicle and the historical remaining range value corresponding to each trip;
and training to obtain the preset machine learning model through the data set.
3. The method of claim 2, wherein constructing the data set based on the historical driving data sample, the vehicle data sample, and the historical environmental data sample for each trip of the sample vehicle, and the historical remaining range value for each trip, comprises:
extracting stroke energy consumption per kilometer, a historical current average value, a historical vehicle speed average value and acceleration from the historical driving data sample; the method comprises the steps of,
extracting battery health, maximum temperature of a battery pack, minimum temperature of the battery pack, battery charge state and battery voltage from the vehicle data sample; the method comprises the steps of,
extracting the outdoor temperature of the vehicle from the historical environmental data sample;
and taking the extracted energy consumption per kilometer of each travel, the battery charge state, the battery voltage, the historical current average value, the historical vehicle speed average value, the acceleration, the battery health, the maximum battery pack temperature, the minimum battery pack temperature and the outdoor vehicle temperature of each travel as sample characteristic data of the data set, and taking a historical remaining range value corresponding to each travel as sample label information corresponding to the sample characteristic data.
4. The method of claim 2, wherein the obtaining the historical remaining range value for each trip of the sample vehicle comprises:
by the formulaCalculating a pure electric remaining range value of the sample vehicle at tm, wherein the sample vehicle starts from t1 to tn from a pure electric mode in each journey, t1<tm<tn，Label tm For the pure electric remaining range value of the sample vehicle at the moment tm, S is the actual travel distance of the sample vehicle in the time interval from the moment tm to the moment tn, and SOE tn And e is the average energy consumption per kilometer of the sample vehicle in a time interval from the time t1 to the time tn.
5. The method of claim 2, wherein training the pre-set machine learning model from the data set comprises:
the data set is partitioned into training and validation sets using a cross-validation method, and subsequent cross-validation.
6. The method of claim 2, wherein training the pre-set machine learning model from the data set comprises:
performing parameter adjustment on model parameters based on the data set by utilizing an Xgboost integrated learning algorithm so as to select an optimal parameter combination;
and finishing model training after the optimal parameter combination is determined.
7. The utility model provides a prediction device of pure electric endurance mileage which characterized in that includes:
an acquisition module configured to acquire travel data of a vehicle to be predicted, vehicle data, and travel environment information;
the input module is configured to extract characteristic data from the driving data, the vehicle data and the driving environment information, input the characteristic data into a preset machine learning model for calculation, and obtain pure electric residual mileage values corresponding to the same vehicle type of the vehicle to be predicted according to the characteristic data; the characteristic data are used for representing the state of a vehicle battery and the dynamic performance of the vehicle, and the preset machine learning model is obtained by training according to a historical driving data sample of a sample vehicle data set and a historical environment data sample corresponding to the historical driving data sample;
and the determining module is configured to determine the pure electric residual mileage value output by the preset machine learning model as the pure electric continuous mileage predicted value of the vehicle to be predicted.
8. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the method of any one of claims 1 to 6.
9. An electronic device comprising a storage medium, a processor and a computer program stored on the storage medium and executable on the processor, characterized in that the processor implements the method of any one of claims 1 to 6 when executing the computer program.
10. A vehicle, characterized by comprising: the electronic device of claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210795441.XA CN117400785A (en) | 2022-07-07 | 2022-07-07 | Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210795441.XA CN117400785A (en) | 2022-07-07 | 2022-07-07 | Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117400785A true CN117400785A (en) | 2024-01-16 |
Family
ID=89496687
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210795441.XA Pending CN117400785A (en) | 2022-07-07 | 2022-07-07 | Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117400785A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117851788A (en) * | 2024-03-04 | 2024-04-09 | 泓浒(苏州)半导体科技有限公司 | Automatic wafer transmission scheduling method and system based on crown block |
-
2022
- 2022-07-07 CN CN202210795441.XA patent/CN117400785A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117851788A (en) * | 2024-03-04 | 2024-04-09 | 泓浒(苏州)半导体科技有限公司 | Automatic wafer transmission scheduling method and system based on crown block |
| CN117851788B (en) * | 2024-03-04 | 2024-05-07 | 泓浒(苏州)半导体科技有限公司 | Automatic wafer transmission scheduling method and system based on crown block |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108896914B (en) | Gradient lifting tree modeling and prediction method for health condition of lithium battery | |
| CN111381170A (en) | Electric vehicle battery pack health state prediction method and system based on big data | |
| CN110806541B (en) | AD-BAS-based lithium battery model parameter identification method | |
| US11704570B2 (en) | Learning device, learning system, and learning method | |
| CN111999649A (en) | XGboost algorithm-based lithium battery residual life prediction method | |
| CN111079780A (en) | Training method of space map convolution network, electronic device and storage medium | |
| CN109598566A (en) | Lower list prediction technique, device, computer equipment and computer readable storage medium | |
| CN110658460A (en) | Battery life prediction method and device for battery pack | |
| CN113887643B (en) | New dialogue intention recognition method based on pseudo tag self-training and source domain retraining | |
| CN113125960A (en) | Vehicle-mounted lithium ion battery charge state prediction method based on random forest model | |
| CN117400785A (en) | Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle | |
| CN115730635A (en) | Electric vehicle load prediction method | |
| CN112305441A (en) | Power battery health state assessment method under integrated clustering | |
| CN114609994A (en) | Fault diagnosis method and device based on multi-granularity regularization rebalance incremental learning | |
| CN109543922B (en) | Time-period borrowing and returning amount prediction method for single-vehicle station group shared by piles | |
| CN116106761A (en) | Real-time lithium ion battery electric quantity estimation method based on typical correlation analysis | |
| CN117074951A (en) | Lithium battery state of charge estimation method, device, equipment and readable storage medium | |
| CN115691140B (en) | Analysis and prediction method for space-time distribution of automobile charging demand | |
| CN112801271B (en) | Method for generating neural network, data processing method and intelligent driving control method | |
| CN112651443A (en) | Driving style identification model evaluation method, device, medium and equipment based on machine learning | |
| CN112348224A (en) | Method and device for predicting service life of vehicle battery, electronic equipment and storage medium | |
| CN113469244B (en) | Volkswagen app classification system | |
| CN117648631B (en) | Power battery health state estimation method for electric automobile group | |
| CN113468771B (en) | Vibration estimation method using structure intrinsic parameters | |
| CN117841689A (en) | Power battery remaining endurance prediction method based on time sequence large model algorithm |
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 |