CN110706039A - Electric vehicle residual value rate evaluation system, method, equipment and medium - Google Patents
Electric vehicle residual value rate evaluation system, method, equipment and medium Download PDFInfo
- Publication number
- CN110706039A CN110706039A CN201910984012.5A CN201910984012A CN110706039A CN 110706039 A CN110706039 A CN 110706039A CN 201910984012 A CN201910984012 A CN 201910984012A CN 110706039 A CN110706039 A CN 110706039A
- Authority
- CN
- China
- Prior art keywords
- data
- residual value
- rate evaluation
- residual
- evaluation model
- 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
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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0278—Product appraisal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Software Systems (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Educational Administration (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a system, a method, equipment and a medium for evaluating the residual value rate of an electric automobile, wherein the system comprises the following components: a historical data acquisition module; the evaluation model training module is used for constructing at least one residual value rate evaluation model based on a machine learning algorithm and training the residual value rate evaluation model by adopting the training set; evaluating a model library; the vehicle data acquisition module is used for acquiring the characteristic data of the electric vehicle to be evaluated; and the residual value rate evaluation module is used for inputting the characteristic data of the electric automobile to be evaluated into the residual value rate evaluation model to obtain the residual value rate output by the residual value rate evaluation model. The method is based on the big data technology and the machine learning technology, automatically predicts the residual value of the electric automobile, reduces the manual detection cost, comprehensively considers the characteristic data of the electric automobile in the prediction, and predicts the residual value according to the automobile condition in real time, so that the accuracy of residual value prediction is improved, and the practicability is high.
Description
Technical Field
The invention relates to the technical field of vehicle data processing, in particular to a system, a method, equipment and a medium for evaluating the residual value rate of an electric vehicle.
Background
The new energy electric vehicle industry is one of seven strategic emerging industries in China, and the industrial scale development is rapid in recent years. In 2009, China starts to formally start the popularization demonstration work of 'ten cities and thousands of vehicles' of new energy electric vehicles, and as late as 2018, more than 550 thousands of new energy electric vehicles are sold globally in an accumulated mode, the percentage of China exceeds 53%, and 2015-charge 2018 is the first to sell for 4 consecutive years. In contrast, consumers often consider uncertainty of automobile residual values when purchasing new energy automobiles nowadays, which adversely affects further popularization of new energy automobiles. The automobile residual value is the remaining use value within the specified reasonable service life of the automobile. The insurance industry also troubles the pure electric vehicles with incomplete commercial insurance, and the problems hinder the popularization of the pure electric vehicles to a certain extent. Therefore, a set of evaluation methods for the residual value of the pure electric vehicle is urgently needed to be established.
There are generally 3 traditional methods for finding a residual value evaluation through investigation: a reset cost law, a current market price law, and a clearing price law. 1) The replacement cost method is a method of replacing the current price of the vehicle to be evaluated by the replacement cost × the new rate. The reset cost is the lowest cost paid to purchase a new vehicle identical to the vehicle being evaluated. The renewal rate is (1-used age/specified used age) × adjustment factor × 100%. The defects are that the objectivity is lacked, and the influence of subjective factors is large when the success rate is determined. 2) The current market price method selects one or more assets which are the same as or similar to the evaluation object as comparison objects through market adjustment, and analyzes the current transaction price and the transaction condition of the comparison objects. A method of estimating the value of an asset to be assessed. The method has the disadvantages that the method has higher market requirement, a fully active second-hand car market needs to exist, and vehicles in the market are comprehensive enough and large in quantity so as to provide similar reference vehicles in the aspects of service life, regions, car types, colors and the like for the evaluated vehicles; the market division does not exist in the market, and the buyer and the seller are equal in status, so that the contingency of the transaction can be avoided. 3) The clearing price method has no special evaluation method theoretically. However, in practice, the following methods are mostly used: the method gives a lowest discount rate which can be paid, and when the discount rate given by the buyer is not lower than the lowest discount rate, the transaction can be made. The product of the evaluation value and the discount rate is the clearing price. The defect is that the method is mainly used for special conditions of enterprise bankruptcy, backlog, voluntary stop, unfit credit and the like. Due to the large restrictions of the applicable conditions, the clearing price is significantly lower than the general residual value.
In addition, in the prior art, an automobile residual value prediction method based on big data is also available, which is mainly characterized in that a causal model is used to establish the relationship between a single factor and a residual value under a big data set, and then the relationships are simply combined and processed to obtain a final result. But the relationship among a plurality of factors is ignored in the mode, and the interpretability is not strong.
In conclusion, the existing automobile residual value evaluation method is simple, cannot fully utilize data, seriously depends on manual detection (time consumption, high cost and strong subjectivity), and causes large estimation error. The existing evaluation methods are basically suitable for fuel vehicles, no scheme specially aiming at electric vehicles is provided, and the electric vehicles are greatly different from the traditional fuel vehicles, so that the residual value prediction of the electric vehicles cannot follow the residual value evaluation scheme of the fuel vehicles.
Disclosure of Invention
Aiming at the problems in the prior art, the invention aims to provide a system, a method, equipment and a medium for evaluating the residual value rate of an electric vehicle, and solves the problems that in the prior art, the use of information is less, manual detection is seriously relied on, modeling cannot be performed under a small data set, and the evaluation accuracy rate is low.
The embodiment of the invention provides an electric automobile residual value rate evaluation system, which comprises:
the historical data acquisition module is used for acquiring historical transaction data of the electric automobile and forming a training set, wherein the historical transaction data comprises characteristic data and residual value rate data of the electric automobile;
the evaluation model training module is used for constructing at least one residual value rate evaluation model based on a machine learning algorithm and training the residual value rate evaluation model by adopting the training set, wherein the input of the residual value rate evaluation model is the characteristic data of the electric automobile, and the output of the residual value rate evaluation model is the residual value rate of the electric automobile;
the evaluation model library is used for storing the trained residual value rate evaluation model;
the vehicle data acquisition module is used for acquiring the characteristic data of the electric vehicle to be evaluated;
and the residual value rate evaluation module is used for inputting the characteristic data of the electric automobile to be evaluated into the residual value rate evaluation model to obtain the residual value rate output by the residual value rate evaluation model.
Optionally, the historical data acquiring module acquires historical transaction data of the electric vehicle, and includes acquiring historical order data by the historical data acquiring module, analyzing the historical order data to obtain feature data and residual value rate data of the electric vehicle, and acquiring sensor detection data of the electric vehicle by the historical data acquiring module from a vehicle-mounted information processing device of the electric vehicle.
Optionally, the evaluation model training module constructs at least one residual value rate evaluation model, and includes that the evaluation model training module constructs a plurality of residual value rate evaluation models corresponding to vehicle type categories one by one according to preset vehicle type classifications;
the evaluation model training module trains the residual value rate evaluation model by adopting the training set, and comprises an evaluation model training module which classifies data in the training set according to preset vehicle type classification and inputs training data of different vehicle type categories into corresponding residual value rate evaluation models respectively.
Optionally, the residual value rate evaluation module inputs the feature data of the electric vehicle to be evaluated to the residual value rate evaluation model, and the method includes the steps of judging the vehicle type category of the electric vehicle to be evaluated by the residual value rate evaluation module, and inputting the feature data of the electric vehicle to be evaluated to the residual value rate evaluation model corresponding to the vehicle type category of the electric vehicle to be evaluated.
Optionally, the characteristic data of the electric automobile comprises vehicle basic characteristic data, vehicle sales characteristic data, electric component characteristic data and use habit characteristic data;
the vehicle basic characteristic data comprises one or more of cities, vehicle types, models, annual money, mileage, card-serving time, appearance information, transaction types, transaction times and use properties;
the vehicle sales characteristic data comprises one or more of brand grade, public praise grade, preferential information and sales information;
the electric power component characteristic data comprises one or more of power battery information, motor information and vehicle control unit information;
the usage habit characteristic data includes one or more of driving behavior information, charging behavior information, maintenance records, and fault records.
Optionally, after the historical data acquisition module acquires historical transaction data of the electric vehicle, the historical data acquisition module sorts the historical transaction data to obtain a plurality of pieces of historical data, and each piece of historical data includes a plurality of features and corresponding feature values.
Optionally, setting a first threshold and a second threshold, which are respectively a specific value between 1 and 0; the evaluation model training module judges whether the feature value missing proportion of each piece of historical data is larger than a first threshold value or not for each piece of historical data, and if so, deletes the piece of historical data;
if the missing proportion of the characteristic value of the historical data is between a first threshold value and a second threshold value, a model prediction method is adopted to supplement the missing characteristic value of the historical data;
and if the missing proportion of the characteristic value of the historical data is smaller than a second threshold value, supplementing the missing characteristic value of the historical data by adopting an averaging method or an interpolation method according to the data corresponding to the missing characteristic value in other historical data.
Optionally, the supplementing the missing feature value of the piece of historical data by using the model prediction method includes determining a type of the missing feature value of the piece of historical data, if the missing feature value is a feature value of a discrete feature, supplementing the missing feature value of the piece of historical data by using a support vector machine, Logistic regression analysis or a tree classification model, and if the missing feature value is a feature value of a continuous feature, supplementing the missing feature value of the piece of historical data by using a linear regression support vector machine or a LightGBM regression model.
Optionally, the evaluation model training module is further configured to filter features of each piece of historical data by using the following steps:
calculating the variance of the feature value of each feature, and if the variance of the feature value of a feature is smaller than 0.1 after normalization processing, deleting the feature and the corresponding feature value in each piece of historical data;
and calculating the correlation between each feature and the residual value rate by adopting a Pearson coefficient correlation matrix method, namely the correlation coefficient corresponding to each feature, and deleting the feature and the corresponding feature value in each historical data if the correlation coefficient of one feature is less than 0.1.
Optionally, the evaluation model training module is further configured to filter features of each piece of historical data by using the following steps:
and performing multiple rounds of training on one base model by adopting the training set, removing the characteristic and the characteristic value with the minimum absolute weight after each round of training, and performing next round of training on the basis of the training set with the characteristic and the characteristic value removed until the residual characteristic quantity meets the preset characteristic quantity requirement.
Optionally, the evaluation model training module further splits the training set obtained by the historical data obtaining module into a training set and a verification set by using a cross validation method, trains the residual value rate evaluation model by using the split training set, and verifies the generalization performance of the trained residual value rate evaluation model by using the verification set.
Optionally, the evaluation model training module trains the residual rate evaluation model by using a regression tree model algorithm.
The embodiment of the invention also provides an electric vehicle residual value rate evaluation method, which adopts the electric vehicle residual value rate evaluation system and comprises the following steps:
the vehicle data acquisition module acquires characteristic data of the electric vehicle to be evaluated;
the residual value rate evaluation module selects a residual value rate evaluation model from the evaluation model library;
and the residual value rate evaluation module inputs the characteristic data of the electric automobile to be evaluated into the residual value rate evaluation model to obtain the residual value rate output by the residual value rate evaluation model.
The embodiment of the invention also provides an electric vehicle residual value rate evaluation device, which comprises:
a processor;
a memory having stored therein executable instructions of the processor;
wherein the processor is configured to execute the steps of the electric vehicle residual value rate evaluation method via executing the executable instructions.
The embodiment of the invention also provides a computer-readable storage medium for storing a program, and the program realizes the steps of the electric vehicle residual value rate evaluation method when being executed.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
The system, the method, the equipment and the medium for evaluating the residual value rate of the electric automobile have the following advantages that:
the method is based on the big data technology and the machine learning technology, automatically predicts the residual value of the electric automobile, reduces the manual detection cost, comprehensively considers the characteristic data of the electric automobile in the prediction, and predicts the residual value according to the automobile condition in real time, so that the accuracy of residual value prediction is improved, the method is more suitable for the electric automobile, has strong practicability, and is beneficial to popularization and application of the electric automobile.
Drawings
Other features, objects and advantages of the present invention will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, with reference to the accompanying drawings.
FIG. 1 is a schematic diagram illustrating a residual value rate evaluation system of an electric vehicle according to an embodiment of the present invention;
FIG. 2 is a flow chart of modeling of an electric vehicle residual rating system according to an embodiment of the present invention;
FIG. 3 is a schematic diagram illustrating a modeling process of an electric vehicle residual value rate evaluation system according to an embodiment of the present invention;
FIG. 4 is a flowchart illustrating a method for evaluating a residual value rate of an electric vehicle according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of an apparatus for estimating a residual value rate of an electric vehicle according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a computer-readable storage medium according to an embodiment of the present invention.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Furthermore, the drawings are merely schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.
As shown in fig. 1, in order to solve the technical problems of the prior art, the present invention provides an electric vehicle residual value rate evaluation system, which includes:
the historical data acquisition module M100 is used for acquiring historical transaction data of the electric automobile and forming a training set, wherein the historical transaction data comprises characteristic data and residual value rate data of the electric automobile;
the evaluation model training module M200 is used for constructing at least one residual value rate evaluation model based on a machine learning algorithm and training the residual value rate evaluation model by adopting the training set, wherein the input of the residual value rate evaluation model is the characteristic data of the electric automobile, and the output of the residual value rate evaluation model is the residual value rate of the electric automobile;
the evaluation model library M300 is used for storing the trained residual value rate evaluation model;
the vehicle data acquisition module M400 is used for acquiring the characteristic data of the electric vehicle to be evaluated;
and the residual value rate evaluation module M500 is used for inputting the characteristic data of the electric vehicle to be evaluated into the residual value rate evaluation model to obtain the residual value rate of the electric vehicle output by the residual value rate evaluation model.
Therefore, the electric vehicle residual value rate evaluation system provided by the invention models the residual value rate evaluation by adopting the functional modules based on a big data technology and a machine learning technology, automatically predicts the residual value of the electric vehicle, reduces the manual detection cost, comprehensively considers the characteristic data of the electric vehicle in the prediction process, and predicts the residual value according to the vehicle condition in real time, thereby improving the accuracy of the residual value prediction.
In this embodiment, the historical data acquiring module acquires historical transaction data of the electric vehicle, and includes acquiring historical order data by the historical data acquiring module, analyzing the historical order data to obtain feature data and residual value rate data of the electric vehicle, and acquiring sensor detection data of the electric vehicle by the historical data acquiring module from a vehicle-mounted information processing device of the electric vehicle. The historical data acquisition module can establish connection with a sales order management system, directly acquire electronic historical order data and analyze the order to obtain required data. And the detection data of the sensors are acquired by vehicle-mounted information processing equipment such as a CAN bus, a vehicle-mounted T-BOX and the like, so that the data acquisition is completed.
In this embodiment, the characteristic data of the electric vehicle includes vehicle basic characteristic data, vehicle sales characteristic data, electric component characteristic data, and usage habit characteristic data;
the vehicle basic characteristic data comprises one or more of city, model, year, mileage, time on sale, appearance information (such as color, etc.), transaction type (person to company, company to person, company to company, person to person, etc.), transaction number and use property (lease, self-use, operation, etc.);
the vehicle sales characteristic data comprises one or more of brand grade, public praise grade, preferential information and sales information; the preferential information can comprise new car preferential information and preferential policy information, and the sales information can further comprise sales conditions (selling speed, sales amount and the like), company production and operation conditions (business amount, profit and the like), light and busy seasons and the like;
the electric power component characteristic data comprises one or more of power battery information, motor information and vehicle control unit information;
the usage habit characteristic data includes one or more of driving behavior information, charging behavior information, maintenance records, and fault records.
In one particular embodiment, the data features are divided into basic features, economic features, trilinear system features, and usage habit features. The basic features include: cities, vehicle types, models, annual fees, mileage, year and month of sale, colors, transaction types, number of times of passing households and use properties; economic features include: brand grade, public praise grade, policy preference strength, sales condition, company production and operation condition, light and busy season and new vehicle preference; the three electrical system features include: the method comprises the following steps of (1) power battery freshness, motor electric control and a vehicle control unit; usage habit features include: driving behavior, charging behavior, maintenance, service, and fault logging. A total of 25 features. The sequence is denoted as α ═ { α 1, α 2, …, α 25 }.
Before feature selection, the quantification of word-of-mouth coefficients, policy benefits, maintenance, service and fault records in this embodiment is described.
Public praise is an important mark that enterprises and products precipitate and coagulate for many years, and influence the purchase of consumers in intense market competition, and public praise also has a great influence on the value retention rate of automobile products. Taking public praise scores and user comments in the automobile website as data, extracting the theme of the LDA (Latent Dirichlet Allocation, document theme generation model), and correspondingly obtaining five grades as a brand public praise coefficient.
The maintenance record of the vehicle records a text description of repair, maintenance, and the like of the vehicle in a 4S shop since the purchase of the vehicle, and describes in detail inspection and component replacement of the vehicle, which implies vehicle condition information. After a large number of used vehicle maintenance record texts are subjected to text preprocessing, an LDA-Gibbs topic model is adopted to obtain the distribution of maintenance records corresponding to each sample vehicle on three topics of maintenance, maintenance and accident, and the distribution is used as an important calculation factor in a subsequent vehicle residual value evaluation model.
Policy preference: the pure electric vehicle enjoys subsidy and tax deduction in a purchasing link, and the subsidy amount plus the tax deduction amount are taken as quantitative representation of policy preference strength.
The depreciation of the power battery has a remarkable influence on the physical depreciation of the pure electric automobile. The reliability study of the power battery is divided into a state of charge (SOC) estimation and a state of health (SOH) estimation. The SOH is the ratio of the discharged capacity of the power battery discharged from a fully charged state to a cut-off voltage at a certain rate under standard conditions to the corresponding nominal capacity, and is the key for evaluating the health degree of the power battery. This example is characterized primarily by the SOH index. In other alternative embodiments, other criteria may be selected instead of the power cell criteria.
As shown in fig. 2, in this embodiment, after the historical transaction data of the electric vehicle is collected, the feature data is preprocessed. Specifically, after the historical data acquisition module M100 acquires historical transaction data of the electric vehicle, a plurality of pieces of historical data are obtained by sorting, where each piece of historical data includes a plurality of features and corresponding feature values. Each piece of historical data is a data sample.
After the historical data acquisition module M100 acquires the historical transaction data of the electric vehicle, the data is first collated into a format that is easy for machine learning modeling. The format is as follows: each column is a feature, and the column name is the name of the feature. The last column is the target variable, i.e., the residual rate. The storage format is preferably a csv file or an sql file. Storage using sql files is recommended if the amount of data is large. And then cleaning useless data, and processing missing values or abnormal values in the process of acquiring the data.
Missing values refer to clustering, grouping, pruning, or truncation of data in coarse data due to lack of information. It means that the value of a certain feature or features in the existing dataset is incomplete. The causes of the occurrence of the deficiency value are mainly classified into mechanical causes and artificial causes. The missing data may reduce the fit of the model or may lead to model bias.
The missing value processing method can be as follows: simple deletion is to delete the entire sample where the missing value is located. The method is simple to operate, but the number of samples is reduced; multiple interpolation, filling with mean, median, mode or adjacent values; and (3) traversing the features by using a model prediction method, assuming t features, dividing the data set into two groups, wherein the 1 st group has no missing value, the 2 nd group has a missing value, creating a prediction model by using the 1 st group, and predicting the missing data of the 2 nd group by using the model. For features with missing values less than 10%, mean or interpolation padding may be used. Features less than 1/3, model prediction is used. For discrete features, according to the characteristic of feature data (discrete special type), the discrete features can be selected from SVM, Logistic Regression and tree classification models; for continuous features, the linear regression SVR and LightGBM regression models can be selected based on the feature data characteristics. Variables with missing values above 2/3 are discarded.
Specifically, in a specific embodiment, a first threshold and a second threshold are set, the first threshold and the second threshold are respectively a specific value between 1 and 0, and the values of the first threshold and the second threshold can be set and selected as required; wherein the first threshold is greater than the second threshold.
For each piece of historical data, the evaluation model training module M200 determines whether the feature value missing proportion of the piece of historical data is greater than a first threshold, and if so, deletes the piece of historical data;
if the missing proportion of the characteristic value of the historical data is between a first threshold value and a second threshold value, a model prediction method is adopted to supplement the missing characteristic value of the historical data; the feature value missing proportion refers to the ratio of the feature quantity of the missing feature value to all the feature quantities of the historical data;
in this embodiment, the supplementing the missing feature value of the piece of historical data by using the model prediction method includes determining a type of the missing feature value of the piece of historical data, if the missing feature value is a feature value of a discrete feature, supplementing the missing feature value of the piece of historical data by using a support vector machine, Logistic regression analysis, or a tree classification model, and if the missing feature value is a feature value of a continuous feature, supplementing the missing feature value of the piece of historical data by using a linear regression support vector machine or a LightGBM regression model.
And if the missing proportion of the characteristic value of the historical data is smaller than a second threshold value, supplementing the missing characteristic value of the historical data by adopting an averaging method or an interpolation method according to the data corresponding to the missing characteristic value in other historical data. The averaging method may be to average feature values of features of the other pieces of history data corresponding to the missing feature value, and use the average as the missing feature value, and the interpolation method may be to construct a continuous function according to feature values of features of the other pieces of history data corresponding to the missing feature value, and then take corresponding points on the continuous function as the missing feature value.
In this embodiment, the evaluation model training module M200 constructs at least one residual rate evaluation model, including the evaluation model training module M200 respectively constructing a plurality of residual rate evaluation models corresponding to vehicle type categories one to one according to preset vehicle type classifications;
the evaluation model training module M200 trains the residual rate evaluation model by using the training set, including the evaluation model training module M200 classifying data in the training set according to preset vehicle type classifications, and inputting training data of different vehicle type categories into corresponding residual rate evaluation models respectively.
In this embodiment, the residual value rate evaluation module M500 inputs the feature data of the electric vehicle to be evaluated into the residual value rate evaluation model, and includes that the residual value rate evaluation module M500 determines the vehicle type category of the electric vehicle to be evaluated, and inputs the feature data of the electric vehicle to be evaluated into the residual value rate evaluation model corresponding to the vehicle type category.
The vehicle type category may be several preset categories, for example, the vehicle type may be divided into several categories such as large-sized vehicle, small-sized vehicle, and medium-sized vehicle, or other category division methods may also be adopted, and all of them fall within the protection scope of the present invention.
Further, the invention also comprises the detection and processing of abnormal values in the historical data. The abnormal value (outlier) refers to a measurement value having a deviation of more than two standard deviations from the mean value among a group of measurement values, and a measurement value having a deviation of more than three standard deviations from the mean value, and is referred to as an abnormal value of height abnormality. The causes of the disease are classified into artificial and natural factors. Outliers can dramatically change the results of data analysis and statistical modeling. Outliers in the data set have many adverse effects.
The abnormal value detection calculation method adopted in the invention can comprise the following steps: (1) visualization methods, such as boxplots (boxplots), also known as box-whisker plots, box plots, or box plots, are a statistical chart used to display a set of data scatter profiles. Abnormal values are detected using the interquartile range (IQR) of the box plot, where the interquartile range is the difference between the upper quartile and the lower quartile, and IQR that exceeds the upper quartile by 1.5 times and the lower quartile by-1.5 times is regarded as abnormal values. (2) The special abnormal value detection algorithm is, for example, One Class Support Vector Machine (SVM) (especially suitable for medium and small data detection) and Iforest (linear time complexity, suitable for processing massive data). In a specific embodiment, a boxplot is used for a preliminary test to remove samples with significant abnormalities. Outliers were analyzed synthetically using One Class SVM and Iforest on the basis of the remaining samples.
Further, as shown in fig. 2, the present invention also includes a step of screening the characteristics. The feature selection makes the model simpler and more accurate, and the good features are the most useful information extracted from the data for the prediction result. The feature selection method may include: the Filter (variance selection method, correlation coefficient method, chi-square test and mutual information method) is used for sorting and selecting the importance of the features, and is independent of a model and the like.
In this embodiment, the evaluation model training module M200 performs an exploration data analysis (Explore DataAnalysis) to analyze the interrelationship between variables. A scatter diagram of each characteristic variable and the residual value rate can be drawn, the relation between each characteristic and the residual value rate is observed, and weakly related characteristics are deleted. And calculating the variance of each feature, and deleting the features with the excessively small variances. Specifically, the evaluation model training module M200 is configured to filter features of each piece of historical data by using the following steps:
the third threshold is set to 0.1. Calculating the variance of the characteristic value of each characteristic, and if the variance of the characteristic value of a characteristic is smaller than a third threshold value after normalization processing, deleting the characteristic and the corresponding characteristic value in each historical data;
the fourth threshold is set to 0.1. And calculating the correlation between each feature and the residual value rate by adopting a Pearson coefficient correlation matrix method, namely the correlation coefficient corresponding to each feature, and deleting the feature and the corresponding feature value in each historical data if the correlation coefficient of one feature is smaller than a fourth threshold value. For example, let the correlation coefficient be p, where p e [ -1,1], the closer p is to 1, which indicates that the characteristic variable is positively correlated with the residual value rate. Otherwise, it becomes negative correlation. Features with | ρ | <0.1 are deleted. The remaining feature set α is { α 1, α 2, …, α n }, and n represents the number of features remaining after the feature is deleted.
Based on the above processing, the filtering of the important features can be further continued by using a Recursive feature elimination method (Recursive feature elimination). And (3) performing multiple rounds of training by using a base model (e.g. LightGBM) at alpha { alpha 1, alpha 2, …, alpha n }, removing the features with the minimum absolute weight after each round of training, and performing the next round of training based on the remaining feature set until the remaining features are in the required number, so that the feature screening is completed.
Specifically, in this embodiment, the evaluation model training module is further configured to filter features of each piece of historical data by using the following steps:
and performing multiple rounds of training on one base model by adopting the training set, removing the characteristic and the characteristic value with the minimum absolute weight after each round of training, and performing next round of training on the basis of the training set with the characteristic and the characteristic value removed until the residual characteristic quantity meets the preset characteristic quantity requirement.
In this embodiment, the evaluation model training module M200 selects an appropriate machine learning method to build different models according to each vehicle type data, and the machine learning method corresponding to each vehicle type data may be the same or different. As shown in fig. 3, in order to verify the accuracy of the constructed model, the evaluation model training module M200 further splits the training set obtained by the historical data obtaining module into a training set and a verification set by using a cross-validation method, trains the residual value rate evaluation model by using the split training set, and verifies the generalization performance of the trained residual value rate evaluation model by using the verification set, thereby preventing overfitting of the model.
In this embodiment, the index between the predicted value and the true value plus the regularization term is minimized by the optimization algorithm. The optimization algorithm may be represented by the following equation, where MSE represents the squared error:
the learning curve graph is drawn by adopting a 10-fold cross validation method and using a LightGBM (here, the LightGBM is taken as an example) algorithm to train a model. By observing the learning graph, it can be seen that the model is either over-fit or under-fit. Finally, a cross-validation optimal result is obtained and is marked as train _ score. If a single model fails to meet the requirements, multi-model fusion may be selected.
In this embodiment, the evaluation model training module trains the residual rate evaluation model using a regression tree model algorithm. Here, the regression tree model is taken as an example to introduce the training process of the residual value rate evaluation model of the vehicle type a. Firstly, inputting training data of a vehicle type A as follows: t { (x)1,y1),(x2,y2),…,(xN,yN) }) in which xiCharacteristic data, y, representing vehicle type AiRepresenting the residual value rate of model a. The training process is as follows:
(1) initialization f0(x) When the value is 0, initializing a first base classifier;
(2) for M is 1,2 … M;
(a) calculating a residual value rmi＝yi-fm-1(xi),i＝1,2,…,N；
(b) Residual value of fit rmiLearning a regression tree to obtain
(c) Updating
(3) Obtaining a regression problem promotion tree
Each base classifier (CART tree model) is constructed as follows: the output Y is the residual value rate, the input feature space is divided into K regions which are respectively R1,R2,…,RKThe output value of each region is: c. C1,c2,…,cm. The loss function takes:
and (3.1) traversing each feature j in turn, calculating a loss function of each dividing point (j, s) by each value s of the feature, and selecting the dividing point with the minimum loss function.
Wherein c is1,c2Are each R1,R2Average output value in interval.
And (3.2) dividing the current input space into two parts by using the segmentation points obtained in the previous step.
(3.3) then calculating the dividing point of the divided two parts again, and the like until the division cannot be continued.
(3.4) finally dividing the input space into K regions R1,R2,…,RKThe generated decision tree is:
wherein c ismIs the average of the output values of the located region.
(3.5) test and verification: and verifying the generalization performance of the model by using the test set, and testing the accuracy of the model on the residual value rate prediction. And d, predicting the data of the test set by using the cross-validation model obtained in the step d to obtain test _ score, and comparing the test _ score with the test _ score to judge whether the model is good or bad.
And (3.6) if the model in the step e meets the requirement, storing the model corresponding to the vehicle type A into an evaluation model library M300.
Each residual rate evaluation model in the evaluation model library M300 may be deployed on a local machine (such as a desktop computer or a mobile device) or a cloud computing server, and the front-end presentation may be in the form of APP (Application) or a web page. In the model using stage, the vehicle data obtaining module M400 obtains data of the vehicle to be evaluated, which is input by the user, or directly obtains sensor detection data of the vehicle through communication with a data processing device of the vehicle, then selects a suitable model in the evaluation model library M300, performs residual value prediction on the input vehicle type information, and finally outputs a vehicle residual value report.
In summary, the method adopts a big data technology, a data mining algorithm and an NLP (natural language processing technology) to model the problem of residual rate evaluation, can model under small sample data, and can fully mine the rules from massive data, thereby obtaining higher evaluation accuracy, and can also obtain higher evaluation instantaneity through real-time detection of vehicle conditions. The system for evaluating the residual value rate of the electric vehicle can provide reference for a consumer to buy the electric vehicle, a second-hand electric vehicle trading market and an electric vehicle leasing market.
As shown in fig. 4, an embodiment of the present invention further provides an electric vehicle residual value rate evaluation method, where the electric vehicle residual value rate evaluation system is adopted, and the method includes:
s100: the vehicle data acquisition module acquires characteristic data of the electric vehicle to be evaluated;
s200: the residual value rate evaluation module selects a residual value rate evaluation model from the evaluation model library;
s300: and the residual value rate evaluation module inputs the characteristic data of the electric automobile to be evaluated into the residual value rate evaluation model to obtain the residual value rate output by the residual value rate evaluation model.
Therefore, the method for evaluating the residual value rate of the electric automobile automatically predicts the residual value of the electric automobile by adopting the steps and based on a big data technology and a machine learning technology, reduces the manual detection cost, comprehensively considers the characteristic data of the electric automobile in the prediction, and predicts the residual value according to the automobile condition in real time, thereby improving the accuracy of the residual value prediction.
In this embodiment, each step in the electric vehicle residual value rate evaluation method may be specifically implemented according to a function implementation manner of each module in the electric vehicle residual value rate evaluation system, for example, step S100 may be implemented according to a specific function implementation manner of the vehicle data acquisition module M400, and step S200 and step S300 may be implemented according to a specific function implementation manner of the residual value rate evaluation module M500, which is not described herein again.
Further, the method for evaluating the residual value rate of the electric vehicle may further include the step of building an evaluation model, that is, the historical data obtaining module M100 obtains the historical transaction data of the electric vehicle, and the evaluation model training module M200 builds a residual value rate evaluation model according to the historical transaction data of the electric vehicle, and stores the residual value rate evaluation model in the evaluation model library M300. The step of constructing the evaluation model may be implemented by using the specific functional implementation manners of the historical data obtaining module M100 and the evaluation model training module M200, which are not described herein again.
The embodiment of the invention also provides electric automobile residual value rate evaluation equipment, which comprises a processor; a memory having stored therein executable instructions of the processor; wherein the processor is configured to execute the steps of the electric vehicle residual value rate evaluation method via executing the executable instructions.
As will be appreciated by one skilled in the art, aspects of the present invention may be embodied as a system, method or program product. Thus, various aspects of the invention may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" platform.
An electronic device 600 according to this embodiment of the invention is described below with reference to fig. 5. The electronic device 600 shown in fig. 5 is only an example and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 5, the electronic device 600 is embodied in the form of a general purpose computing device. The combination of the electronic device 600 may include, but is not limited to: at least one processing unit 610, at least one memory unit 620, a bus 630 connecting different platform combinations (including memory unit 620 and processing unit 610), a display unit 640, etc.
Wherein the storage unit stores program code executable by the processing unit 610 to cause the processing unit 610 to perform steps according to various exemplary embodiments of the present invention described in the above-mentioned electronic prescription flow processing method section of the present specification. For example, the processing unit 610 may perform the steps as shown in fig. 1.
The storage unit 620 may include readable media in the form of volatile memory units, such as a random access memory unit (RAM)6201 and/or a cache memory unit 6202, and may further include a read-only memory unit (ROM) 6203.
The memory unit 620 may also include programs/utilities 6204 including one or more program modules 6205, such program modules 6205 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.
The electronic device 600 may also communicate with one or more external devices 700, and the external devices 700 may be one or more of a keyboard, a pointing device, a bluetooth device, and the like. The external devices 700 enable a user to interactively communicate with the electronic device 600. The electronic device 600 may also be capable of communicating with one or more other computing devices, including routers, modems. Such communication may occur via an input/output (I/O) interface 650. Also, the electronic device 600 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network such as the Internet) via the network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via the bus 630. It should be appreciated that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, to name a few.
The embodiment of the invention also provides a computer-readable storage medium for storing a program, and the program realizes the steps of the electric vehicle residual value rate evaluation method when being executed. In some possible embodiments, aspects of the present invention may also be implemented in the form of a program product comprising program code for causing a terminal device to perform the steps according to various exemplary embodiments of the present invention described in the above-mentioned electronic prescription flow processing method section of this specification, when the program product is run on the terminal device.
Referring to fig. 6, a program product 800 for implementing the above method according to an embodiment of the present invention is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited in this regard and, in the present document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list here) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device, such as through the internet using an internet service provider.
In summary, compared with the prior art, the system, the method, the device and the medium for evaluating the residual value rate of the electric vehicle provided by the invention have the following advantages:
the method is based on the big data technology and the machine learning technology, automatically predicts the residual value of the electric automobile, reduces the manual detection cost, comprehensively considers the characteristic data of the electric automobile in the prediction, and predicts the residual value according to the automobile condition in real time, so that the accuracy of residual value prediction is improved, the method is more suitable for the electric automobile, has strong practicability, and is beneficial to popularization and application of the electric automobile.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.
Claims (15)
1. An electric vehicle residual value rate evaluation system, characterized in that the system comprises:
the historical data acquisition module is used for acquiring historical transaction data of the electric automobile and forming a training set, wherein the historical transaction data comprises characteristic data and residual value rate data of the electric automobile;
the evaluation model training module is used for constructing at least one residual value rate evaluation model based on a machine learning algorithm and training the residual value rate evaluation model by adopting the training set, wherein the input of the residual value rate evaluation model is the characteristic data of the electric automobile, and the output of the residual value rate evaluation model is the residual value rate of the electric automobile;
the evaluation model library is used for storing the trained residual value rate evaluation model;
the vehicle data acquisition module is used for acquiring the characteristic data of the electric vehicle to be evaluated;
and the residual value rate evaluation module is used for inputting the characteristic data of the electric automobile to be evaluated into the residual value rate evaluation model to obtain the residual value rate output by the residual value rate evaluation model.
2. The electric vehicle residual value rate evaluation system according to claim 1, wherein the historical data acquisition module acquires historical transaction data of an electric vehicle, and comprises the historical data acquisition module acquires historical order data, the historical order data is analyzed to obtain characteristic data and residual value rate data of the electric vehicle, and the historical data acquisition module acquires sensor detection data of the electric vehicle from an on-board information processing device of the electric vehicle.
3. The electric vehicle residual value rate evaluation system of claim 1, wherein the evaluation model training module constructs at least one residual value rate evaluation model, including the evaluation model training module respectively constructing a plurality of residual value rate evaluation models corresponding to vehicle type categories one by one according to preset vehicle type classifications;
the evaluation model training module trains the residual value rate evaluation model by adopting the training set, and comprises an evaluation model training module which classifies data in the training set according to preset vehicle type classification and inputs training data of different vehicle type categories into corresponding residual value rate evaluation models respectively.
4. The system of claim 3, wherein the residual rate evaluation module inputs the feature data of the electric vehicle to be evaluated into the residual rate evaluation model, and the system further comprises the residual rate evaluation module determines the type of the electric vehicle to be evaluated and inputs the feature data of the electric vehicle to be evaluated into the residual rate evaluation model corresponding to the type of the electric vehicle.
5. The electric vehicle residual rate evaluation system according to claim 1, wherein the characteristic data of the electric vehicle includes vehicle basic characteristic data, vehicle sales characteristic data, electric component characteristic data, and usage habit characteristic data;
the vehicle basic characteristic data comprises one or more of cities, vehicle types, models, annual money, mileage, card-serving time, appearance information, transaction types, transaction times and use properties;
the vehicle sales characteristic data comprises one or more of brand grade, public praise grade, preferential information and sales information;
the electric power component characteristic data comprises one or more of power battery information, motor information and vehicle control unit information;
the usage habit characteristic data includes one or more of driving behavior information, charging behavior information, maintenance records, and fault records.
6. The system of claim 1, wherein the historical data obtaining module obtains historical transaction data of the electric vehicle and then arranges the historical transaction data into a plurality of pieces of historical data, and each piece of historical data comprises a plurality of features and corresponding feature values.
7. The electric vehicle residual rate evaluation system according to claim 6, wherein a first threshold value and a second threshold value are set to be a specific value between 1 and 0; the evaluation model training module judges whether the feature value missing proportion of each piece of historical data is larger than a first threshold value or not for each piece of historical data, and if so, deletes the piece of historical data;
if the missing proportion of the characteristic value of the historical data is between a first threshold value and a second threshold value, a model prediction method is adopted to supplement the missing characteristic value of the historical data;
and if the missing proportion of the characteristic value of the historical data is smaller than a second threshold value, supplementing the missing characteristic value of the historical data by adopting an averaging method or an interpolation method according to the data corresponding to the missing characteristic value in other historical data.
8. The system of claim 7, wherein the supplementing the missing feature value of the historical data using the model prediction method comprises determining a type of the missing feature value of the historical data, and if the missing feature value is a discrete feature value, supplementing the missing feature value of the historical data using a support vector machine, Logistic regression analysis, or a tree classification model, and if the missing feature value is a continuous feature value, supplementing the missing feature value of the historical data using a linear regression support vector machine or a LightGBM regression model.
9. The electric vehicle residual rate evaluation system of claim 6, wherein the evaluation model training module is further configured to filter the characteristics of each historical datum by:
calculating the variance of the feature value of each feature, and if the variance of the feature value of a feature is smaller than 0.1 after normalization processing, deleting the feature and the corresponding feature value in each piece of historical data;
and calculating the correlation between each feature and the residual value rate by adopting a Pearson coefficient correlation matrix method, namely the correlation coefficient corresponding to each feature, and deleting the feature and the corresponding feature value in each historical data if the correlation coefficient of one feature is less than 0.1.
10. The system of claim 9, wherein the evaluation model training module is further configured to filter the characteristics of each historical datum by:
and performing multiple rounds of training on one base model by adopting the training set, removing the characteristic and the characteristic value with the minimum absolute weight after each round of training, and performing next round of training on the basis of the training set with the characteristic and the characteristic value removed until the residual characteristic quantity meets the preset characteristic quantity requirement.
11. The electric vehicle residual rate evaluation system of claim 1, wherein the evaluation model training module further splits the training set obtained by the historical data acquisition module into a training set and a validation set by using a cross validation method, trains the residual rate evaluation model by using the split training set, and verifies the generalization performance of the trained residual rate evaluation model by using the validation set.
12. The electric vehicle residual value rate evaluation system according to claim 1, wherein the evaluation model training module trains the residual value rate evaluation model using a regression tree model algorithm.
13. An electric vehicle residual value rate evaluation method, characterized in that the electric vehicle residual value rate evaluation system of any one of claims 1 to 12 is adopted, the method comprising:
the vehicle data acquisition module acquires characteristic data of the electric vehicle to be evaluated;
the residual value rate evaluation module selects a residual value rate evaluation model from the evaluation model library;
and the residual value rate evaluation module inputs the characteristic data of the electric automobile to be evaluated into the residual value rate evaluation model to obtain the residual value rate output by the residual value rate evaluation model.
14. An electric vehicle residual value rate evaluation apparatus, characterized by comprising:
a processor;
a memory having stored therein executable instructions of the processor;
wherein the processor is configured to perform the steps of the electric vehicle residual value rate evaluation method of claim 13 via execution of the executable instructions.
15. A computer-readable storage medium storing a program, wherein the program is executed to implement the steps of the electric vehicle remaining value rate evaluation method according to claim 13.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910984012.5A CN110706039A (en) | 2019-10-16 | 2019-10-16 | Electric vehicle residual value rate evaluation system, method, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910984012.5A CN110706039A (en) | 2019-10-16 | 2019-10-16 | Electric vehicle residual value rate evaluation system, method, equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110706039A true CN110706039A (en) | 2020-01-17 |
Family
ID=69199975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910984012.5A Pending CN110706039A (en) | 2019-10-16 | 2019-10-16 | Electric vehicle residual value rate evaluation system, method, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110706039A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523945A (en) * | 2020-04-27 | 2020-08-11 | 众能联合数字技术有限公司 | Computing system for electrically driving and leasing high-altitude vehicle residual value based on data of Internet of things |
| CN111624433A (en) * | 2020-07-08 | 2020-09-04 | 深圳技术大学 | State evaluation method and system for pure electric vehicle and computer readable storage medium |
| CN111859294A (en) * | 2020-07-09 | 2020-10-30 | 北理新源(佛山)信息科技有限公司 | Electric vehicle evaluation method and system based on big data |
| CN112016964A (en) * | 2020-08-27 | 2020-12-01 | 李忠耘 | Second-hand vehicle dynamic pricing method and device, electronic equipment and storage medium |
| CN112257224A (en) * | 2020-09-11 | 2021-01-22 | 上海发电设备成套设计研究院有限责任公司 | Method, system and terminal for overhauling state of steam turbine generator |
| CN112508213A (en) * | 2020-12-25 | 2021-03-16 | 武汉理工大学 | Method and equipment for evaluating residual value of running pure electric automobile |
| CN112950030A (en) * | 2021-03-02 | 2021-06-11 | 爱驰汽车有限公司 | Residual error evaluation method and device for electric vehicle, electronic equipment and storage medium |
| CN113052633A (en) * | 2021-03-26 | 2021-06-29 | 中国第一汽车股份有限公司 | Vehicle residual value evaluation method, device, equipment and medium |
| CN113724417A (en) * | 2020-05-22 | 2021-11-30 | 丰田自动车株式会社 | Condition evaluation system and condition evaluation method for vehicle-mounted equipment |
| CN116125325A (en) * | 2022-12-06 | 2023-05-16 | 北汽福田汽车股份有限公司 | Method and device for detecting consistency of battery cells, vehicle and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106204138A (en) * | 2016-07-11 | 2016-12-07 | 上海安吉星信息服务有限公司 | A kind of vehicle salvage value rate appraisal procedure and device |
| CN108154275A (en) * | 2017-12-29 | 2018-06-12 | 广东数鼎科技有限公司 | Automobile residual value prediction model and Forecasting Methodology based on big data |
| CN108564415A (en) * | 2018-04-25 | 2018-09-21 | 济南浪潮高新科技投资发展有限公司 | A kind of method of intelligent predicting used car price |
| CN109767251A (en) * | 2017-11-09 | 2019-05-17 | 优估（上海）信息科技有限公司 | A kind of used car valuation methods and appraisal system |
-
2019
- 2019-10-16 CN CN201910984012.5A patent/CN110706039A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106204138A (en) * | 2016-07-11 | 2016-12-07 | 上海安吉星信息服务有限公司 | A kind of vehicle salvage value rate appraisal procedure and device |
| CN109767251A (en) * | 2017-11-09 | 2019-05-17 | 优估（上海）信息科技有限公司 | A kind of used car valuation methods and appraisal system |
| CN108154275A (en) * | 2017-12-29 | 2018-06-12 | 广东数鼎科技有限公司 | Automobile residual value prediction model and Forecasting Methodology based on big data |
| CN108564415A (en) * | 2018-04-25 | 2018-09-21 | 济南浪潮高新科技投资发展有限公司 | A kind of method of intelligent predicting used car price |
Non-Patent Citations (3)
| Title |
|---|
| JOE-HAN: "数据预处理与特征选择", 《HTTPS://BLOG.CSDN.NET/U010089444/ARTICLE/DETAILS/70053104》 * |
| 刘丽丽: "《个体学习和知识共享对团队结构及成员创新行为的影响研究》", 30 November 2017, 中国经济出版社 * |
| 陈宸: "如何破除新能源汽车残值评估的困境", 《HTTPS://WWW.SOHU.COM/A/249295634_100039896》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523945A (en) * | 2020-04-27 | 2020-08-11 | 众能联合数字技术有限公司 | Computing system for electrically driving and leasing high-altitude vehicle residual value based on data of Internet of things |
| CN113724417A (en) * | 2020-05-22 | 2021-11-30 | 丰田自动车株式会社 | Condition evaluation system and condition evaluation method for vehicle-mounted equipment |
| CN113724417B (en) * | 2020-05-22 | 2023-05-02 | 丰田自动车株式会社 | Condition evaluation system and condition evaluation method for in-vehicle apparatus |
| CN111624433A (en) * | 2020-07-08 | 2020-09-04 | 深圳技术大学 | State evaluation method and system for pure electric vehicle and computer readable storage medium |
| WO2022007236A1 (en) * | 2020-07-08 | 2022-01-13 | 深圳技术大学 | Battery electric vehicle state evaluation method and system, and computer-readable storage medium |
| CN111859294A (en) * | 2020-07-09 | 2020-10-30 | 北理新源(佛山)信息科技有限公司 | Electric vehicle evaluation method and system based on big data |
| CN112016964A (en) * | 2020-08-27 | 2020-12-01 | 李忠耘 | Second-hand vehicle dynamic pricing method and device, electronic equipment and storage medium |
| CN112257224B (en) * | 2020-09-11 | 2022-12-30 | 上海发电设备成套设计研究院有限责任公司 | Method, system and terminal for overhauling state of steam turbine generator |
| CN112257224A (en) * | 2020-09-11 | 2021-01-22 | 上海发电设备成套设计研究院有限责任公司 | Method, system and terminal for overhauling state of steam turbine generator |
| CN112508213A (en) * | 2020-12-25 | 2021-03-16 | 武汉理工大学 | Method and equipment for evaluating residual value of running pure electric automobile |
| CN112950030A (en) * | 2021-03-02 | 2021-06-11 | 爱驰汽车有限公司 | Residual error evaluation method and device for electric vehicle, electronic equipment and storage medium |
| CN112950030B (en) * | 2021-03-02 | 2024-02-06 | 爱驰汽车有限公司 | Residual error assessment method and device for electric automobile, electronic equipment and storage medium |
| CN113052633A (en) * | 2021-03-26 | 2021-06-29 | 中国第一汽车股份有限公司 | Vehicle residual value evaluation method, device, equipment and medium |
| CN116125325A (en) * | 2022-12-06 | 2023-05-16 | 北汽福田汽车股份有限公司 | Method and device for detecting consistency of battery cells, vehicle and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110706039A (en) | Electric vehicle residual value rate evaluation system, method, equipment and medium | |
| US20150221040A1 (en) | Residual risk analysis system, method and computer program product therefor | |
| US11367142B1 (en) | Systems and methods for clustering data to forecast risk and other metrics | |
| US11367141B1 (en) | Systems and methods for forecasting loss metrics | |
| CN112116184A (en) | Factory risk estimation using historical inspection data | |
| US20190378180A1 (en) | Method and system for generating and using vehicle pricing models | |
| CN111079941B (en) | Credit information processing method, credit information processing system, terminal and storage medium | |
| US20170270546A1 (en) | Service churn model | |
| CN115526652A (en) | Client loss early warning method and system based on machine learning | |
| CN111626855A (en) | Bond credit interest difference prediction method and system | |
| CN113674040A (en) | Vehicle quotation method, computer device and computer-readable storage medium | |
| Hu | Predicting and improving invoice-to-cash collection through machine learning | |
| CN112116185A (en) | Test risk estimation using historical test data | |
| CN114118793A (en) | Local exchange risk early warning method, device and equipment | |
| CN112037006A (en) | Credit risk identification method and device for small and micro enterprises | |
| Gleue et al. | Decision support for the automotive industry: Forecasting residual values using artificial neural networks | |
| US11935075B2 (en) | Card inactivity modeling | |
| CN115809930A (en) | Anti-fraud analysis method, device, equipment and medium based on data fusion matching | |
| CN111401329B (en) | Information flow direction identification method, device, equipment and storage medium | |
| CN114612239A (en) | Stock public opinion monitoring and wind control system based on algorithm, big data and artificial intelligence | |
| CN113065683A (en) | Price prediction method, device, equipment and storage medium for vehicle pledge | |
| Kim et al. | Trustworthy residual vehicle value prediction for auto finance | |
| Hemendiran et al. | Predicting the Prices of the Used Cars using Machine Learning for Resale | |
| Collard | Price prediction for used cars: a comparison of machine learning regression models | |
| CN111047438A (en) | Data processing method, device and computer readable storage medium |
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 |
Application publication date: 20200117 |
|
| RJ01 | Rejection of invention patent application after publication |