Insurance evaluation method and system for new energy automobile and storage medium

Abstract

The invention belongs to the technical field of data processing, and particularly relates to an insurance evaluation method and system for a new energy automobile and a storage medium, and the method comprises the steps: collecting the driving record data and vehicle data of a driver through a vehicle-mounted terminal of the new energy automobile, collecting the feature data of the driver, verifying the data through a data verification module, and carrying out the verification of the data. Deleting error data, performing first analysis and second analysis on the driving record data, the vehicle data and the driver feature data to obtain analysis results, inputting the analysis results as input data into a machine learning model for model training, and generating a risk assessment model; the risk assessment value is obtained based on the target driving record data, the target vehicle data, the target driver feature data and the risk assessment model, the vehicle loss rate and the target vehicle total score are also obtained, the insurance premium is set based on the risk assessment value, the vehicle loss rate and the target vehicle total score, and the more reasonable insurance premium can be set.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and in particular relates to an insurance assessment method, system and storage medium for new energy vehicles. Background Art

[0002] With the rapid development of new energy vehicles, there are significant differences in structure and usage characteristics between them and traditional fuel vehicles, such as battery life, charging mode, energy consumption, etc. These factors have an important impact on insurance assessment. However, the existing insurance assessment method fails to fully consider these factors, resulting in inaccurate insurance pricing. A similar prior art Chinese patent application with publication number CN117893333A provides a new energy vehicle insurance method, medium and device based on historical driving data, the method comprising: obtaining historical driving behavior data of a target driver in a plurality of consecutive set time periods before a target time period; obtaining historical driving habit data and historical driving skill level of the target driver in each set time period according to each of the historical driving behavior data; predicting the expected driving habit data and expected driving skill level of the target driver in the target time period according to each of the historical driving skill levels and each of the historical driving habit data when each of the historical driving skills does not meet the preset technical stability conditions; obtaining the vehicle attribute information of the target vehicle, and using the UBI insurance model, obtaining the insurance pricing strategy of the target vehicle in the target time period according to the vehicle attribute information, the expected driving skill level and the expected driving habit data. However, this method does not consider the vehicle loss rate and battery loss rate when formulating the insurance pricing strategy. Similar prior art also includes a Chinese patent application with publication number CN117893333A, which discloses a method, device and system for constructing a car insurance pricing model based on vehicle mileage. The method includes: obtaining the original data of the vehicle in a historical time period; processing the original data to obtain a single mileage factor and a dependent variable; constructing multiple mileage interval factors based on the single mileage factor; multiple mileage interval factors are used as independent variables and are placed in a preset model together with the dependent variable to establish a mileage pricing model. However, this method only considers a single mileage factor when establishing the pricing model, and does not consider other factors that affect the premium, such as vehicle loss. Therefore, the present invention provides an insurance assessment method, system and storage medium for new energy vehicles. Summary of the invention

[0003] The present invention collects data through a data collection module, verifies the data through a data verification module to delete erroneous data, trains a risk assessment model through a model training module, obtains a risk assessment value through the risk assessment model, and also obtains a vehicle loss rate and a vehicle total score, and sets a reasonable premium based on the risk assessment value, vehicle loss rate and vehicle total score.

[0004] In order to achieve the above-mentioned purpose of the invention, the present invention provides the following insurance assessment method for new energy vehicles, which is mainly implemented by executing the following steps:

[0005] Collecting the driver's driving record data and vehicle data through the vehicle terminal of the new energy vehicle, and also collecting the driver's characteristic data, the driving record data at least includes mileage data, speed data, risk record, driving time data, acceleration and deceleration data, regional data and weather data, and the vehicle data at least includes vehicle model, engine data, vehicle fault data and battery data;

[0006] Preprocessing the driving record data to remove invalid and erroneous data, converting the driving record data, the vehicle data and the driver characteristic data into data in a standardized format, acquiring key data from other information providing terminals, acquiring relevant data corresponding to the key data from the driving record data, comparing the key data and the relevant data to determine whether they are the same, if they are the same, retaining the relevant data, if they are different, deleting the relevant data from the driving record data;

[0007] Performing a first analysis on the driving record data, the vehicle data, and the driver characteristic data to obtain risk characteristic data items related to risk assessment, using a second analysis method to analyze each of the risk characteristic data items to obtain an analysis result, and inputting the analysis result as input data into a machine learning model to perform model training and generate a risk assessment model;

[0008] Use the model training module to perform a first analysis and a second analysis on the target driving record data, the target vehicle data and the target driver characteristic data to obtain corresponding target risk characteristic data, input the target risk characteristic data into the risk assessment model to obtain a risk assessment value, obtain the vehicle loss rate of the target vehicle based on the target vehicle data, obtain the target vehicle total score based on the target vehicle's driving route history data, and set insurance premiums based on the risk assessment value, the vehicle loss rate and the target vehicle total score.

[0009] As a preferred technical solution of the present invention, performing a first analysis on the driving record data, the vehicle data and the driver characteristic data comprises the following steps:

[0010] Perform statistical analysis on each trip, calculate the statistical data amount of each risk characteristic data item, obtain the risk record therein based on the driving record data, obtain all first data items corresponding to the time point when the risk record occurs, obtain a time series distribution graph of the first data item, obtain the time point of the first data item in the time series distribution graph, obtain multiple second data items within a preset time period before the time point, obtain multiple third data items within a preset time period after the time point, calculate the average value of the first data item, multiple second data items and multiple third data items as the average value, calculate the standard deviation based on the average value, obtain the trip standard deviation corresponding to the risk record based on the statistical data amount, obtain the difference between the standard deviation and the trip standard deviation, and if the difference is greater than or equal to a preset first threshold, use the first data item as the risk characteristic data item corresponding to the risk record, obtain all the risk characteristic data items corresponding to the risk record, and associate the risk record with the risk characteristic data item and save them.

[0011] As a preferred technical solution of the present invention, using the second analysis method to analyze each of the risk characteristic data items comprises the following steps:

[0012] A rectangular coordinate system is established, with the time point as the horizontal axis and each of the risk characteristic data items as the vertical axis, a scatter plot of each of the risk characteristic data items is obtained, two adjacent points are connected to each other based on the scatter plot to obtain a line graph, the line graph is fitted to obtain a corresponding curve graph, and based on the curve graph, it is determined whether the risk characteristic data items have the same change trend. If so, the correlation coefficient between each of the risk characteristic data items is calculated using a correlation coefficient calculation method, and the absolute values ​​of all the correlation coefficients are obtained. The absolute values ​​of all the correlation coefficients are divided into three levels according to a preset second threshold and a third threshold, and those greater than or equal to the second threshold are divided into a first level, those less than the second threshold and greater than or equal to the third threshold are divided into a second level, and those less than the third threshold are divided into a third level. The risk characteristic data items corresponding to the correlation coefficients of the first level are collectively referred to as first risk characteristic data items, the risk characteristic data items corresponding to the correlation coefficients of the second level are collectively referred to as second risk characteristic data items, and the risk characteristic data items corresponding to the correlation coefficients of the third level are collectively referred to as third risk characteristic data items.

[0013] As a preferred technical solution of the present invention, obtaining the vehicle loss rate of the target vehicle based on the target vehicle data includes the following steps:

[0014] Obtain battery data of a target vehicle, calculate first data of the battery based on the battery data, the first data refers to the ratio of the current storage capacity of the battery when it is fully charged to the storage capacity of the battery when it leaves the factory, obtain battery-related data of the target vehicle, the battery data includes the use environment of the vehicle, the initial storage capacity of the battery, the mileage data of the vehicle, the charge and discharge cycle of the battery, and the storage capacity of the battery after each charge and discharge, calculate the first data of the battery after each charge and discharge, use the battery-related data and the corresponding first data as learning data, train the learning data to generate a battery evaluation model, use the battery evaluation model to estimate the loss rate of the battery after a preset time, obtain relevant data of key components of the target vehicle, obtain the loss rate of each key component based on the relevant data of the key components, and calculate the loss rate of the target vehicle using a first formula based on the loss rate of the battery and the loss rate of each key component, the first formula is: Where B is the battery loss rate, W b is the weight of the battery, N is the total number of key components, C i is the loss rate of the i-th key component, W i is the weight of the i-th key component.

[0015] As a preferred technical solution of the present invention, obtaining the total score of the target vehicle based on the driving route history data of the target vehicle includes the following steps:

[0016] Obtain the driving route history data of the target vehicle, perform statistical analysis on the driving route history data, obtain the accident-prone areas of the corresponding route, pre-store an accident information storage table, the accident information storage table stores the accident location, accident type and accident level, compare the driving route history data with the accident location in the accident information storage table, determine whether the target vehicle has passed the accident location in the accident information storage table, and if so, set a vehicle score for the target vehicle based on the accident level, and after comparing all the accident locations in the accident information storage table, obtain the total vehicle score of the target vehicle.

[0017] As a preferred technical solution of the present invention, setting a vehicle score for the target vehicle based on the accident level includes the following steps:

[0018] Preset multiple levels of accident levels, and set corresponding scores for each of the accident levels. Initialize the total vehicle score of the target vehicle to zero. Based on the driving route history data, determine whether the target vehicle has had an accident at the accident site. If not, obtain the weather conditions and time period when the target vehicle passes the accident site. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, increase the score corresponding to the accident level for the target vehicle by half, otherwise increase the score corresponding to the accident level for the target vehicle. If yes, obtain the weather conditions and time period when the target vehicle passes the accident. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, subtract the score corresponding to the accident level from the target vehicle, otherwise subtract half of the score corresponding to the accident level from the target vehicle.

[0019] The present invention also provides an insurance assessment system for new energy vehicles, comprising the following modules:

[0020] A data collection module, used to collect the driver's driving record data and vehicle data through the vehicle terminal of the new energy vehicle, and also collect the driver's characteristic data, the driving record data at least includes mileage data, speed data, risk record, driving time data, acceleration and deceleration data, regional data and weather data, and the vehicle data at least includes vehicle model, engine data, vehicle fault data and battery data;

[0021] a data verification module, used to pre-process the driving record data, remove invalid and erroneous data, convert the driving record data, the vehicle data and the driver characteristic data into data in a standardized format, obtain key data from other information providing terminals, obtain relevant data corresponding to the key data from the driving record data, compare the key data and the relevant data, determine whether they are the same, if they are the same, retain the relevant data, if they are different, delete the relevant data in the driving record data;

[0022] A model training module, configured to perform a first analysis on the driving record data, the vehicle data, and the driver characteristic data to obtain risk characteristic data items related to risk assessment, analyze each of the risk characteristic data items using a second analysis method to obtain an analysis result, and input the analysis result as input data into a machine learning model to perform model training and generate a risk assessment model;

[0023] A premium setting module is used to use the model training module to perform a first analysis and a second analysis on the target driving record data, the target vehicle data and the target driver characteristic data to obtain corresponding target risk characteristic data, input the target risk characteristic data into the risk assessment model, obtain a risk assessment value, obtain the vehicle loss rate of the target vehicle based on the target vehicle data, obtain the total score of the target vehicle based on the driving route history data of the target vehicle, and set the insurance premium based on the risk assessment value, the vehicle loss rate and the total score of the target vehicle.

[0024] The present invention also provides a storage medium, which stores program instructions, wherein when the program instructions are executed, the device where the storage medium is located is controlled to execute any one of the above-mentioned insurance assessment methods for new energy vehicles.

[0025] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0026] In the present invention, firstly, all relevant data required for use are collected through the data collection module to provide sufficient data support for subsequent data analysis and model training. For data collected from different terminals, loss or errors may occur during the transmission process, so the collected data is verified by the data verification module to ensure the accuracy of the data. After that, the collected relevant data is fully analyzed, and the analysis results are trained using a machine learning algorithm to generate a risk assessment model that can reflect the risk, so that driving habits, driving behaviors, etc. can be fully analyzed to obtain an accurate risk assessment model. Finally, by using the insurance setting module, the insurance premium is set by comprehensively considering multiple factors such as driving behavior, loss rate of the target vehicle, and accident occurrence of the target vehicle, so that the insurance premium is more in line with the actual situation and more personalized. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A flowchart of the steps of the insurance assessment method for new energy vehicles of the present invention;

[0028] Figure 2 The figure is a structural diagram of the insurance assessment system for new energy vehicles of the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] It is understood that the terms "first", "second", etc. used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, without departing from the scope of this application, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script.

[0031] The present invention provides Figure 1 The insurance assessment method for new energy vehicles shown is mainly implemented by executing the following steps:

[0032] Step S1, collecting the driver's driving record data and vehicle data through the on-board terminal of the new energy vehicle, and also collecting the driver's characteristic data, the driving record data at least includes mileage data, speed data, risk record, driving time data, acceleration and deceleration data, regional data and weather data, and the vehicle data at least includes vehicle model, engine data, vehicle fault data and battery data.

[0033] Specifically, in order to design more reasonable insurance premiums, the vehicle terminal is used to collect driving record data of each vehicle trip. Vehicle data is also collected. Mileage information refers to the total mileage data of the driver driving the vehicle, mileage data within a preset time period, average mileage data of each trip, maximum mileage data and minimum mileage data. Speed ​​information includes the maximum driving speed, minimum driving speed and average driving speed of the driver's vehicle. Risk records include accident records, violation records and insurance payment records, etc. Violation data refers to all violation records in the historical driving data, such as violation of parking signs, speed limit violations and traffic light violations, etc. Driving time data refers to the time when the driver usually drives the vehicle. The time period of the day, for example, whether the vehicle is in the morning or evening rush hour or other time periods; acceleration and deceleration data refers to the number of operations such as rapid acceleration, rapid deceleration, sharp turns and emergency stops; regional data refers to the main driving sections and regional information of the vehicle, such as expressways, national roads and mountain roads; weather data refers to the weather data when the vehicle is driving, such as temperature, humidity, rain, ice and water accumulation; vehicle data includes various data related to the vehicle, among which battery data includes battery usage time, usage temperature, current and voltage, etc.; driver characteristic data includes the driver's gender, age, occupation and place of residence, etc. The above method is used to collect various data used by the prediction model.

[0034] Step S2: pre-process the driving record data to remove invalid and erroneous data, convert the driving record data, vehicle data and driver characteristic data into data in a standardized format, obtain key data from other information providing terminals, obtain relevant data corresponding to the key data from the driving record data, compare the key data and the relevant data, and determine whether they are the same. If they are the same, retain the relevant data; if they are different, delete the relevant data in the driving record data.

[0035] Specifically, the data is first preprocessed to delete obvious invalid and erroneous data such as missing data and abnormal data values. Standardized format refers to normalization processing to facilitate subsequent analysis. In the process of data generation and transmission, inaccurate data may occur due to sensor failure, communication interference, network delay or data packet loss. In order to ensure the accuracy of data collection, key data of driving data, such as driving speed, driving distance, etc., are obtained from other information providing terminals, such as traffic information providing terminals, and relevant data corresponding to the key data in the driving record data are obtained. By comparing data from different sources, the collected data is verified and data that may be erroneous is discarded to ensure the accuracy and reliability of the collected driving record data, providing an accurate data basis for subsequent analysis steps.

[0036] Step S3: perform a first analysis on the driving record data, vehicle data and driver characteristic data to obtain risk characteristic data items related to risk assessment, use a second analysis method to analyze each risk characteristic data item to obtain an analysis result, and input the analysis result as input data into a machine learning model for model training to generate a risk assessment model.

[0037] Specifically, in order to make full use of the vehicle's driving information in the calculation of insurance premiums so as to achieve the purpose of providing personalized insurance products, a first analysis is performed based on the driving record data, vehicle data and driver characteristic data to obtain risk characteristic data items related to risk assessment, such as vehicle speed, acceleration, driving road condition information and driving time period. Thereafter, a second analysis method is used to further analyze the risk characteristic data items to obtain the correlation between the risk characteristic data items. A machine learning algorithm training model is used to perform a multi-dimensional analysis of driving behavior to generate a risk assessment model. The risk assessment model can make full use of driving record data to analyze driving behavior and assess driving risks, and can more accurately assess the driver's risk level. Subsequently, a driving risk assessment value is obtained based on the risk assessment model, providing a basis for setting fairer and more reasonable premiums in the future.

[0038] Step S4, use the model training module to perform a first analysis and a second analysis on the target driving record data, the target vehicle data and the target driver characteristic data to obtain corresponding target risk characteristic data, input the target risk characteristic data into the risk assessment model, obtain a risk assessment value, obtain the vehicle loss rate of the target vehicle based on the target vehicle data, obtain the target vehicle total score based on the target vehicle's driving route history data, and set the insurance premium based on the risk assessment value, the vehicle loss rate and the target vehicle total score.

[0039] Specifically, after the risk assessment model is generated, the risk assessment model is used to evaluate the risk assessment value of the target driver. The higher the risk assessment value, the higher the premium should be set when setting the premium. If the risk assessment value is low, it means that the probability of an accident is also lower. In this case, a relatively low premium should be set for the user when setting the premium. In order to set a more realistic and reasonable premium, the target vehicle has a larger loss, and the possibility of failure and accident is relatively greater. Therefore, the loss rate of the target vehicle is also obtained. The loss rate of the target vehicle is based on the target vehicle. A more accurate reference is provided for insurance pricing. The higher the loss rate, the higher the corresponding premium should be set. In addition, the historical data of the target vehicle's driving route is also considered. The driving history that can reflect the actual accident is obtained based on the historical data of the driving route to set the premium. By considering factors such as the vehicle's driving route, the accident site, the accident level, and the weather conditions when passing the accident site, if the vehicle passes through a high-incidence area without an accident, the corresponding vehicle score is higher, indicating that the vehicle's driving skills are relatively more stable and safer, so a lower premium should be set.

[0040] In the above method, firstly, all relevant data needed are collected through the data collection module to provide sufficient data support for subsequent data analysis and model training. For data collected from different terminals, loss or errors may occur during the transmission process, so the collected data is verified by the data verification module to ensure the accuracy of the data. After that, the collected relevant data is fully analyzed, and the analysis results are trained using a machine learning algorithm to generate a risk assessment model that can reflect the risk, so that driving habits, driving behaviors, etc. can be fully analyzed to obtain an accurate risk assessment model. Finally, by using the insurance setting module, the driving behavior, the loss rate of the target vehicle and the accident occurrence of the target vehicle are comprehensively considered to set the insurance premium more accurately, so that the insurance premium is more in line with the actual situation, while encouraging safe driving.

[0041] Further, a first analysis is performed on the driving record data, the vehicle data and the driver characteristic data, including the following steps:

[0042] Perform statistical analysis on each trip, calculate the statistical data amount of each risk characteristic data item, obtain the risk record in the driving record data based on the driving record data, obtain all first data items corresponding to the time point when the risk record occurs, obtain a time series distribution graph of the first data item, obtain the time point of the first data item in the time series distribution graph, obtain multiple second data items within a preset time period before the time point, obtain multiple third data items within a preset time period after the time point, calculate the average value of the first data item, multiple second data items and multiple third data items as the average value, calculate the standard deviation based on the average value, obtain the trip standard deviation corresponding to the risk record based on the statistical data amount, obtain the difference between the standard deviation and the trip standard deviation, if the difference is greater than or equal to a preset first threshold, use the first data item as the risk characteristic data item corresponding to the risk record, obtain all risk characteristic data items corresponding to the risk record, and associate the risk record with the risk characteristic data item and save them.

[0043] Specifically, firstly, a statistical analysis is performed on the driving record data of each trip, and the statistical data quantity of each data item in the driving record data corresponding to the trip is calculated, and the statistical data quantity includes the mean, variance and standard deviation. In order to obtain data items related to risk assessment, usually, a violation of the vehicle indicates that there is a high possibility of insurance compensation, so the risk record is obtained based on the driving record data, and then all the first data items when the risk record occurs are obtained based on the time point of the risk record. In order to determine whether the first data item is related to the occurrence of the risk, a time series distribution graph of the first data item is obtained, and based on the time series distribution graph, the time point of the first data item in the time series distribution graph is obtained, and multiple data items before and after the time point are obtained. Taking the vehicle speed as an example, it is assumed that the time point when the risk occurs corresponds to the first vehicle speed (corresponding to the first data item), and the second vehicle speed within 5 minutes before the time point (corresponding to the second data item) is obtained to obtain Take the third vehicle speed (corresponding to the third data item) within 5 minutes after this time point. If the vehicle speed is collected every half minute, 10 second vehicle speeds and 10 third vehicle speeds can be obtained. Then, the standard deviation of the first, second and third vehicle speeds is calculated, as well as the trip standard deviation of the corresponding data item. The calculated standard deviation and the trip standard deviation are compared by the above method. If the difference between the two is greater than the preset first threshold, it means that the current first data item is significantly different from the data item of the entire trip, and the correlation with the current risk record may be relatively large. Therefore, the corresponding first data item is used as the risk characteristic data item of the risk record. There may be multiple risk characteristic data items corresponding to the risk record. The same method is used to obtain all risk characteristic data items corresponding to the risk record, and the risk record and the corresponding risk characteristic data item are associated and saved to provide a basis for subsequent analysis. The above method is used to preliminarily determine which data items may be related to the risk record.

[0044] Further, using the second analysis method to analyze the first relationship between the risk feature data items includes the following steps:

[0045] A rectangular coordinate system is established with the time point as the horizontal axis and each risk characteristic data item as the vertical axis, a scatter plot of each risk characteristic data item is obtained, and based on the scatter plot, two adjacent points are connected to obtain a line graph, and the line graph is fitted to obtain the corresponding curve graph. Based on the curve graph, it is determined whether the risk characteristic data items have the same change trend. If so, the correlation coefficient calculation method is used to calculate the correlation coefficient between each risk characteristic data item, and the absolute values ​​of all correlation coefficients are obtained. The absolute values ​​of all correlation coefficients are divided into three levels according to the preset second threshold and third threshold, and those greater than or equal to the second threshold are divided into the first level, those less than the second threshold and greater than or equal to the third threshold are divided into the second level, and those less than the third threshold are divided into the third level. The risk characteristic data items corresponding to the correlation coefficients of the first level are collectively referred to as first risk characteristic data items, the risk characteristic data items corresponding to the correlation coefficients of the second level are collectively referred to as second risk characteristic data items, and the risk characteristic data items corresponding to the correlation coefficients of the third level are collectively referred to as third risk characteristic data items.

[0046] Specifically, in order to further analyze the relationship between each risk feature data item and the risk record, first use visualization technology to obtain the correlation between each risk feature data item. By establishing a rectangular coordinate system, with the time point as the horizontal axis and each risk feature data item as the vertical axis, the risk feature data items are marked in the rectangular coordinate system to obtain a scatter plot. For more intuitive observation, the two scatter plots of adjacent time points are connected to each other to obtain a line graph. The corresponding curve graph is obtained by fitting the line graph. Based on the changing trend of the curve graph, it is judged whether the risk feature data items have the same changing trend. If the risk feature data items have the same changing trend, it proves that there is a high possibility of correlation between them. All risk feature data items with the same changing trend are obtained, and the correlation coefficient between the risk feature data items is calculated using the correlation coefficient calculation method. The specific correlation coefficient calculation method includes the Pearson correlation coefficient calculation method, etc., and the corresponding After calculating the correlation coefficient, the absolute values ​​of all correlation coefficients are obtained, and the correlation coefficients are graded according to the preset second threshold and third threshold. For example, the second threshold can be set to 0.8, the third threshold can be set to 0.5, the absolute value of the correlation coefficient greater than or equal to 0.8 is set to the first level, the absolute value of the correlation coefficient less than 0.8 and greater than or equal to 0.5 is set to the second level, and the absolute value of the correlation coefficient less than 0.5 is set to the third level. The risk characteristic data items corresponding to the correlation coefficients of the first level are obtained and collectively referred to as first risk characteristic data items. The correlation coefficients between the first risk characteristic data items are large, so they have a high correlation. The correlation coefficients between the second risk characteristic data items are relatively small, so the correlation between the second risk characteristic data items is not too high or too low. The correlation coefficients between the third risk characteristic data items are low, so the correlation between the third risk characteristic data items is also low.

[0047] The risk feature data items are divided into three levels through the above method. Subsequently, the model can be trained separately based on the risk feature data of each level, so that the model training can make full use of the combination of different data items and improve the performance and accuracy of the model.

[0048] Further, obtaining the vehicle loss rate of the target vehicle based on the target vehicle data includes the following steps:

[0049] Obtain the battery data of the target vehicle, calculate the first data of the battery based on the battery data, the first data refers to the ratio of the current storage capacity of the battery when it is fully charged to the storage capacity of the battery when it leaves the factory, obtain the battery-related data of the target vehicle, the battery data includes the use environment of the vehicle, the initial storage capacity of the battery, the mileage data of the vehicle, the charge and discharge cycle of the battery and the storage capacity of the battery after each charge and discharge, calculate the first data of the battery after each charge and discharge, use the battery-related data and the corresponding first data as learning data, train the learning data to generate a battery evaluation model, use the battery evaluation model to estimate the battery loss rate after a preset time, obtain the relevant data of the key components of the target vehicle, obtain the loss rate of each key component based on the relevant data of the key components, and calculate the loss rate of the target vehicle using the first formula based on the loss rate of the battery and the loss rate of each key component. The first formula is: Where B is the battery loss rate, W b is the weight of the battery, N is the total number of key components, C i is the loss rate of the i-th key component, W i is the weight of the i-th key component.

[0050] Specifically, in order to obtain the vehicle loss rate of the target vehicle so that a more reasonable premium can be set when setting the premium, since the vehicle is a new energy vehicle, the battery loss is the key. First, the battery data of the target vehicle is obtained, and the first data of the battery is calculated based on the battery data. The first data refers to the ratio of the current storage capacity of the battery when it is fully charged to the storage capacity of the battery when it leaves the factory. In order to obtain a more accurate battery loss rate, the relevant data of the battery is obtained, and the first data of the battery after each charge and discharge is calculated based on the relevant data. The relevant data and the first data are used as learning data to train the model to generate a battery evaluation model. Vehicle insurance premiums are generally paid in years, so the battery evaluation model is used to estimate the battery loss rate after one year. In order to obtain the vehicle loss rate more accurately, the relevant data of other key components of the vehicle, such as the engine and brakes, are also considered. Different weight values ​​are set for each key component based on the degree of influence of each key component on the vehicle. Since it is a new energy vehicle, assuming that the key components are the engine and brakes, the weight of the battery can be set to 0.6, the weight of the engine can be set to 0.3, and the weight of the brake can be set to 0.1. The vehicle loss rate is calculated by the above first formula. The above method can more accurately calculate the vehicle loss rate.

[0051] Furthermore, the target vehicle's driving route history data is used to obtain the target vehicle's total score, including the following steps:

[0052] Obtain the target vehicle's driving route history data, perform statistical analysis on the driving route history data, obtain accident-prone areas on the corresponding route, pre-store an accident information storage table that stores the accident location, accident type, and accident level, compare the driving route history data with the accident location in the accident information storage table, and determine whether the target vehicle has passed the accident location in the accident information storage table. If so, set a vehicle score for the target vehicle based on the accident level. After comparing all accident locations in the accident information storage table, obtain the target vehicle's total vehicle score.

[0053] Specifically, in order to obtain the total score of the target vehicle, the target vehicle's driving route history data is considered, and an accident information storage table is pre-set to store accident-related information of high-incidence accidents such as accident locations and accident levels. The target vehicle's driving route history data is compared with the accident location in the accident information storage table to determine whether the target vehicle has passed the accident location in the accident information storage table. If it has passed but no accident has occurred, the score of the target vehicle is increased. If it has passed but an accident has occurred, the score of the target vehicle is reduced. After all the accident locations in the accident information storage table are determined through the above steps, the target vehicle's total vehicle score is obtained. Finally, the vehicle total score can be used to determine whether the target vehicle often has accidents, and the vehicle total score can also be used to reflect the target vehicle's driving level. Subsequently, a more reasonable premium can be set based on the vehicle total score. If the vehicle total score is high, it means that the target vehicle may have a low frequency of accidents and a corresponding high driving level, and the premium can be appropriately reduced. If the vehicle total score is low, it means that the target vehicle may have a high frequency of accidents, and the premium can be appropriately increased. This calculation method can also encourage and urge drivers to drive safely.

[0054] Further, setting a vehicle score for the target vehicle based on the accident level includes the following steps:

[0055] Preset multiple levels of accident levels, and set corresponding scores for each accident level. Initialize the total vehicle score of the target vehicle to zero. Based on the historical driving route data, determine whether the target vehicle has had an accident at the accident site. If not, obtain the weather conditions and time period when the target vehicle passes the accident site. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, increase the score corresponding to the accident level for the target vehicle by half, otherwise increase the score corresponding to the accident level for the target vehicle. If yes, obtain the weather conditions and time period when the target vehicle passes the accident. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, subtract the score corresponding to the accident level from the target vehicle, otherwise subtract half of the score corresponding to the accident level from the target vehicle.

[0056] Specifically, in order to set a reasonable vehicle score for the target vehicle, multiple levels of accident grades are set in advance for the accidents in the accident storage table, and corresponding scores are set for each accident grade. The total vehicle score of the target vehicle is initialized to zero. For each accident location in the accident storage table, it is determined whether the target vehicle has had an accident at the accident location based on the driving route history data. If not, the weather conditions and time period when the target vehicle passes the accident location are obtained. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, for example, the weather conditions are clear, there is no wind and visibility is high, and the time period is during the daytime of the season, such as 6 am to 5 pm, it means that the weather conditions and time period are conditions where accidents are not likely to occur, and the probability of an accident is low. Therefore, half of the corresponding accident grade score is added to the target vehicle. Otherwise, it means that the current weather conditions are not good, visibility may be low, or it may not be within the preset time. In this period, affected by the external environment, the probability of an accident will also be higher. Under such conditions, the target vehicle still has no accident, indicating that the target vehicle has good driving skills and driving habits, and the corresponding probability of an accident is low. Therefore, a higher score is set for the target vehicle, that is, the score of the target vehicle is set to the score corresponding to the accident level. If an accident has occurred, the weather conditions and time periods are compared. If the weather conditions are good and the time period is daytime when accidents are not likely to occur, but the target vehicle still has an accident, it means that the driving skills and driving habits of the target vehicle are relatively unstable, so a higher score is subtracted from the target vehicle. If the weather conditions are bad or the time period is night when accidents are prone to occur, the target vehicle has an accident due to the influence of external environmental factors, so a lower score is subtracted from the target vehicle. Setting a score for the target vehicle by the above method can more reasonably estimate the score of the target vehicle, which provides a basis for setting a reasonable premium in the future.

[0057] According to another aspect of the embodiment of the present invention, Figure 2 As shown, an insurance assessment system for new energy vehicles is also provided, including a data collection module, a data verification module, a model training module and a premium setting module, which are used to implement the insurance assessment method for new energy vehicles described above. The specific functions of each module are as follows:

[0058] A data collection module is used to collect the driver's driving record data and vehicle data through the vehicle terminal of the new energy vehicle, and also collect the driver's characteristic data. The driving record data at least includes mileage data, speed data, risk record, driving time data, acceleration and deceleration data, regional data and weather data. The vehicle data at least includes vehicle model, engine data, vehicle fault data and battery data;

[0059] A data verification module is used to pre-process the driving record data, remove invalid and erroneous data, convert the driving record data, vehicle data and driver characteristic data into data in a standardized format, obtain key data from other information providing terminals, obtain relevant data corresponding to the key data from the driving record data, compare the key data and the relevant data, and determine whether they are the same. If they are the same, the relevant data is retained; if they are different, the relevant data in the driving record data is deleted;

[0060] A model training module is used to perform a first analysis on the driving record data, vehicle data and driver characteristic data to obtain risk characteristic data items related to risk assessment, use a second analysis method to analyze each risk characteristic data item to obtain an analysis result, and input the analysis result as input data into a machine learning model to perform model training and generate a risk assessment model;

[0061] The premium setting module is used to use the model training module to perform a first analysis and a second analysis on the target driving record data, the target vehicle data and the target driver characteristic data to obtain the corresponding target risk characteristic data, input the target risk characteristic data into the risk assessment model, obtain the risk assessment value, obtain the vehicle loss rate of the target vehicle based on the target vehicle data, obtain the target vehicle total score based on the target vehicle's driving route history data, and set the insurance premium based on the risk assessment value, the vehicle loss rate and the target vehicle total score.

[0062] According to another aspect of an embodiment of the present invention, a storage medium is also provided, wherein the storage medium stores program instructions, wherein when the program instructions are executed, the device where the storage medium is located is controlled to execute any one of the above-mentioned insurance assessment methods for new energy vehicles.

[0063] In summary, the insurance assessment method, system and storage medium of the new energy vehicle of the present invention include collecting the driver's driving record data and vehicle data through the on-board terminal of the new energy vehicle, collecting the driver's characteristic data, verifying the data through a data verification module, deleting erroneous data, performing a first analysis and a second analysis on the driving record data, vehicle data and driver characteristic data to obtain analysis results, using a machine learning algorithm to perform model training on the analysis results as input data, generating a risk assessment model, obtaining a risk assessment value based on the target driving record data, target vehicle data, target driver characteristic data and the risk assessment model, and obtaining a vehicle loss rate and a target vehicle total score. The insurance premium is set based on the risk assessment value, the vehicle loss rate and the target vehicle total score. The present invention can set a more reasonable premium.

[0064] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0065] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The above-mentioned program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0066] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0067] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0068] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An insurance assessment method for new energy vehicles, characterized in that: The steps include: Collecting the driver's driving record data and vehicle data through the vehicle terminal of the new energy vehicle, and also collecting the driver's characteristic data, the driving record data at least includes mileage data, speed data, risk record, driving time data, acceleration and deceleration data, regional data and weather data, and the vehicle data at least includes vehicle model, engine data, vehicle fault data and battery data; Preprocessing the driving record data to remove invalid and erroneous data, converting the driving record data, the vehicle data and the driver characteristic data into data in a standardized format, acquiring key data from other information providing terminals, acquiring relevant data corresponding to the key data from the driving record data, comparing the key data and the relevant data to determine whether they are the same, if they are the same, retaining the relevant data, if they are different, deleting the relevant data from the driving record data; Performing a first analysis on the driving record data, the vehicle data, and the driver characteristic data to obtain risk characteristic data items related to risk assessment, using a second analysis method to analyze each of the risk characteristic data items to obtain an analysis result, and inputting the analysis result as input data into a machine learning model to perform model training and generate a risk assessment model; Use the model training module to perform a first analysis and a second analysis on the target driving record data, the target vehicle data and the target driver characteristic data to obtain corresponding target risk characteristic data, input the target risk characteristic data into the risk assessment model to obtain a risk assessment value, obtain the vehicle loss rate of the target vehicle based on the target vehicle data, obtain the target vehicle total score based on the target vehicle's driving route history data, and set insurance premiums based on the risk assessment value, the vehicle loss rate and the target vehicle total score.

2. The method according to claim 1, characterized in that Performing a first analysis on the driving record data, the vehicle data and the driver characteristic data comprises the following steps: Perform statistical analysis on each trip, calculate the statistical data amount of each risk characteristic data item, obtain the risk record therein based on the driving record data, obtain all first data items corresponding to the time point when the risk record occurs, obtain a time series distribution graph of the first data item, obtain the time point of the first data item in the time series distribution graph, obtain multiple second data items within a preset time period before the time point, obtain multiple third data items within a preset time period after the time point, calculate the average value of the first data item, multiple second data items and multiple third data items as the average value, calculate the standard deviation based on the average value, obtain the trip standard deviation corresponding to the risk record based on the statistical data amount, obtain the difference between the standard deviation and the trip standard deviation, and if the difference is greater than or equal to a preset first threshold, use the first data item as the risk characteristic data item corresponding to the risk record, obtain all the risk characteristic data items corresponding to the risk record, and associate the risk record with the risk characteristic data item and save them.

3. The method according to claim 2, characterized in that Analyzing each of the risk characteristic data items using the second analysis method comprises the following steps: A rectangular coordinate system is established, with the time point as the horizontal axis and each of the risk characteristic data items as the vertical axis, a scatter plot of each of the risk characteristic data items is obtained, two adjacent points are connected to each other based on the scatter plot to obtain a line graph, the line graph is fitted to obtain a corresponding curve graph, and based on the curve graph, it is determined whether the risk characteristic data items have the same change trend. If so, the correlation coefficient between each of the risk characteristic data items is calculated using a correlation coefficient calculation method, and the absolute values ​​of all the correlation coefficients are obtained. The absolute values ​​of all the correlation coefficients are divided into three levels according to a preset second threshold and a third threshold, and those greater than or equal to the second threshold are divided into a first level, those less than the second threshold and greater than or equal to the third threshold are divided into a second level, and those less than the third threshold are divided into a third level. The risk characteristic data items corresponding to the correlation coefficients of the first level are collectively referred to as first risk characteristic data items, the risk characteristic data items corresponding to the correlation coefficients of the second level are collectively referred to as second risk characteristic data items, and the risk characteristic data items corresponding to the correlation coefficients of the third level are collectively referred to as third risk characteristic data items.

4. The method according to claim 1, characterized in that: Acquiring the vehicle loss rate of the target vehicle based on the target vehicle data includes the following steps: Obtain battery data of a target vehicle, calculate first data of the battery based on the battery data, the first data refers to the ratio of the current storage capacity of the battery when it is fully charged to the storage capacity of the battery when it leaves the factory, obtain battery-related data of the target vehicle, the battery data includes the use environment of the vehicle, the initial storage capacity of the battery, the mileage data of the vehicle, the charge and discharge cycle of the battery, and the storage capacity of the battery after each charge and discharge, calculate the first data of the battery after each charge and discharge, use the battery-related data and the corresponding first data as learning data, train the learning data to generate a battery evaluation model, use the battery evaluation model to estimate the loss rate of the battery after a preset time, obtain relevant data of key components of the target vehicle, obtain the loss rate of each key component based on the relevant data of the key components, and calculate the loss rate of the target vehicle using a first formula based on the loss rate of the battery and the loss rate of each key component, the first formula is: Where B is the battery loss rate, W b is the weight of the battery, N is the total number of key components, C i is the loss rate of the i-th key component, W i is the weight of the ith key component.

5. The method according to claim 1, characterized in that The method of obtaining the total score of the target vehicle based on the driving route history data of the target vehicle comprises the following steps: Obtain the driving route history data of the target vehicle, perform statistical analysis on the driving route history data, obtain the accident-prone areas of the corresponding route, pre-store an accident information storage table, the accident information storage table stores the accident location, accident type and accident level, compare the driving route history data with the accident location in the accident information storage table, determine whether the target vehicle has passed the accident location in the accident information storage table, and if so, set a vehicle score for the target vehicle based on the accident level, and after comparing all the accident locations in the accident information storage table, obtain the total vehicle score of the target vehicle.

6. The method according to claim 1, characterized in that Setting a vehicle score for the target vehicle based on the accident level includes the following steps: Preset multiple levels of accident levels, and set corresponding scores for each of the accident levels. Initialize the total vehicle score of the target vehicle to zero. Based on the driving route history data, determine whether the target vehicle has had an accident at the accident site. If not, obtain the weather conditions and time period when the target vehicle passes the accident site. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, increase the score corresponding to the accident level for the target vehicle by half, otherwise increase the score corresponding to the accident level for the target vehicle. If yes, obtain the weather conditions and time period when the target vehicle passes the accident. If the weather conditions meet the preset weather conditions and the time period is within the preset time period, subtract the score corresponding to the accident level from the target vehicle, otherwise subtract half of the score corresponding to the accident level from the target vehicle.

7. An insurance assessment system for new energy vehicles, used to implement the insurance assessment method for new energy vehicles as described in any one of claims 1 to 6, characterized in that: Includes the following modules: A data collection module, used to collect the driver's driving record data and vehicle data through the vehicle terminal of the new energy vehicle, and also collect the driver's characteristic data, the driving record data at least includes mileage data, speed data, risk record, driving time data, acceleration and deceleration data, regional data and weather data, and the vehicle data at least includes vehicle model, engine data, vehicle fault data and battery data; a data verification module, used to pre-process the driving record data, remove invalid and erroneous data, convert the driving record data, the vehicle data and the driver characteristic data into data in a standardized format, obtain key data from other information providing terminals, obtain relevant data corresponding to the key data from the driving record data, compare the key data and the relevant data, determine whether they are the same, if they are the same, retain the relevant data, if they are different, delete the relevant data in the driving record data; A model training module, configured to perform a first analysis on the driving record data, the vehicle data, and the driver characteristic data to obtain risk characteristic data items related to risk assessment, analyze each of the risk characteristic data items using a second analysis method to obtain an analysis result, and input the analysis result as input data into a machine learning model to perform model training and generate a risk assessment model; A premium setting module is used to use the model training module to perform a first analysis and a second analysis on the target driving record data, the target vehicle data and the target driver characteristic data to obtain corresponding target risk characteristic data, input the target risk characteristic data into the risk assessment model, obtain a risk assessment value, obtain the vehicle loss rate of the target vehicle based on the target vehicle data, obtain the total score of the target vehicle based on the driving route history data of the target vehicle, and set the insurance premium based on the risk assessment value, the vehicle loss rate and the total score of the target vehicle.

8. A storage medium, characterized in that: The storage medium stores program instructions, wherein when the program instructions are executed, the device where the storage medium is located is controlled to execute the insurance assessment method for new energy vehicles described in any one of claims 1-6.

Images