CN112232849A

CN112232849A - Accident vehicle evaluation method and system based on intelligent algorithm

Info

Publication number: CN112232849A
Application number: CN202010932013.8A
Authority: CN
Inventors: 孙伟
Original assignee: Fuzhou Benmo Zhiyuan Technology Co ltd
Current assignee: Fuzhou Benmo Zhiyuan Technology Co ltd
Priority date: 2020-09-08
Filing date: 2020-09-08
Publication date: 2021-01-15
Also published as: CN113674015A

Abstract

The invention discloses an accident vehicle evaluation method based on an intelligent algorithm, which comprises the steps of obtaining a driving license picture, a license date and collision position information of a target vehicle, which are input by a user; identifying and obtaining the frame number information according to the driving license photo of the user to obtain the province and the city; obtaining a second-hand car price model according to the car frame number information and the positioning province, city and sale dates; acquiring an accident vehicle collision model according to collision information input by a user; obtaining the price of the second-hand vehicle according to the second-hand vehicle price model and obtaining the vehicle maintenance price according to the accident vehicle collision model; according to the second-hand car price model and the accident car collision model, outputting a corresponding algorithm; and calculating the residual value price of the current vehicle. The invention realizes real-time update by combining historical transaction cases, meets the changing market environment, and realizes scientific prediction and accurate evaluation.

Description

Accident vehicle evaluation method and system based on intelligent algorithm

Technical Field

The invention relates to the field of automobile evaluation, in particular to an accident vehicle evaluation method and system based on an intelligent algorithm.

Background

In recent years, insurance loss material accident cars are more and more going to an auction mode, the sales volume of which is increasing, and massive transaction data are accumulated in the accident car market, but after the vehicles are out of insurance, an insurance company has no support of the data in the process of determining a scheme for deciding whether to perform maintenance or presume total loss auction. In the current mode, the cooperative auction companies are allowed to evaluate the vehicle photos and the vehicle conditions through uploading the photos. The auction company evaluates the vehicle completely by manpower, and judges whether the price is high or low by using the maintenance experience. And the evaluation price can be given only in 24 hours. Insurance companies have a certain influence on the treatment time of vehicles. Because a processing scheme cannot be given in time, the vehicle can also generate a plurality of greasy cats in the parking process. Therefore, the traditional manual evaluation mode lacks objectivity, rationality and scientificity.

Secondly, the existing manual price checking and evaluating modes all need special professionals, all need to know vehicle types and accessory prices, and the possible hidden damage condition can be analyzed and predicted. Therefore, only professionals can complete the work, and the current market demand cannot be met.

Disclosure of Invention

In order to solve the defects of the prior art and meet the current market demand, the invention provides an accident vehicle evaluation method and system based on an intelligent algorithm.

The invention adopts the following technical scheme:

an accident vehicle evaluation method based on an intelligent algorithm comprises the following steps:

s1, acquiring a driving license picture, a boarding date and collision part information of the target vehicle input by a user;

s2, identifying and acquiring the frame number information according to the driving license photo of the user, positioning according to the mobile phone, transmitting the positioning information to the Baidu map AIP, and acquiring the province and the city;

s3, obtaining a second-hand car price model according to the car frame number information and the positioning province, city and sale dates;

s4, acquiring an accident vehicle collision model according to the collision information input by the user;

s5, acquiring the price of the second-hand car according to the price model of the second-hand car and acquiring the maintenance price of the vehicle according to the collision model of the accident vehicle;

s6, outputting a corresponding algorithm according to the second-hand car price model and the accident car collision model;

and S7, calculating the residual value price of the current vehicle.

Further, the used-vehicle price model in step S3 is packaged based on parameters required by the vehicle 300 used-vehicle value evaluation platform, and the used-vehicle price is obtained through the following steps:

s3.1, obtaining unique automobile brand, automobile series, automobile type and annual money according to the identified frame number;

s3.2, acquiring provinces and cities of the vehicles according to the positioning of the mobile phone;

s3.3, estimating the number of kilometers driven according to the license date of the vehicle;

s3.4, establishing a second-hand car price model, transmitting the model to a vehicle 300 second-hand car value evaluation platform, and calling an algorithm of the vehicle 300 second-hand car value evaluation platform to obtain the current vehicle second-hand car price, and transmitting the current vehicle second-hand car price through an interface;

and S3.5, carrying out persistence on the second-hand car price obtained by the model.

Further, the estimation interface of the vehicle 300 acquires, and needs the parameters:

1. the vehicle type is obtained through VIN, and if a plurality of vehicle types exist, one vehicle type is manually confirmed;

2. the year and month of the year of the sale are obtained by inputting the initial registration date;

3. the area is according to the positioning area;

4. the kilometers are calculated as the current date minus the registration date by 2 kilometers per year.

Further, the accident vehicle collision model in step S4 is obtained based on years of experience, the damage level is set according to the vehicle collision type, and the maintenance price of each collision model is obtained through analysis of massive transaction data, and the accident vehicle collision model establishing step is as follows:

s4.1, classifying and classifying all possible collision modes of the vehicle;

s4.2, classifying the characteristics of each collision mode, and dividing the characteristics into four grades of micro, light, medium and heavy;

s4.3, combining each collision part with various collision grades, wherein one collision part plus one collision grade is a collision model;

and S4.4, analyzing mass transaction data to obtain the maintenance price of each collision model.

The accident collision model is established by combining and defining the actual collision position of the vehicle and the collision degree, and a plurality of collision models exist simultaneously.

Further, the estimated residual value price in step S5 is the used car price, the maintenance price queried by the system, the logistics cost, the customer passing cost, the profit of the repair shop, the reserved profit of the auction company, and the loss rate.

The accident vehicle evaluation method based on the intelligent algorithm is implemented by relying on an accident vehicle evaluation system, and the accident vehicle evaluation system comprises:

an image processing module: the system is used for carrying out character recognition on the imported driving license photo and reading out the number of the frame and the number plate;

an information receiving module: the mobile phone positioning system is used for receiving vehicle information such as a frame number, a license plate and the like, mobile phone positioning information and vehicle collision information;

a data acquisition module: obtaining vehicle type information and transaction case data by using the obtained vehicle frame number information and positioning information;

a data processing module: processing the vehicle type information, the used vehicle transaction data and the collision information to obtain case data of the used vehicle price;

a model training module: the vehicle model information and the collision information which are processed are used as training samples and input into a machine learning algorithm for learning and training to obtain parameter information of an estimation model;

a data persistence module: the obtained model parameter information is stored, so that the accident vehicle price evaluation model can be conveniently called;

an estimation output module: the system is used for calling a second-hand car price model and an accident car valuation model according to the information of the target vehicle and calculating the residual valuation price of the accident car of the target vehicle;

an external display module: and displaying the accident vehicle residual value price and the analysis information of the target vehicle.

Preferably, the data acquisition module is further configured to:

obtaining vehicle type information according to the identified vehicle frame number information;

obtaining the information of province and city of the vehicle according to the mobile phone positioning information;

obtaining the driving mileage of the vehicle according to the processing result of the data processing module;

acquiring price data of a second-hand car price interval according to the information of the car type, the information of the province and the city and the information of the driving mileage;

and obtaining the maintenance price data of the vehicle according to the collision grade information and the vehicle type information.

Preferably, the data processing module is further configured to:

processing image information according to the driving license picture, and identifying a frame number and a license plate;

processing the vehicle use duration data according to the initial date;

processing collision grade data, vehicle breakage data and second-hand vehicle price data according to the vehicle type information and the collision information;

historical repair price data is processed based on the collision information.

Preferably, the estimation model parameters include a discount proportion, a reserved profit, a passing household cost and a logistics cost;

the transaction case data comprises second-hand car case data and accident car maintenance case data;

the target vehicle information comprises vehicle type identification information and state information;

the vehicle type identification information of the vehicle comprises a brand, a vehicle series and a vehicle type;

the status information of the vehicle includes a boarding time (initial date).

The beneficial technical effects obtained by adopting the technical scheme are as follows:

the accident vehicle evaluation method based on the intelligent algorithm is based on the machine learning algorithm, is combined with historical transaction cases, realizes real-time updating, meets the changing market environment, and achieves scientific prediction and accurate evaluation. The method is based on massive transaction case data, utilizes a big data technology to summarize the potential characteristics of the accident vehicle, establishes models for the characteristics, and comprehensively evaluates the vehicle conditions by combining market change conditions and corresponding to a plurality of pieces of historical transaction data, so that an evaluation system is more scientific, reasonable and intelligent.

Drawings

Fig. 1 is a flowchart of an accident vehicle evaluation method based on an intelligent algorithm.

Fig. 2 is a flowchart of a used vehicle price obtaining method of the present invention.

FIG. 3 is a flow chart of a process for building an intelligent algorithm based accident vehicle valuation model.

Fig. 4 is a structural framework diagram of the intelligent algorithm-based accident vehicle evaluation system of the invention.

Fig. 5 is a screenshot of an accident vehicle assessment system, according to an embodiment.

Detailed Description

The technical solutions in the embodiments of the present invention are clearly and completely described with reference to fig. 1 to 5 and tables 1 to 9, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The data acquisition module is further configured to: obtaining vehicle type information according to the identified vehicle frame number information; obtaining the information of province and city of the vehicle according to the mobile phone positioning information; obtaining the driving mileage of the vehicle according to the processing result of the data processing module; acquiring price data of a second-hand car price interval according to the information of the car type, the information of the province and the city and the information of the driving mileage; and obtaining the maintenance price data of the vehicle according to the collision grade information and the vehicle type information.

The data processing module is further configured to: processing image information according to the driving license picture, and identifying a frame number and a license plate; processing the vehicle use duration data according to the initial date; processing collision grade data, vehicle breakage data and second-hand vehicle price data according to the vehicle type information and the collision information; historical repair price data is processed based on the collision information.

The parameters of the valuation model comprise a depreciation proportion, a reserved profit, a passing household cost and a logistics cost.

The transaction case data comprises second-hand car case data and accident car maintenance case data.

The target vehicle information includes vehicle type identification information and state information.

The model identification information of the vehicle includes a brand, a vehicle series, and a model.

The state information of the vehicle includes a boarding time (initial date).

The accident vehicle evaluation method based on the intelligent algorithm comprises the following steps:

and S7, calculating the residual value price of the current vehicle.

In the step S3, the used vehicle price model is packaged based on parameters required by the vehicle 300 used vehicle value evaluation platform, and the used vehicle price is obtained through the following steps:

The accident vehicle collision model in the step S4 is obtained based on years of experience, the damage level is set according to the vehicle collision type, and the maintenance price of each collision model is obtained through analysis of massive transaction data, and the accident vehicle collision model establishment steps are as follows:

s4.1, classifying and classifying all possible collision modes of the vehicle;

The accident collision model is established by combining and defining the actual collision position of the vehicle and the collision degree, for example, the collision form is a frontal collision, and the collision degree is micro, so that the combination is a collision model. During a vehicle collision, multiple collision models may exist simultaneously. As in table 1, a partial collision model is presented.

TABLE 1 automotive crash model graphic Table

The accident vehicle estimation model is established by combining the experience of an industry top-level price checker for more than ten years, summarizing and classifying the characteristics of the accident vehicle, and utilizing mass transaction data through data transformation and discretization processing.

The accident vehicle estimation model establishment method comprises the following steps:

(1) deducing, analyzing and summarizing collision modes by using tens of thousands of collision cases;

(2) analyzing the same type of collision condition, analyzing the loss degree and setting the collision grade;

(3) listing the accessory loss conditions of each collision grade, and analyzing the loss characteristics of the same grade;

(4) and combining the collision mode, the collision grade and the loss characteristics, wherein one combination is just one model.

In step S5, the estimated residual value price is the second-hand car price, the maintenance price queried by the system, the logistics cost, the customer-passing cost, the profit of the repair shop, the reserved profit of the auction company, and the discount rate.

The second-hand car price in the formula is obtained through the valuation interface of the car 300, each car series indicates whether the car is sold successfully or is ordinary or cold, and the corresponding price is obtained according to the hot selling degree of the car series, as shown in the following table 2.

TABLE 2 price corresponding table for vehicle type of second-hand vehicle

Freely-sold vehicle type	Get 300 outstanding car commercial purchases price
		Common vehicle type	Good vehicle commercial purchase price for taking 300 vehicles
Cold-door vehicle type	General vehicle commercial purchase price for taking 300 vehicles
		Super-cooled vehicle model	General vehicle commercial purchase price for taking 300 vehicles

The vehicle 300 valuation interface obtains, requires parameters:

3. the area is according to the positioning area;

The maintenance price inquired by the system is analyzed through mass historical transaction data, each collision model of each vehicle type has a corresponding maintenance price, and the program can directly obtain the maintenance price according to the corresponding collision model.

The second-hand vehicle price is 500 yuan, 2000 yuan and 5000 yuan, wherein the second-hand vehicle price is less than 10 ten thousand, the second-hand vehicle price is 10 to 20 ten thousand, and the second-hand vehicle price is more than 20 ten thousand.

Regarding the user fee, the user fee is 800 yuan less than 10 ten thousand for the second-hand vehicle, 1500 yuan less than 10 to 20 ten thousand for the user fee, and 3000 yuan less than 20 ten thousand for the user fee.

The profit of the repair shop is the lowest profit of the reserved repair shop of 2000 percent, and the price of the second-hand car is 10 percent.

Regarding auction company profit, the auction company profit is charged in sections for used car prices, and the profit criteria are made as per table 3.

TABLE 3 auction company auction price sectional toll collection table

The depreciation rate relates to depreciation range degree grading, depreciation of special conditions, depreciation of major accidents, depreciation of secondary accidents, depreciation of additional flooding accidents and the like.

The lossy extent rankings are shown in table 4.

TABLE 4 damaged Return Range grade Scale

Heavy load	Severe collision
		In	Moderate collision
Light and lightweight	Light collision
		Micro-meter	Micro collision

And traversing all the breaking conditions by using a backtracking algorithm in the breaking calculation mode, obtaining corresponding breaking proportions, and taking the maximum value, wherein if the maximum breaking proportion plus the water-flooded breaking proportion is met, the maximum breaking proportion is taken.

First, the special case compromises are shown in table 5.

TABLE 5 Special cases discount watch

Damaged frame number	Alternative to engine number change	20% (second-hand vehicle price)
			Engine number change	And the damaged frame number is selected from	20％

Regarding major accident damage, micro collision is considered to have no damage.

One type of major accident breakage ratio is shown in table 6.

TABLE 6 proportion table for breaking-down of major accident

Tip over (roof all touching ground)	Heavy (hurt A \ B \ C column)	20％
			High-altitude pendant	Heavy (hurt A \ B \ C column)	20％
Front collision with truck	Heavy (hurt A \ B \ C column)	20％
			Side collision with truck	Heavy (hurt A \ B \ C column)	20％

The ratio of the second type of accident damage is shown in Table 7.

TABLE 7 proportion table for breakdown of second-class accident

The damage of the secondary accident is shown in Table 8. And if the secondary accident is damaged, the special condition is damaged, the major accident is damaged, and if the secondary accident is overlapped, the highest damage is selected.

TABLE 8 breakdown watch on secondary accident

Freely-sold vehicle type	10％
		Common vehicle type	15％
Cold-door vehicle type	20％
		Super-cooled vehicle model	20％

Additional flooding incident penalties are shown in table 9. If the collision vehicle has a water flooding accident, the proportion of the damage is added on the basis of the original damage.

Table 9 is an additional flooding accident damage table.

Freely-sold vehicle type	15％
		Common vehicle type	20％
Cold-door vehicle type	25％
		Super-cooled vehicle model	25％

Taking a 2018-model 1.5L CVT comfortable skylight-version accident vehicle as an example, the intelligent algorithm-based accident vehicle evaluation method estimates the residual value, as shown in FIG. 5, where: kangdong Qingdao, VIN code of car: LHGGK5869K8092072, model: a 2018 version of a comfortable skylight for a femoris 1.5L CVT, with an initial date: 2019-09, special case: no, collision mode: the two vehicles collide with each other at the front-in-the-middle.

Residual value evaluated price: 4.29 ten thousand.

And 7, a freely sold vehicle type. And 7, a freely sold vehicle type. There are special situations: none. Collision mode: the two vehicles collide with each other at the front-in-the-middle.

It should be understood, however, that the description herein of specific embodiments is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.

Claims

1. An accident vehicle evaluation method based on an intelligent algorithm is characterized by comprising the following steps:

s2, identifying and acquiring the frame number information according to the driving license photo of the user, positioning according to the mobile phone, and transmitting the positioning information to a Baidu map API to acquire province and city;

and S7, calculating the residual value price of the current vehicle.

2. The accident vehicle evaluation method based on intelligent algorithm according to claim 1, wherein the used vehicle price model in step S3 is packaged based on parameters required by a vehicle 300 used vehicle value evaluation platform, and the used vehicle price is obtained by the following steps:

3. The intelligent algorithm-based accident vehicle assessment method according to claim 1, wherein the accident vehicle collision model in step S4 is obtained based on years of experience, is set according to vehicle collision type and damage level, and obtains the maintenance price of each collision model through mass transaction data analysis, and the accident vehicle collision model establishing step is as follows:

s4.1, classifying and classifying all possible collision modes of the vehicle;

4. The intelligent algorithm-based accident vehicle assessment method according to claim 1, wherein the residual estimated price is the price of used vehicle-the maintenance price queried by the system-the logistics cost-the expense of passing the house-the profit of the repair shop-the reserved profit of the auction company-the discount rate in step S5.

5. The accident vehicle assessment method based on intelligent algorithm according to claim 1, wherein the accident vehicle assessment method is implemented by means of an accident vehicle assessment system, and the accident vehicle assessment system comprises:

6. An intelligent algorithm-based accident vehicle assessment method according to claim 5, wherein said data acquisition module is further configured to:

7. An intelligent algorithm-based accident vehicle assessment method according to claim 5, wherein said data processing module is further configured to:

processing the vehicle use duration data according to the initial date;

historical repair price data is processed based on the collision information.

8. The intelligent algorithm-based accident vehicle assessment method according to claim 5, wherein the estimation model parameters comprise discount rate, reserved profit, house passing cost, and logistics cost;

the status information of the vehicle includes a boarding time (initial date).