CN109658272A - Driving behavior evaluation system and Insurance Pricing system based on driving behavior - Google Patents
Driving behavior evaluation system and Insurance Pricing system based on driving behavior Download PDFInfo
- Publication number
- CN109658272A CN109658272A CN201811600703.2A CN201811600703A CN109658272A CN 109658272 A CN109658272 A CN 109658272A CN 201811600703 A CN201811600703 A CN 201811600703A CN 109658272 A CN109658272 A CN 109658272A
- Authority
- CN
- China
- Prior art keywords
- driving behavior
- driving
- danger
- sample
- data
- 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
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
- G06Q30/0283—Price estimation or determination
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Technology Law (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The Insurance Pricing system based on driving behavior that the invention discloses a kind of, including driving behavior identification module, for identification driving behavior of user；Driving behavior extraction module obtains driving behavior for carrying out feature extraction to the driving behavior；Driving characteristics analysis module does correlation analysis to the driving behavior using the driving behavior coefficient correlation model that is in danger, obtains the coefficient that is in danger of driving behavior；Insurance Pricing module, be in danger coefficient and the driving behavior progress Insurance Pricing in conjunction with other insurance underwriting Pricing Factors to user according to.The present invention is based on the Insurance Pricing systems of driving behavior, driving behavior is identified from the driving behavior data of acquisition, feature extraction is carried out to driving behavior, and to the driving behavior of extraction, then with these features and it is in danger the frequency and the amount of money that is in danger is associated analysis, obtain driving behavior the coefficient that is in danger, root be in danger coefficient carry out Insurance Pricing.
Description
Technical field
The invention belongs to car insurance application fields, and in particular to a kind of driving behavior evaluation system and be based on driving behavior
Insurance Pricing system.
Background technique
Car networking is to be connected internet and loading mobile device on automobile, is realized in the network platform to vapour
The information data of vehicle is acquired, shares and efficiently uses, and carries out real-time management to automobile according to different functional requirements, with
Just the service of diversification is provided for related industry.In insurance field, the most important application of car networking is UBI (Usage Based
Insurance, the insurance based on driving behavior), with the rise of UBI car networking insurance and big data, there is an urgent need to a kind of sections
Method assesses the driving behavior of driver, which can formulate different grades of premium for insurance company
Foundation is provided, the data supporting that test user drives risk is also used as, reminds and supervise car owner to improve driving habits.
Summary of the invention
The Insurance Pricing system based on driving behavior that the technical problem to be solved in the present invention is to provide a kind of, to improve insurance
Price precision reduces Claims Resolution cost, improves driving behavior.
The Insurance Pricing system based on driving behavior that in order to solve the above-mentioned technical problems, the present invention provides a kind of, including,
Driving behavior identification module, driving behavior and traveling for identifying user from vehicle drive and running data are practised
It is used；
Driving behavior extraction module is driven for carrying out feature extraction to the driving behavior and driving habits
Sail behavioural characteristic；
Driving characteristics analysis module does correlation analysis to the driving behavior and the coefficient that is in danger, and obtains and drives
The relevant driving behavior of behavioural characteristic is in danger correlation model, and obtains the coefficient that is in danger；
Insurance Pricing module, according to the driving behavior be in danger correlation model and be in danger coefficient and combine other insurance
Pricing Factor of accepting insurance carries out Insurance Pricing to the driving behavior of user.
It further comprise it further include weight adjustment module in a preferred embodiment of the present invention, the weight adjusts mould
Block is used to adjust the driving behavior and is in danger correlation model and the weight of coefficient of being in danger, and the Insurance Pricing module is weighed according to adjustment
The coefficient that is in danger after weight carries out Insurance Pricing to the driving behavior of user.
It further comprise vehicle control of the driving behavior identification module from user in a preferred embodiment of the present invention
The driving behavior of user is identified in the driving behavior data of bus acquisition；Or the vehicle driving being attached on vehicle from user
Driving behavior and the driving habits of user are identified in the vehicle operation data of data acquisition equipment acquisition.
It further comprise the driving behavior data include but is not limited to vehicle position in a preferred embodiment of the present invention
It sets, speed, acceleration, angular speed, angular acceleration, brake, image/video and obstacle avoidance sensor data；From the driving
The driving behavior identified in behavioral data includes but is not limited to driving range, running region, travel speed, bring to a halt, take a sudden turn,
Anxious acceleration, hypervelocity, fatigue driving, following distance and change lane.
It further comprise it further include driving behavior identification model training module, institute in a preferred embodiment of the present invention
It states driving behavior identification model training module to be trained the driving travel data for being used as sample, obtains driving behavior and identify mould
Type；It includes,
Driving travel data sample processing unit, driving and traveling initial data to acquisition carry out combing cleaning, select
Driving travel data sample obtains sample driving travel data；
Driving behavior identification model training unit, based on cluster and related algorithm and according to the sample driving travel
Data training obtains the driving behavior identification model.
In a preferred embodiment of the present invention, further comprise its further include driving behavior be in danger coefficient correlation model training
Module, the driving behavior coefficient correlation model training module that is in danger drive the sample and running data is trained, obtain
The driving behavior is in danger coefficient correlation model；It includes,
Driving behavior analytical unit carries out sample driving travel data and is in danger number and the amount of money that is in danger is associated with
Property analysis, obtain sample driving behavior, the sample driving behavior include but is not limited to cluster and classification feature；
Correlation model training unit, based on related algorithm and according to the sample driving behavior and the frequency of being in danger
And the degree of association training for the amount of money that is in danger obtains the driving behavior and is in danger coefficient correlation model.
In a preferred embodiment of the present invention, further comprise analyze the sample driving behavior and be in danger the frequency and
Be in danger the amount of money correlation model when consider probability that the sample driving behavior occurs in various driving habits；Wherein,
The driving habit including but not limited to drives period, road type, total driving range and always drives duration.
In order to solve the above-mentioned technical problems, the present invention provides a kind of driving behavior evaluation systems, it is characterised in that: including
Driving behavior identification module, for identifying the driving behavior of user from the vehicle drive and running data of acquisition；
Driving behavior extraction module obtains driving behavior for carrying out feature extraction to the driving behavior；
Driving behavior analysis module, using driving behavior be in danger coefficient correlation model the driving behavior is done it is more
Target correlation analysis, obtains the prediction data of driving behavior, and the prediction data includes but is not limited to predict accident rate, thing
Therefore loss late, vehicle loss rate, oil consumption rate and rate violating the regulations；
Marking evaluation module, carries out marking evaluation to the driving behavior of user according to the prediction data.
It further comprise it further include weight adjustment module in a preferred embodiment of the present invention, the weight adjusts mould
Block is used to adjust the weight of the prediction data, and the marking evaluation module is according to the prediction data after adjustment weight to user's
Driving behavior carries out marking evaluation.
It further comprise vehicle control of the driving behavior identification module from user in a preferred embodiment of the present invention
The driving behavior of user is identified in the driving behavior data of bus acquisition；Or the vehicle being attached on vehicle from the user
The driving behavior of user is identified in the vehicle operation data of traveling acquisition equipment acquisition.
It further comprise the vehicle drive and running data includes but is not limited to vehicle in a preferred embodiment of the present invention
Position, speed, acceleration, angular speed, angular acceleration, brake, image video and obstacle avoidance sensor information；From described
The driving behavior identified in vehicle drive and running data includes but is not limited to driving range, travel speed, brings to a halt, racing
Curved, anxious acceleration, hypervelocity, fatigue driving, following distance and change lane.
It further comprise it further include driving behavior identification model training module, institute in a preferred embodiment of the present invention
It states driving behavior identification model training module to be trained the driving travel data for being used as sample, obtains the driving behavior and know
Other model；It includes,
Driving travel data sample processing unit, driving and traveling initial data to acquisition carry out combing cleaning, select
Driving travel data sample obtains sample driving travel data；
Driving behavior identification model training unit is instructed based on related algorithm and according to the sample driving travel data
Practice and obtains the driving behavior identification model.
In a preferred embodiment of the present invention, further comprise its further include driving behavior be in danger coefficient correlation model training
Module, the driving behavior coefficient correlation model training module that is in danger are trained the sample driving travel data, obtain
The driving behavior is in danger coefficient correlation model；It includes,
Driving behavior analytical unit carries out sample driving travel data and is in danger number and the amount of money that is in danger is associated with
Property analysis, obtain sample driving behavior, the sample driving behavior include but is not limited to cluster and classification feature；
Correlation model training unit, based on related algorithm and according to the sample driving behavior and the frequency of being in danger
And the degree of association training for the amount of money that is in danger obtains the driving behavior and is in danger coefficient correlation model.
In a preferred embodiment of the present invention, further comprise analyze the sample driving behavior and the frequency violating the regulations,
Consider the sample driving behavior in various driving habits when the degree of association of penalty score, accident frequency and accident severity
The probability of middle appearance；Wherein, the driving habit includes but is not limited to drive the period, and road type and is always driven total driving range
Sail duration.
Compared to traditional Insurance Pricing system, the present invention is based on the Insurance Pricing systems of driving behavior with following excellent
Gesture:
1, compared to traditional from factors such as vehicle, vehicle age, mileage number, driver's age, gender, marital status
Insurance Pricing, Insurance Pricing method of the invention is from really to the strongest factor of vehicle insurance forecasting of cost ability --- and " driving behavior "
It sets out, increases very crucial risk information on the basis of traditional vehicle insurance Pricing Factor, so that vehicle insurance price be made to become more
It is accurate to add, and can increase substantially the price precision of vehicle insurance.
2, using driving behavior as the foundation factors of Insurance Pricing, driving behavior is better, and premium is lower.Therefore, compared to
The bad client of driving behavior, it is believed that the client that oneself driving behavior is good, risk is low is more likely to selection based on driving behavior
Insurance products, on short terms, the insurance products based on driving behavior can motivate the self selective of top-tier customer, public for insurance
Department attracts the driver of more and more low-risks.With gradually increasing for top-tier customer, vehicle insurance accident will be greatly reduced, insurance
The Claims Resolution spending of company also will accordingly decline therewith.
3, another aspect by the real time monitoring to client's driving behavior and is from the point of long-term development safe driving
Premium discount is provided, effectively can form a kind of restraining force to customer action, to gradually improve the driving behavior of user, lower
It is in danger frequency, fundamentally reduces Claims Resolution cost.
Detailed description of the invention
Fig. 1 is the structural block diagram of driving behavior evaluation system in the preferred embodiment of the present invention；
Fig. 2 is the structural block diagram of the Insurance Pricing system in the preferred embodiment of the present invention based on driving behavior；
Fig. 3 is that training obtains driving behavior and is in danger the structural block diagram of coefficient correlation model in the preferred embodiment of the present invention.
Specific embodiment
The present invention will be further explained below with reference to the attached drawings and specific examples, so that those skilled in the art can be with
It more fully understands the present invention and can be practiced, but illustrated embodiment is not as a limitation of the invention.
Embodiment one
As shown in Figure 1, present embodiment discloses a kind of driving behavior evaluation system, including the training of driving behavior identification model
Module, driving behavior are in danger coefficient correlation model training module and driving behavior evaluation, driving behavior are in danger coefficient correlation model
Training module is trained sample driving travel data, obtains driving behavior and is in danger coefficient correlation model；Driving behavior evaluation
Marking evaluation is carried out to the practical driving behavior of user using the driving behavior of the acquisition coefficient correlation model that is in danger.
Above-mentioned driving behavior identification model training module is trained the driving travel data for being used as sample, is driven
Activity recognition model.Specifically, the driving behavior data for being used as sample include but is not limited to vehicle location, speed, acceleration, angle
Speed, angular acceleration, brake, image video and obstacle avoidance sensor data；Know from above-mentioned sample driving behavior data
Other driving behavior includes but is not limited to driving range, travel speed, bring to a halt, take a sudden turn, anxious acceleration, hypervelocity, fatigue driving,
Following distance and change lane.
Driving behavior identification model training module includes driving travel data sample processing unit and driving behavior identification mould
Type training unit:
Driving travel data sample processing unit, driving and traveling initial data to acquisition carry out combing cleaning, select
Driving travel data sample obtains sample driving travel data；
Driving behavior identification model training unit drives and goes based on cluster and related algorithm and according to above-mentioned sample
It sails data training and obtains above-mentioned driving behavior identification model.
The above-mentioned driving behavior coefficient correlation model training module that is in danger is trained above-mentioned sample driving travel data, obtains
Driving behavior is in danger coefficient correlation model.
As shown in figure 3, driving behavior is in danger coefficient correlation model training module include driving behavior analytical unit and
Correlation model training unit.
Above-mentioned driving behavior analytical unit carries out and be in danger number and the amount of money that is in danger sample driving travel data
Association analysis, obtains sample driving behavior, and above-mentioned sample driving behavior includes but is not limited to cluster and classify
Feature.Specifically, above-mentioned driving behavior analytical unit drives another set sample using above-mentioned driving behavior identification model
It sails running data and carries out feature extraction, obtain sample driving behavior.Specifically, combining road condition carries out sample driving behavior
The driving behavior of objective reality is rejected in feature extraction, for example, current road segment very congestion, bringing to a halt at this time, it is normal to belong to
The sample driving behavior for belonging to normally travel is rejected, is driven to subjective sample by driving behavior, driving behavior extraction unit
It sails behavioural characteristic to extract, for example, present road is unobstructed, brings to a halt, takes a sudden turn at this time, anxious acceleration, fatigue driving, becoming lane
Equal driving behaviors are to be considered as subjective sample driving behavior.
Above-mentioned correlation model training unit, based on cluster and related algorithm and according to the sample driving behavior
The driving behavior is obtained with the training of the degree of association of the be in danger frequency and the amount of money that is in danger to be in danger coefficient correlation model.Sample is driven and is gone
It is characterized and is in danger the frequency and the amount of money that is in danger does association analysis respectively, obtain sample driving behavior and the frequency of being in danger, be in danger
The degree of association of the amount of money.For example, the degree of association of the sample driving behavior of zig zag and the frequency of being in danger is 0.8 and the gold that is in danger
The degree of association of volume * * * * ten thousand is 0.6.
In addition, being examined when it should be noted that calculating above-mentioned sample driving behavior with the frequency of being in danger, the amount of money degree of association of being in danger
Consider the probability that above-mentioned sample driving behavior occurs in various driving habits；Wherein, above-mentioned driving habit includes but unlimited
In driving period, road type, total driving range drives duration with total.It brings to a halt 10 times for example, user one opens 1000Km, user
Two, which open 100Km, brings to a halt 2 times, after considering the number that driving behavior occurs in total driving range or total driving time, uses
The frequency that family one is brought to a halt is 0.1, and the frequency that user two brings to a halt is 0.2.
Specifically, related algorithm includes but is not limited to covariance, defences matrix, simple regression, multiple regression, convolution mind jointly
Through network algorithm, deep learning, decision tree, classification method based on probability theory etc..
Above-mentioned driving behavior evaluation system includes driving behavior identification module, driving behavior extraction module, drives row
Module and marking evaluation module are adjusted for analysis module, weight.
Above-mentioned driving behavior identification module identifies the driving behavior of user from vehicle drive and running data.Specifically,
The driving behavior of user, or being attached to from user are identified from the driving behavior data that the vehicle control bus of user acquires
The driving behavior of user is identified in the vehicle operation data of vehicle operation data acquisition equipment acquisition on vehicle, above-mentioned vehicle is driven
It sails and running data includes but is not limited to vehicle location, speed, acceleration, angular speed, angular acceleration, brake and anticollision
Sensing data；The driving behavior identified from above-mentioned driving behavior data includes but is not limited to driving range, travel speed, urgency
Brake, zig zag, anxious acceleration, hypervelocity, fatigue driving, following distance and change lane.
Above-mentioned driving behavior analysis module does more mesh to driving behavior using the driving behavior coefficient correlation model that is in danger
Correlation analysis is marked, the prediction data of driving behavior is obtained, above-mentioned prediction data includes but is not limited to predict accident rate, accident
Loss late, vehicle loss rate, oil consumption rate and rate violating the regulations.
Above-mentioned weight adjustment module is used to adjust the weight of above-mentioned prediction data.For example, the risk of zig zag is big,
Weight is increased to it through being in danger on the basis of the frequency or/and the amount of money that is in danger of obtaining, the evaluation score value of zig zag is reduced with this.
Above-mentioned marking evaluation module carries out marking evaluation to the driving behavior of user according to the prediction data after adjustment weight.
Driving behavior evaluation system of the invention can be used for the driving evaluation of the single driving behavior of user, can also
With multiple driving behaviors for overall evaluation user, specifically:
When the single driving behavior of A → evaluation: being in danger coefficient correlation model to single driving row using driving behavior
It is characterized and does association analysis, obtain the prediction data of driving behavior, according to prediction data to single driving behavior
Carry out marking evaluation.For example, the frequency of being in danger of zig zag is 8 times, the amount of money that is in danger is $6000, according to the frequency of being in danger of zig zag
And the amount of money that is in danger carries out marking evaluation to this driving behavior of taking a sudden turn.
In addition, it should be noted that when carrying out marking evaluation to current driving behavior, it is also necessary to consider the frequency of being in danger
Or/and the weight for the amount of money that is in danger.
When the multiple driving behaviors of B → evaluation user: the corresponding one group of prediction data of a driving behavior, to each
The whole driving behavior to user after the corresponding prediction data adjustment weight of a driving behavior is given a mark evaluation.
Embodiment two
As shown in Fig. 2, present embodiment discloses a kind of Insurance Pricing system based on driving behavior, including driving behavior are known
Other model training module, driving behavior be in danger coefficient correlation model training module and based on driving behavior evaluation Insurance Pricing.
The driving behavior coefficient correlation model training module that is in danger is trained sample driving travel data, obtains driving behavior and is in danger and is
Number correlation model；Driving behavior evaluation is in danger practical driving behavior of the coefficient correlation model to user using the driving behavior of acquisition
Marking evaluation is carried out, the Insurance Pricing of driving behavior is based on driving behavior evaluation and carries out Insurance Pricing.
Above-mentioned driving behavior identification model training module is trained the driving travel data for being used as sample, is driven
Activity recognition model.Specifically, the driving behavior data for being used as sample include but is not limited to vehicle location, speed, acceleration, angle
Speed, angular acceleration, brake and obstacle avoidance sensor data；The driving row identified from above-mentioned sample driving behavior data
For include but is not limited to driving range, travel speed, bring to a halt, take a sudden turn, anxious acceleration, hypervelocity, fatigue driving, following distance and
Become lane.
Driving behavior identification model training module includes driving travel data sample processing unit and driving behavior identification mould
Type training unit:
Driving travel data sample processing unit, driving and traveling initial data to acquisition carry out combing cleaning, select
Driving travel data sample obtains sample driving travel data；
Driving behavior identification model training unit is instructed based on related algorithm and according to above-mentioned sample driving travel data
Practice and obtains above-mentioned driving behavior identification model.
The above-mentioned driving behavior coefficient correlation model training module that is in danger is trained above-mentioned sample driving travel data, obtains
Driving behavior is in danger coefficient correlation model.
As shown in figure 3, driving behavior is in danger coefficient correlation model training module include driving behavior analytical unit and
Correlation model training unit.
Above-mentioned driving behavior analytical unit carries out and be in danger number and the amount of money that is in danger sample driving travel data
Association analysis, obtains sample driving behavior, and above-mentioned sample driving behavior includes but is not limited to cluster and classify
Feature.Specifically, above-mentioned driving behavior analytical unit drives another set sample using above-mentioned driving behavior identification model
It sails running data and carries out feature extraction, obtain sample driving behavior.Specifically, combining road condition carries out sample driving behavior
The driving behavior of objective reality is rejected in feature extraction, for example, current road segment very congestion, bringing to a halt at this time, it is normal to belong to
The sample driving behavior for belonging to normally travel is rejected, is driven to subjective sample by driving behavior, driving behavior extraction unit
It sails behavioural characteristic to extract, for example, present road is unobstructed, brings to a halt, takes a sudden turn at this time, anxious acceleration, fatigue driving, becoming lane
Equal driving behaviors are to be considered as subjective sample driving behavior.
Above-mentioned correlation model training unit, based on cluster and related algorithm and according to the sample driving behavior
The driving behavior is obtained with the training of the degree of association of the be in danger frequency and the amount of money that is in danger to be in danger coefficient correlation model.Sample is driven and is gone
It is characterized and is in danger the frequency and the amount of money that is in danger does association analysis respectively, obtain sample driving behavior and the frequency of being in danger, be in danger
The degree of association of the amount of money.For example, the degree of association of the sample driving behavior of zig zag and the frequency of being in danger is 0.8 and the gold that is in danger
The degree of association of volume * * * * ten thousand is 0.6.
In addition, being examined when it should be noted that calculating above-mentioned sample driving behavior with the frequency of being in danger, the amount of money degree of association of being in danger
Consider the probability that above-mentioned sample driving behavior occurs in various driving habits；Wherein, above-mentioned driving habit includes but unlimited
In driving period, road type, total driving range drives duration with total.It brings to a halt 10 times for example, user one opens 1000Km, user
Two, which open 100Km, brings to a halt 2 times, after considering the number that driving behavior occurs in total driving range or total driving time, uses
The frequency that family one is brought to a halt is 0.1, and the frequency that user two brings to a halt is 0.2.
Specifically, related algorithm includes but is not limited to covariance, defences matrix, simple regression, multiple regression, convolution mind jointly
Through network algorithm, deep learning, decision tree, classification method based on probability theory etc..
The above-mentioned Insurance Pricing system based on driving behavior includes driving behavior identification module, driving behavior extraction mould
Block, driving characteristics analysis module, weight adjustment module and Insurance Pricing module.
Above-mentioned driving behavior identification module from driving and running data for identifying the driving behavior of user.Specifically,
The driving behavior of user is identified from the driving behavior data that the vehicle control bus of user acquires；Or being attached to from user
The driving behavior of user is identified in the vehicle operation data of vehicle operation data acquisition equipment acquisition on vehicle.The driving row
It include but is not limited to vehicle location, speed, acceleration, angular speed, angular acceleration, brake and obstacle avoidance sensor for data
Information；The driving behavior identified from the driving behavior data includes but is not limited to driving range, travel speed, bring to a halt,
Zig zag, anxious acceleration, hypervelocity, fatigue driving, following distance and change lane.
Above-mentioned driving behavior extraction module is used to carry out feature extraction to above-mentioned driving behavior, and it is special to obtain driving behavior
Sign.Specifically, combining road condition carries out feature extraction to driving behavior, the driving behavior of objective reality is rejected, for example, currently
Section very congestion, brings to a halt belong to normal driving behavior at this time, and driving behavior extraction unit will belong to normally travel
Driving behavior is rejected, and is extracted to subjective driving behavior, for example, present road is unobstructed, is brought to a halt at this time, racing
Curved, anxious acceleration, becomes the driving behavior that the driving behaviors such as lane are considered as subjectivity at fatigue driving.
Above-mentioned driving characteristics analysis module does correlation to driving behavior using the driving behavior coefficient correlation model that is in danger
Analysis, obtains the coefficient that is in danger of driving behavior.
Above-mentioned weight adjustment module is used to adjust the weight for the coefficient that is in danger, and is suitable for adjustment member or whole driving behaviors
The weight of feature.For example, zig zag risk it is big, obtained be in danger coefficient on the basis of weight is increased to it, with
This Insurance Pricing to improve zig zag.
Above-mentioned Insurance Pricing module according to adjustment weight after be in danger coefficient and combine other insurance underwriting Pricing Factors pair
The driving behavior of user carries out Insurance Pricing.
The present invention is based on the Insurance Pricing method, systems of driving behavior, can be used for the insurance of the single driving behavior of user
Price can be used for carrying out Insurance Pricing to multiple driving behaviors of user, specifically:
A → to single driving behavior carry out Insurance Pricing when: be in danger coefficient correlation model to list using driving behavior
One driving behavior does association analysis, obtains the coefficient that is in danger of driving behavior, single to this according to the coefficient that is in danger
Driving behavior carry out Insurance Pricing.
When B → carry out Insurance Pricing to multiple driving behaviors of user: one corresponding one group of driving behavior goes out
Dangerous coefficient, the whole Insurance Pricing after coefficient adjustment weight of being in danger corresponding to each driving behavior.
Embodiment described above is only to absolutely prove preferred embodiment that is of the invention and being lifted, protection model of the invention
It encloses without being limited thereto.Those skilled in the art's made equivalent substitute or transformation on the basis of the present invention, in the present invention
Protection scope within.Protection scope of the present invention is subject to claims.
Claims (14)
1. a kind of Insurance Pricing system based on driving behavior, it is characterised in that: including,
Driving behavior identification module, for identifying driving behavior and the driving habits of user from vehicle drive and running data；
Driving behavior extraction module obtains for carrying out feature extraction to the driving behavior and driving habits and drives row
It is characterized；
Driving characteristics analysis module does correlation analysis, acquisition and driving behavior to the driving behavior and the coefficient that is in danger
The relevant driving behavior of feature is in danger correlation model, and obtains the coefficient that is in danger；
Insurance Pricing module, according to the driving behavior be in danger correlation model and be in danger coefficient and combine other insurance underwritings
Pricing Factor carries out Insurance Pricing to the driving behavior of user.
2. the Insurance Pricing system based on driving behavior as described in claim 1, it is characterised in that: it further includes weight adjustment
Module, weight adjustment module are used to adjust the driving behavior and are in danger the weight of correlation model and the coefficient that is in danger, the guarantor
Dangerous pricing module carries out Insurance Pricing to the driving behavior of user according to the coefficient that is in danger after adjustment weight.
3. the Insurance Pricing system based on driving behavior as claimed in claim 1 or 2, it is characterised in that: the driving behavior
Identification module identifies the driving behavior of user from the driving behavior data that the vehicle control bus of user acquires；Or from user
Be attached on vehicle vehicle operation data acquisition equipment acquisition vehicle operation data in identify user driving behavior and
Driving habits.
4. the Insurance Pricing system based on driving behavior as claimed in claim 3, it is characterised in that: the driving behavior data
Including but not limited to vehicle location, speed, acceleration, angular speed, angular acceleration, brake, image/video and anticollision pass
Sensor data；The driving behavior identified from the driving behavior data includes but is not limited to driving range, running region, traveling
Speed brings to a halt, takes a sudden turn, anxious accelerations, hypervelocity, fatigue driving, following distance and change lane.
5. the Insurance Pricing system based on driving behavior as described in claim 1, it is characterised in that: it further includes driving behavior
Identification model training module, the driving behavior identification model training module instruct the driving travel data for being used as sample
Practice, obtains driving behavior identification model；It includes,
Driving travel data sample processing unit, driving and traveling initial data to acquisition carry out combing cleaning, and selection drives
Running data sample obtains sample driving travel data；
Driving behavior identification model training unit, based on cluster and related algorithm and according to the sample driving travel data
Training obtains the driving behavior identification model.
6. the Insurance Pricing system based on driving behavior as claimed in claim 5, it is characterised in that: it further includes driving behavior
Be in danger coefficient correlation model training module, the driving behavior be in danger coefficient correlation model training module to the sample drive and
Running data is trained, and is obtained the driving behavior and is in danger coefficient correlation model；It includes,
Driving behavior analytical unit carries out and number and the amount of money relevance of being in danger point of being in danger sample driving travel data
Analysis, obtains sample driving behavior, and the sample driving behavior includes but is not limited to the feature for clustering and classifying；
Correlation model training unit, based on related algorithm and according to the sample driving behavior and be in danger the frequency and go out
The degree of association training of the dangerous amount of money obtains the driving behavior and is in danger coefficient correlation model.
7. the Insurance Pricing system based on driving behavior as claimed in claim 6, it is characterised in that: analyze the sample and drive
Consider that the sample driving behavior is practised in various travelings when the correlation model of behavioural characteristic and the be in danger frequency and the amount of money that is in danger
The probability occurred in used；Wherein, the driving habit including but not limited to drive the period, road type, total driving range and
It is total to drive duration.
8. a kind of driving behavior evaluation system, it is characterised in that: including
Driving behavior identification module, for identifying the driving behavior of user from vehicle drive and running data；
Driving behavior extraction module obtains driving behavior for carrying out feature extraction to the driving behavior；
Driving behavior analysis module does multiple target to the driving behavior using the driving behavior coefficient correlation model that is in danger
Correlation analysis obtains the prediction data of driving behavior, and the prediction data includes but is not limited to predict accident rate, accident damage
Mistake rate, vehicle loss rate, oil consumption rate and rate violating the regulations；
Marking evaluation module, carries out marking evaluation to the driving behavior of user according to the prediction data.
9. driving behavior evaluation system as claimed in claim 8, it is characterised in that: it further includes weight adjustment module, described
Weight adjustment module is used to adjust the weight of the prediction data, and the marking evaluation module is according to the prediction number after adjustment weight
Marking evaluation is carried out according to the driving behavior to user.
10. driving behavior evaluation system as claimed in claim 8 or 9, it is characterised in that: the driving behavior identification module from
The driving behavior of user is identified in the driving behavior data of the vehicle control bus acquisition of user；Or adding from the user
The driving behavior of user is identified in the vehicle operation data of vehicle driving acquisition equipment acquisition on vehicle.
11. driving behavior evaluation system as claimed in claim 10, it is characterised in that: the vehicle drive and running data packet
Include but be not limited to vehicle location, speed, acceleration, angular speed, angular acceleration, brake, image video and anticollision sensing
Device information；The driving behavior identified from the vehicle drive and running data include but is not limited to driving range, travel speed,
It brings to a halt, take a sudden turn, anxious accelerations, hypervelocity, fatigue driving, following distance and change lane.
12. driving behavior evaluation system as claimed in claim 8, it is characterised in that: it further includes driving behavior identification model
Training module, the driving behavior identification model training module are trained the driving travel data for being used as sample, obtain institute
State driving behavior identification model；It includes,
Driving travel data sample processing unit, driving and traveling initial data to acquisition carry out combing cleaning, and selection drives
Running data sample obtains sample driving travel data；
Driving behavior identification model training unit is obtained based on related algorithm and according to sample driving travel data training
Obtain the driving behavior identification model.
13. driving behavior evaluation system as claimed in claim 12, it is characterised in that: it further includes that driving behavior is in danger coefficient
Correlation model training module, the driving behavior be in danger coefficient correlation model training module to the sample driving travel data into
Row training, obtains the driving behavior and is in danger coefficient correlation model；It includes,
Driving behavior analytical unit carries out and number and the amount of money relevance of being in danger point of being in danger sample driving travel data
Analysis, obtains sample driving behavior, and the sample driving behavior includes but is not limited to the feature for clustering and classifying；
Correlation model training unit, based on related algorithm and according to the sample driving behavior and be in danger the frequency and go out
The degree of association training of the dangerous amount of money obtains the driving behavior and is in danger coefficient correlation model.
14. driving behavior evaluation system as claimed in claim 13, it is characterised in that: analyze the sample driving behavior
Consider that the sample driving behavior exists when with the degree of association of the frequency violating the regulations, penalty score, accident frequency and accident severity
The probability occurred in various driving habits；Wherein, the driving habit includes but is not limited to drive the period, and road type is always driven
Sail mileage and total driving duration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811600703.2A CN109658272A (en) | 2018-12-26 | 2018-12-26 | Driving behavior evaluation system and Insurance Pricing system based on driving behavior |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811600703.2A CN109658272A (en) | 2018-12-26 | 2018-12-26 | Driving behavior evaluation system and Insurance Pricing system based on driving behavior |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109658272A true CN109658272A (en) | 2019-04-19 |
Family
ID=66115574
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811600703.2A Pending CN109658272A (en) | 2018-12-26 | 2018-12-26 | Driving behavior evaluation system and Insurance Pricing system based on driving behavior |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109658272A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110288096A (en) * | 2019-06-28 | 2019-09-27 | 江苏满运软件科技有限公司 | Prediction model training and prediction technique, device, electronic equipment, storage medium |
| CN110490752A (en) * | 2019-08-21 | 2019-11-22 | 福州大学 | Car insurance analysis and automatic recommendation service system and its working method based on driving behavior data |
| CN110738574A (en) * | 2019-09-06 | 2020-01-31 | 中国平安财产保险股份有限公司 | Data prediction method, device, equipment and storage medium |
| CN111651473A (en) * | 2020-05-29 | 2020-09-11 | 泰康保险集团股份有限公司 | Policy generation method, device, equipment and computer readable storage medium |
| CN112070553A (en) * | 2020-09-15 | 2020-12-11 | 成都雅信安科技服务有限公司 | Vehicle insurance pricing method and device based on vehicle-mounted data and assisted with insurance company |
| CN112200685A (en) * | 2020-09-30 | 2021-01-08 | 宝能(西安)汽车研究院有限公司 | Vehicle premium calculation method, device, medium, and electronic apparatus |
| CN112288334A (en) * | 2020-11-27 | 2021-01-29 | 上海评驾科技有限公司 | Lightgbm-based car networking risk factor extraction method |
| CN112597835A (en) * | 2020-12-11 | 2021-04-02 | 国汽(北京)智能网联汽车研究院有限公司 | Driving behavior evaluation method and device, electronic equipment and readable storage medium |
| CN112613998A (en) * | 2020-12-16 | 2021-04-06 | 深圳市麦谷科技有限公司 | Vehicle insurance premium pricing method and system based on driving behavior scoring model |
| CN112989069A (en) * | 2021-05-10 | 2021-06-18 | 苏州博宇鑫交通科技有限公司 | Traffic violation analysis method based on knowledge graph and block chain |
| CN113135225A (en) * | 2021-05-28 | 2021-07-20 | 山东新一代信息产业技术研究院有限公司 | Front-turning rear-driving robot chassis and moving robot |
-
2018
- 2018-12-26 CN CN201811600703.2A patent/CN109658272A/en active Pending
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110288096B (en) * | 2019-06-28 | 2021-06-08 | 满帮信息咨询有限公司 | Prediction model training method, prediction model training device, prediction model prediction method, prediction model prediction device, electronic equipment and storage medium |
| CN110288096A (en) * | 2019-06-28 | 2019-09-27 | 江苏满运软件科技有限公司 | Prediction model training and prediction technique, device, electronic equipment, storage medium |
| CN110490752A (en) * | 2019-08-21 | 2019-11-22 | 福州大学 | Car insurance analysis and automatic recommendation service system and its working method based on driving behavior data |
| CN110738574A (en) * | 2019-09-06 | 2020-01-31 | 中国平安财产保险股份有限公司 | Data prediction method, device, equipment and storage medium |
| CN111651473A (en) * | 2020-05-29 | 2020-09-11 | 泰康保险集团股份有限公司 | Policy generation method, device, equipment and computer readable storage medium |
| CN111651473B (en) * | 2020-05-29 | 2024-03-01 | 泰康保险集团股份有限公司 | Policy generation method, device, equipment and computer readable storage medium |
| CN112070553A (en) * | 2020-09-15 | 2020-12-11 | 成都雅信安科技服务有限公司 | Vehicle insurance pricing method and device based on vehicle-mounted data and assisted with insurance company |
| CN112200685A (en) * | 2020-09-30 | 2021-01-08 | 宝能(西安)汽车研究院有限公司 | Vehicle premium calculation method, device, medium, and electronic apparatus |
| CN112288334A (en) * | 2020-11-27 | 2021-01-29 | 上海评驾科技有限公司 | Lightgbm-based car networking risk factor extraction method |
| CN112288334B (en) * | 2020-11-27 | 2024-04-16 | 上海评驾科技有限公司 | Method for extracting Internet of vehicles risk factors based on lightgbm |
| CN112597835A (en) * | 2020-12-11 | 2021-04-02 | 国汽(北京)智能网联汽车研究院有限公司 | Driving behavior evaluation method and device, electronic equipment and readable storage medium |
| CN112613998A (en) * | 2020-12-16 | 2021-04-06 | 深圳市麦谷科技有限公司 | Vehicle insurance premium pricing method and system based on driving behavior scoring model |
| CN112989069A (en) * | 2021-05-10 | 2021-06-18 | 苏州博宇鑫交通科技有限公司 | Traffic violation analysis method based on knowledge graph and block chain |
| CN113135225A (en) * | 2021-05-28 | 2021-07-20 | 山东新一代信息产业技术研究院有限公司 | Front-turning rear-driving robot chassis and moving robot |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109658272A (en) | Driving behavior evaluation system and Insurance Pricing system based on driving behavior | |
| Arbabzadeh et al. | A data-driven approach for driving safety risk prediction using driver behavior and roadway information data | |
| Ma et al. | Driving style recognition and comparisons among driving tasks based on driver behavior in the online car-hailing industry | |
| US11407410B2 (en) | Method and system for estimating an accident risk of an autonomous vehicle | |
| Chen et al. | A graphical modeling method for individual driving behavior and its application in driving safety analysis using GPS data | |
| JP6961012B2 (en) | Customized safe speed determination for the vehicle | |
| EP3578433B1 (en) | Method for estimating an accident risk of an autonomous vehicle | |
| CN109670970B (en) | Driving behavior scoring method and device and computer readable storage medium | |
| US20140180727A1 (en) | System and Method for Classifying and Identifying a Driver Using Driving Performance Data | |
| US9082072B1 (en) | Method for applying usage based data | |
| CN110866677A (en) | Driver relative risk evaluation method based on benchmark analysis | |
| JP6593798B2 (en) | Vehicle, system for communicating with vehicle communication module, and system for communicating with a set of vehicles | |
| CN109840612A (en) | User's driving behavior analysis method and system | |
| CN107784587A (en) | A kind of driving behavior evaluation system | |
| Lyu et al. | Using naturalistic driving data to identify driving style based on longitudinal driving operation conditions | |
| Das et al. | Nonparametric multivariate adaptive regression splines models for investigating lane-changing gap acceptance behavior utilizing strategic highway research program 2 naturalistic driving data | |
| CN108091132A (en) | A kind of traffic flow forecasting method and device | |
| Verma et al. | Avoiding stress driving: Online trip recommendation from driving behavior prediction | |
| CN109993966A (en) | A kind of method and device of building user portrait | |
| CN107045786A (en) | The computational methods of traffic accident interval time occur for the driver based on Kaplan Meier methods | |
| EP4126620B1 (en) | Coachable driver risk groups | |
| CN109784586B (en) | Prediction method and system for danger emergence condition of vehicle danger | |
| Keramati et al. | A crash severity analysis at highway-rail grade crossings: The random survival forest method | |
| Stankevich et al. | Usage-based vehicle insurance: Driving style factors of accident probability and severity | |
| Hanafizadeh et al. | A data mining model for risk assessment and customer segmentation in the insurance industry |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190419 |