CN114119256A - UBI dangerous chemical vehicle driving behavior acquisition and analysis system and premium discount method - Google Patents
UBI dangerous chemical vehicle driving behavior acquisition and analysis system and premium discount method Download PDFInfo
- Publication number
- CN114119256A CN114119256A CN202111326613.0A CN202111326613A CN114119256A CN 114119256 A CN114119256 A CN 114119256A CN 202111326613 A CN202111326613 A CN 202111326613A CN 114119256 A CN114119256 A CN 114119256A
- Authority
- CN
- China
- Prior art keywords
- data
- vehicle
- drivers
- driver
- acceleration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Finance (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Accounting & Taxation (AREA)
- Evolutionary Biology (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to a UBI dangerous chemical vehicle driving behavior acquisition and analysis system and a premium discount method. The system comprises a vehicle-mounted terminal and a risk monitoring platform which are arranged on a dangerous chemical transport vehicle, wherein the vehicle-mounted terminal comprises a vehicle driving behavior acquisition module and a driver driving behavior acquisition module, and the risk monitoring platform comprises a data storage module, a data extraction module, a data visualization module and a data analysis module; the premium discount method provides a corresponding premium discount according to the driver classification. Compared with the prior art, the method classifies the drivers by analyzing the driving behaviors of the vehicles and the driving behavior data of the drivers and adopting a machine learning algorithm, gives different drivers different premium discounts according to the classification of the drivers, and achieves the purposes of reasonably and fairly guaranteeing the premium and improving the traffic safety of the dangerous chemical transport vehicles.
Description
Technical Field
The invention relates to the technical field of automobile monitoring management, in particular to a UBI dangerous chemical vehicle driving behavior acquisition and analysis system and a premium discount method.
Background
The UBI (use Based insurance) vehicle insurance is the vehicle insurance cost determined according to driving behaviors and driving mileage, and compared with the traditional vehicle insurance, a driver with poor driving habits has higher insurance cost than a driver with good driving habits. On one hand, UBI is fairer and more reasonable in vehicle insurance, on the other hand, UBI encourages drivers to drive safely, and is helpful for improving traffic safety and is a future trend. However, the existing UBI car insurance is concerned about private cars, or light passenger cars, and is concerned less about dangerous chemical transport vehicles.
Dangerous chemical transport vehicle accidents have the characteristics of high difficulty in rescue and treatment and serious consequences, and the dangerous chemical accidents are likely to occur in all links of production, operation, transportation, storage, use and abandonment, wherein the accidents in the transportation link are the most, and reach 41% of the total number of accidents, so that how to ensure the driving safety of dangerous chemical transport vehicles is extremely important.
However, the conventional UBI risk driver classification system generally adopts a statistical model, classifies the drivers relatively statically based on a fixed index threshold, has a long classification period, and with the increase of computer computing power and the improvement of data analysis technology, a data-driven UBI risk scheme for adaptively classifying the drivers based on driving behavior data is needed.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a UBI dangerous chemical vehicle driving behavior acquisition and analysis system and a premium discount method.
The purpose of the invention can be realized by the following technical scheme:
a UBI dangerous chemical vehicle driving behavior acquisition and analysis system comprises a vehicle-mounted terminal and a risk monitoring platform which are arranged on a dangerous chemical transport vehicle;
the vehicle-mounted terminal includes:
the vehicle driving behavior acquisition module is used for acquiring the driving behaviors of the dangerous chemical transport vehicle, wherein the driving behaviors comprise vehicle positions, vehicle speed acceleration lane departure alarm times, and forward and lateral alarm times of distances between the vehicle and other vehicles;
the driver driving behavior acquisition module is used for acquiring the driving behavior of a driver of the dangerous chemical transport vehicle, and the driving behavior comprises driver identity information, fatigue alarm times, hand-held call receiving and making alarm times and smoking alarm times;
the risk monitoring platform includes:
the data storage module is used for storing historical data collected by the vehicle-mounted terminal;
the data extraction module is used for extracting the data stored by the data storage module;
the data visualization module is used for displaying the extracted data, and playing back and verifying the alarm video of the driver state and visualizing other data;
and the data analysis module is used for preprocessing the extracted data, identifying bad driving behaviors, classifying drivers, outputting corresponding premium discounts according to classification, and displaying the processed premium results in the data visualization module.
The vehicle driving behavior acquisition module comprises a high-frequency GPS (global positioning system) locator used for acquiring the position and the speed of a vehicle, a nine-axis acceleration sensor used for acquiring acceleration and auxiliary driving equipment used for acquiring lane departure alarm times and forward and lateral over-close alarm times with other vehicles.
The driver driving behavior acquisition module comprises driver state monitoring equipment.
A premium discount method for UBI dangerous chemical vehicles, which employs the UBI dangerous chemical vehicle driving behavior collection and analysis system as described above, the method comprising the steps of:
the vehicle driving behavior acquisition module based on the vehicle-mounted terminal acquires the vehicle position, the vehicle speed, the acceleration, the lane departure alarm times and the alarm times data of too close distance between the front direction and the side direction of other vehicles in real time, and simultaneously acquires the driver identity information, the fatigue alarm times, the hand-held call receiving alarm times and the smoking alarm times data in real time based on the driver driving behavior acquisition module;
outputting each collected data to a data storage module of the risk monitoring platform for storage and recording;
a data extraction module of the risk monitoring platform extracts data in the data storage module and outputs the data to a data analysis module;
the data analysis module preprocesses the received data, identifies bad driving behaviors, and divides the drivers into drivers with different safety levels by adopting a center clustering algorithm; outputting premium discount information according to the classification result;
the premium discount is updated for each driver by driving miles per hundred kilometers.
Further, the data analysis module preprocesses the received data, identifies bad driving behaviors, and divides drivers into four types, namely safe drivers, safer drivers, less-risk drivers and more-risk drivers by adopting a center clustering algorithm.
Furthermore, discount information is not generated for drivers with higher risks, nine-fold and five-fold discount information is generated for drivers with lower risks, nine-fold and five-fold discount information is generated for drivers with higher safety, and eight-fold and five-fold discount information is generated for drivers with safety.
Further, the high-frequency GPS locator acquires a first data table, and the head of the first data table sequentially comprises a vehicle number, a timestamp, a longitude and a latitude from left to right; the method comprises the steps that a nine-axis acceleration sensor obtains a second data table, and the head of the second data table sequentially comprises a vehicle number, a timestamp, an x-direction acceleration, a y-direction acceleration, a z-direction acceleration, an x-direction angular velocity, a y-direction angular velocity, a z-direction angular velocity, an x-direction magnetic field intensity, a y-direction magnetic field intensity and a z-direction magnetic field intensity from left to right; the auxiliary driving equipment acquires a third data table, and the head of the third data table sequentially comprises a vehicle number, a timestamp, an alarm category and a remark from left to right; the driver state monitoring equipment acquires a fourth data table, and the table head sequentially comprises a vehicle number, a driver number, a timestamp, an alarm category and notes from left to right.
Further, the data storage module stores the first data table, the second data table, the third data table and the fourth data table, the data extraction module extracts data stored by the data storage module, five columns of a vehicle number, a time stamp, acceleration in the x direction, acceleration in the y direction and acceleration in the z direction are extracted from the second data table acquired by the nine-axis accelerometer to serve as an acceleration data table, the vehicle number and the time stamp are taken as main keys for the first data table, the acceleration data table, the third data table and the fourth data table to be summarized into one table and transmitted to the data analysis module, and redundant data are discarded on the basis of the data table with the lowest sampling frequency.
Further, the data analysis module preprocesses the vehicle position, speed and acceleration data, and identifies the number of times of overspeed, the number of times of driving outside a route, the number of times of rapid acceleration and rapid deceleration and the number of times of frequent lane change for the preprocessed vehicle position, speed and acceleration data; and clustering the obtained data of each time and other collected data into four types by using a central clustering algorithm, namely drivers in safe driving types, drivers in safer driving types, drivers in less risk types and drivers in greater risk types.
Further, a new table is formed by the serial number of the driver, the accumulated driving distance of the driver, the driving times outside the route, the overspeed times, the rapid acceleration and rapid deceleration times, the frequent lane changing times, the lane departure alarming times, the forward and lateral overtime alarming times with other vehicles, the fatigue alarming times, the hand-held call receiving alarming times and the smoking alarming times, and the four types of the signals are segmented into four types according to the center clustering algorithm of the center clustering algorithm, namely drivers in safe driving types, drivers in safer driving types, drivers in small risk types and drivers in large risk types.
Compared with the prior art, the invention at least comprises the following beneficial effects:
1) the driving behavior data of the dangerous chemical transport vehicle and the driving behavior data of a driver are acquired through a vehicle-mounted terminal arranged on the dangerous chemical transport vehicle; the driving mode of the driver is analyzed by utilizing the data, the driver is classified, different drivers are given different premium discounts according to the classification of the driver, and the purposes of reasonably and fairly guaranteeing the premium and improving the traffic safety of the dangerous chemical transport vehicle are achieved;
2) the method adopts a center clustering algorithm in machine learning when classifying the drivers, has short calculation time compared with the traditional statistical method, adaptively classifies the drivers based on the driving behavior data, and more accords with the actual driving operation condition compared with the method that the drivers are classified relatively statically by a fixed index threshold value, and is more accurate and reasonable;
3) according to the premium discount method, the drivers are classified into different types by classifying the driving behaviors, so that the premium customers and the inferior customers can be identified, and different insurance management can be realized for different customers; according to the invention, discount information is generated according to the classification result and is sent to the driver, when different customers carry out insurance management, a certain discount can be given to the high-quality driver according to the discount information, so that the high-quality driver is encouraged to keep good driving behavior, and no discount is given to the poor-quality driver, so that the driving behavior of the poor-quality driver is normalized, the relative fairness and reasonableness of insurance products are realized, and the vehicle insurance cost of the high-quality driver is reduced; in addition, the method can also advocate drivers to travel green and drive civilized, thereby reducing the accident rate of emergency and improving the safety of driving.
Drawings
FIG. 1 is a block diagram of a driving behavior acquisition and analysis system for UBI hazardous chemical vehicles in an embodiment;
FIG. 2 is a schematic flow chart of a premium discount method for UBI hazardous chemical vehicles in an embodiment.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of protection of the present invention.
Examples
As shown in FIG. 1, the invention relates to a UBI dangerous chemical vehicle driving behavior acquisition and analysis system, which comprises a vehicle-mounted terminal and a risk monitoring platform. Vehicle-mounted terminal and risk monitoring platform all set up on dangerous chemical transport vehicle.
The vehicle-mounted terminal comprises a vehicle driving behavior acquisition module and a driver driving behavior acquisition module.
The vehicle driving behavior acquisition module comprises a high-frequency GPS locator, a nine-axis acceleration sensor and driving auxiliary equipment (ADAS);
the vehicle driving behavior acquisition module is used for acquiring the driving behaviors of the dangerous chemical transport vehicle, and comprises vehicle position (longitude and latitude), vehicle speed, acceleration, lane departure alarm times, forward and lateral over-close alarm times between other vehicles and the like, wherein the vehicle position and the vehicle speed are acquired by a high-frequency GPS (global positioning system) positioning instrument, the acceleration is acquired by a nine-axis acceleration sensor, the lane departure alarm times, and the forward and lateral over-close alarm times between other vehicles and auxiliary driving equipment.
The driver driving behavior acquisition module mainly comprises driver state monitoring equipment (DSM);
the driver driving behavior acquisition module is used for acquiring the driving behaviors of the driver of the dangerous chemical transport vehicle, wherein the driving behaviors comprise driver identity information, fatigue alarm times, hand-held call receiving and making alarm times, smoking alarm times and the like.
The risk monitoring platform comprises a data storage module, a data extraction module, a data visualization module and a data analysis module.
The data storage module is connected with the vehicle-mounted terminal and used for storing historical data acquired by the terminal;
the data extraction module is connected with the data storage module and used for extracting the data stored by the data storage module;
the data visualization module is connected with the data extraction module and used for displaying the extracted data; driver status warning videos and other visual data may also be played back, verified, and visualized by the data visualization module as needed.
And the data analysis module is connected with the data visualization module and used for preprocessing the extracted data, identifying bad driving behaviors, classifying drivers, outputting corresponding premium discounts according to classification and displaying the processed premium results in the data visualization module.
In order to achieve the above object, referring to fig. 2, the present invention further provides a premium discount method for a UBI hazardous chemical vehicle, which adopts the UBI hazardous chemical vehicle driving behavior collecting and analyzing system, and specifically includes the following steps:
s1: the vehicle driving behavior acquisition module of the vehicle-mounted terminal of the dangerous chemical transport vehicle acquires data of vehicle position, vehicle speed, acceleration, lane departure alarm times, forward and lateral over-close alarm times with other vehicles in real time, and the driver driving behavior acquisition module acquires data of driver identity information, fatigue alarm times, hand-held call receiving alarm times and smoking alarm times in real time.
S2: and outputting the collected driving behavior data of each vehicle and the driving behavior data of the driver to a data storage module of the risk monitoring platform for storage and recording.
S3: during analysis, the data in the data storage module is extracted by the data extraction module of the risk monitoring platform and is output to the data analysis module.
S4: the data analysis module preprocesses the received data, identifies bad driving behaviors, classifies drivers and outputs premium discounts according to classification; the method comprises the following specific steps:
s41: preprocessing vehicle position, speed and acceleration data, namely removing abnormal values from the received position, speed and acceleration data, and performing interpolation and filtering; specifically, the method comprises the following steps:
s411, setting a difference threshold value between the GPS and two adjacent points of the acceleration, and deleting the difference value larger than the threshold value as an abnormal value;
s412, traversing the GPS and the acceleration data according to time, finding out a position segment of the missing value, and linearly interpolating and filling up the missing value by using a previous value and a next value of the missing segment;
s413, inputting the data sequence after the deficiency value is filled into Kalman filtering to obtain preprocessed position, speed and acceleration data;
s42: identifying overspeed times, off-route driving times, rapid acceleration and rapid deceleration times and frequent lane changing times for the preprocessed position, speed and acceleration data;
further, the specific process of identifying the risky driving behavior in this step is as follows: giving a reference distance, dividing the driving distance outside the specified route by the reference distance to obtain the driving times outside the route, and recording the driving times less than the reference distance as one time; overspeed and rapid acceleration and rapid deceleration are carried out in the same way, a reference distance is given, the overspeed frequency is obtained by dividing the speed-limited driving distance exceeding the road by the reference distance, and only the reference distance is different from the driving outside the route; giving a reference distance, and dividing the driving distance of which the absolute value of the acceleration exceeds a given acceleration threshold value by the reference distance to obtain the number of rapid acceleration and rapid deceleration; the frequent lane change is different from the above, a reference distance is given, and lane change is recorded as a frequent lane change when the vehicle does not travel the full reference distance after the lane change.
S43: clustering the data obtained in the step S43 and other collected data into four types by using a central clustering algorithm, namely drivers in safe driving types, drivers in safer driving types, drivers in less risk types and drivers in greater risk types;
and S44, outputting corresponding premium discounts for various drivers.
S5: the premium discount is updated for each driver every hundred kilometers of travel.
Optionally, the driver status warning video and other data may be played back, verified, and visualized by the data visualization module as needed.
To more clearly illustrate the above process of the present invention, the following is further illustrated with reference to specific examples:
step 1: a data table is obtained by a high-frequency GPS locator, and a table head sequentially comprises a vehicle number, a timestamp, longitude and latitude from left to right. A data table is obtained by a nine-axis accelerometer, and a meter head sequentially comprises a vehicle number, a time stamp, an x-direction acceleration, a y-direction acceleration, a z-direction acceleration, an x-direction angular velocity, a y-direction angular velocity, a z-direction angular velocity, an x-direction magnetic field intensity, a y-direction magnetic field intensity and a z-direction magnetic field intensity from left to right. A data table is obtained by auxiliary driving equipment, and a table head sequentially comprises a vehicle number, a time stamp, an alarm category (lane departure alarm, forward and lateral over-close alarm with other vehicles) and remarks from left to right. The data sheet is obtained by the driver state monitoring equipment, and the sheet head sequentially comprises a vehicle number, a driver number, a timestamp, an alarm category (fatigue alarm, hand-held call receiving and making alarm, smoking alarm) and remarks from left to right.
Step 2: and (3) transmitting the data acquired in the step (1) to a risk monitoring platform, and storing the original data by a data storage module of the risk monitoring platform.
And step 3: and (2) extracting the data stored in the step (2) by a data extraction module of the risk monitoring platform, wherein a data table of the nine-axis accelerometer only extracts five columns of vehicle numbers, timestamps, acceleration in the x direction, acceleration in the y direction and acceleration in the z direction, summarizing the GPS data table, the acceleration data table, the vehicle alarm data table and the driver alarm data table into one table by taking the vehicle numbers and the timestamps as main keys (keys) and transmitting the table to a data analysis module, and abandoning redundant data by taking the data table with the lowest sampling frequency as the standard.
And 4, step 4: receiving a data summary table transmitted by a data extraction module, extracting four rows of data of a vehicle number, a timestamp, longitude and latitude, traversing the data to calculate the distance between two rows, marking a point with the distance being more than 23 meters as an abnormal point, marking a real abnormal point as a previous point and a next point, and completely assigning two rows of longitude and latitude of a continuous point, namely the abnormal point, except a head point and a tail point, with null values; then, traversing and finding out a null value, and using the previous point and the next point of the null value segment to perform linear interpolation to obtain the longitude and the latitude of the null value; the four columns of data are then filtered using a kalman filter algorithm.
The specific filtering steps are initializing a Kalman filter, setting an initial state (an initial error covariance matrix and a covariance matrix), predicting a target position (substituting the initial state and a state transition matrix into a state prediction equation to obtain a predicted value of a current motion state, substituting the initial error covariance matrix, the covariance matrix and the state transition matrix into the error covariance prediction equation to obtain a predicted value of an error covariance), and matching the target position (within a target predicted position range) to search the target by using the position characteristics of the target. If there is only one object in the search range, then the object is the object to be tracked. If a plurality of targets are found in the search range, determining which target is to be tracked according to the minimum value of the target template matching distance), updating filter parameters (solving Kalman gain by using a Kalman gain equation), substituting target coordinates as an observed value into a state correction equation to obtain a corrected state vector, and correcting an error covariance matrix according to the covariance correction equation.
Repeating the steps to obtain four rows of data of the filtered vehicle number, the timestamp, the longitude and the latitude, obtaining the accumulated running distance of the driver number, the off-route running distance and the vehicle speed according to the longitude and the latitude, dividing the off-route running distance by 500 meters, rounding upwards to obtain the off-route running times, obtaining the overspeed running distance according to the vehicle speed, and dividing the overspeed running distance by 500 meters to obtain the overspeed times; for triaxial acceleration data, the longitudinal acceleration is greater than 0.5g and less than-0.5 g, the rapid acceleration and rapid deceleration are regarded as rapid acceleration and rapid deceleration, and the rapid acceleration and rapid deceleration running distance is divided by 5m to obtain the rapid acceleration and rapid deceleration times; the change amplitude of the transverse acceleration exceeds 0.2g and is regarded as a lane change, the lane change for the second time within 50 meters is regarded as a frequent lane change for the first time, the lane change for the third time is regarded as a frequent lane change for the second time, and the rest is done in the same way;
forming a new table by the serial number of a driver, the accumulated driving distance of the driver, the times of driving outside a route, the times of overspeed, the times of rapid acceleration and rapid deceleration, the times of frequent lane change, the times of lane departure alarm, the times of alarm of too close distance between the forward direction and the lateral direction and other vehicles, the times of fatigue alarm, the times of hand-held call receiving and making alarm and the times of smoke alarm, sending the new table into a center clustering algorithm by sections according to one hundred kilometers, clustering drivers into four categories by the center clustering algorithm, and recording the times of driving outside the route as x1The number of overspeed is x2The number of rapid acceleration and rapid deceleration is x3The frequency of lane change is x4The number of times of lane departure warning is x5The number of the over-close alarming times of the distance between the front direction and the side direction and other vehicles is x6The fatigue alarm frequency is x7The number of the hand-held call receiving and making alarm times is x8The smoking alarm frequency is x9As shown in the data schematic table, four types obtained by clustering are respectively calculated:
wherein:
indicating average dangerous driving times, x, of drivers in classijRepresenting the j dangerous driving behavior frequency of the ith driver in the class, j representing the j dangerous driving behavior frequency, i representing the ith driver in the class, and n representing n drivers in the class, for example, in the data representation table in the table 1, the four clustered classes n are respectively 3,2,2 and 3;
TABLE 1 Cluster calculation data schematic
In the invention, n can be automatically obtained by an algorithm or can be artificially counted. The dangerous chemical vehicles tested can be the same or different, the data is sorted according to the serial numbers of the drivers, the same vehicle is driven by different drivers, the data can be counted by different drivers, different drivers drive different vehicles and also count different drivers, the same driver drives different vehicles, and the driver can also count the dangerous chemical vehicles as long as the hardware equipment is installed.
4 pieces to be obtained
In the order from small to large,
the minimum class drivers are safe driving class drivers,
Secondly, the driver is safer, and secondly, the driver is less risky
The largest class is the more risky class of drivers.
The specific steps of the center clustering algorithm are (1) randomly selecting initial centers of four clusters, (2) solving Euclidean distances from each sample point to the four centers, marking the samples as the class of the center closest to the sample point, updating the clustering centers to be the mean value of the sample points of the class to which the clustering centers belong after marking is finished, and repeating (2) (3) and max _ iter times.
Taking the open source programming language python as an example, the code of the clustering part may be:
from sklearn.cluster import KMeans
km＝KMeans(n_clusters＝4,max_iter＝300)
label＝km.fit_predict(data)
wherein the data is the driver number and x1To x9According to a data table formed by hundred kilometers, n _ cluster is a required class number, is fixed to be 4, max _ iter is a required iteration number, can be set to be 300, if higher precision is required, a larger value can be filled, but corresponding calculation time is increased, if the calculation time is too long, a smaller value can be filled, and a specific value is comprehensively determined according to required precision, calculation force and calculation time which can be borne.
And generating premium discount information according to the classified risk condition and sending the premium discount information to a driver of the dangerous goods transport vehicle. As one of the preferable embodiments, the premium can be discounted to drivers with higher risk, nine folds and five folds for drivers with lower risk, nine folds for drivers with higher safety and eight folds and five folds for drivers with safety.
And 5: the premium discount is updated every hundred kilometers.
The driving behavior data of the dangerous chemical transport vehicle and the driving behavior data of a driver are acquired through a vehicle-mounted terminal arranged on the dangerous chemical transport vehicle; the driving mode of the driver is analyzed by utilizing the data, the driver is classified, different premium discounts are given to different drivers according to the classification of the driver, and the purposes of reasonably and fairly guaranteeing the premium and improving the traffic safety of the dangerous chemical transport vehicle are achieved. The method adopts a center clustering algorithm in machine learning when classifying the drivers, has short calculation time compared with the traditional statistical method, adaptively classifies the drivers based on the driving behavior data, and more accords with the actual driving operation condition compared with the condition that the drivers are classified relatively statically by a fixed index threshold value, and is more accurate and reasonable.
While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and those skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A UBI dangerous chemical vehicle driving behavior acquisition and analysis system is characterized by comprising a vehicle-mounted terminal and a risk monitoring platform which are arranged on a dangerous chemical transport vehicle;
the vehicle-mounted terminal includes:
the vehicle driving behavior acquisition module is used for acquiring the driving behaviors of the dangerous chemical transport vehicle, wherein the driving behaviors comprise vehicle positions, vehicle speed acceleration lane departure alarm times, and forward and lateral alarm times of distances between the vehicle and other vehicles;
the driver driving behavior acquisition module is used for acquiring the driving behavior of a driver of the dangerous chemical transport vehicle, and the driving behavior comprises driver identity information, fatigue alarm times, hand-held call receiving and making alarm times and smoking alarm times;
the risk monitoring platform includes:
the data storage module is used for storing historical data collected by the vehicle-mounted terminal;
the data extraction module is used for extracting the data stored by the data storage module;
the data visualization module is used for displaying the extracted data, and playing back and verifying the alarm video of the driver state and visualizing other data;
and the data analysis module is used for preprocessing the extracted data, identifying bad driving behaviors, classifying drivers, outputting corresponding premium discounts according to classification, and displaying the processed premium results in the data visualization module.
2. The UBI hazardous chemicals vehicle driving behavior collection and analysis system according to claim 1, wherein the vehicle driving behavior collection module comprises a high frequency GPS locator to obtain vehicle position and speed, a nine axis acceleration sensor to obtain acceleration, and a driver assistance device to obtain lane departure warning times and forward and lateral over-close warning times to other vehicles.
3. The UBI hazardous chemicals vehicle driving behavior collection and analysis system according to claim 1, characterized in that said driver driving behavior collection module comprises a driver status monitoring device.
4. A premium discount method of a UBI hazardous chemical vehicle using the UBI hazardous chemical vehicle driving behavior collection and analysis system according to any one of claims 1 to 3, comprising the steps of:
the vehicle driving behavior acquisition module based on the vehicle-mounted terminal acquires the vehicle position, the vehicle speed, the acceleration, the lane departure alarm times and the alarm times data of too close distance between the front direction and the side direction of other vehicles in real time, and simultaneously acquires the driver identity information, the fatigue alarm times, the hand-held call receiving alarm times and the smoking alarm times data in real time based on the driver driving behavior acquisition module;
outputting each collected data to a data storage module of the risk monitoring platform for storage and recording;
a data extraction module of the risk monitoring platform extracts data in the data storage module and outputs the data to a data analysis module;
the data analysis module preprocesses the received data, identifies bad driving behaviors, and divides the drivers into drivers with different safety levels by adopting a center clustering algorithm; outputting premium discount information according to the classification result;
the premium discount is updated for each driver by driving miles per hundred kilometers.
5. The method of claim 4, wherein the data analysis module preprocesses the received data to identify poor driving behavior and categorizes drivers into four categories, safe drivers, safer drivers, less risky drivers, and more risky drivers using a center clustering algorithm.
6. The method of claim 5, wherein the discount information is not generated for a driver with a higher risk, the discount information is generated for a driver with a lower risk, the discount information is generated for a nine-fold and a five-fold for a driver with a higher risk, and the discount information is generated for a eight-fold and a five-fold for a driver with a higher safety.
7. The method of discount premium discount on UBI hazardous chemicals vehicle of claim 4, wherein the high frequency GPS locator obtains the first data table whose header is vehicle number, time stamp, longitude and latitude from left to right; the method comprises the steps that a nine-axis acceleration sensor obtains a second data table, and the head of the second data table sequentially comprises a vehicle number, a timestamp, an x-direction acceleration, a y-direction acceleration, a z-direction acceleration, an x-direction angular velocity, a y-direction angular velocity, a z-direction angular velocity, an x-direction magnetic field intensity, a y-direction magnetic field intensity and a z-direction magnetic field intensity from left to right; the auxiliary driving equipment acquires a third data table, and the head of the third data table sequentially comprises a vehicle number, a timestamp, an alarm category and a remark from left to right; the driver state monitoring equipment acquires a fourth data table, and the table head sequentially comprises a vehicle number, a driver number, a timestamp, an alarm category and notes from left to right.
8. The method for discount premiums of UBI dangerous chemical vehicles according to claim 7, wherein the data storage module stores the first data table, the second data table, the third data table and the fourth data table, the data extraction module extracts the data stored in the data storage module, five columns of vehicle numbers, time stamps, x-direction acceleration, y-direction acceleration and z-direction acceleration are extracted from the second data table obtained by the nine-axis accelerometer to serve as acceleration data tables, the vehicle numbers and the time stamps are taken as main keys for the first data table, the acceleration data table, the third data table and the fourth data table to be collected into one table and transmitted to the data analysis module, and redundant data are discarded based on the data table with the lowest sampling frequency.
9. The method of claim 8, wherein the data analysis module preprocesses vehicle position, speed, and acceleration data and identifies, for the preprocessed vehicle position, speed, and acceleration data, the number of speeding, off-route travel, rapid acceleration, rapid deceleration, and frequent lane changes; and clustering the obtained data of each time and other collected data into four types by using a central clustering algorithm, namely drivers in safe driving types, drivers in safer driving types, drivers in less risk types and drivers in greater risk types.
10. The method of claim 9, wherein the driver number, the accumulated driving distance of the driver, the off-route driving times, the overspeed times, the rapid acceleration and deceleration times, the frequent lane change times, the lane departure warning times, the forward and lateral too close to other vehicles warning times, the fatigue warning times, the hand-held call receiving warning times and the smoking warning times are formed into a new table, and the new table is segmented into four categories by a central clustering algorithm of a central clustering algorithm, namely, safe driving drivers, safer drivers, less risky drivers and more risky drivers.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111326613.0A CN114119256A (en) | 2021-11-10 | 2021-11-10 | UBI dangerous chemical vehicle driving behavior acquisition and analysis system and premium discount method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111326613.0A CN114119256A (en) | 2021-11-10 | 2021-11-10 | UBI dangerous chemical vehicle driving behavior acquisition and analysis system and premium discount method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114119256A true CN114119256A (en) | 2022-03-01 |
Family
ID=80378076
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111326613.0A Pending CN114119256A (en) | 2021-11-10 | 2021-11-10 | UBI dangerous chemical vehicle driving behavior acquisition and analysis system and premium discount method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114119256A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115297146A (en) * | 2022-08-04 | 2022-11-04 | 上海移为通信技术股份有限公司 | Driving behavior monitoring device, driving behavior evaluation method and insurance fee determination method |
-
2021
- 2021-11-10 CN CN202111326613.0A patent/CN114119256A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115297146A (en) * | 2022-08-04 | 2022-11-04 | 上海移为通信技术股份有限公司 | Driving behavior monitoring device, driving behavior evaluation method and insurance fee determination method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Brombacher et al. | Driving event detection and driving style classification using artificial neural networks | |
| EP2710572B1 (en) | Vehicle data analysis method and vehicle data analysis system | |
| KR20140031435A (en) | Diagnostic system and method for the analysis of driving behavior | |
| CN111452799A (en) | Driving behavior evaluation method and system | |
| CN110276954A (en) | Vehicle driving behavior integration methods of marking based on BEI-DOU position system | |
| CN114999181B (en) | Highway vehicle speed abnormality identification method based on ETC system data | |
| CN106956680A (en) | A kind of electric automobile driving behavior identifying and analyzing method | |
| EP3578433A1 (en) | Method for estimating an accident risk of an autonomous vehicle | |
| CN113436432A (en) | Method for predicting short-term traffic risk of road section by using road side observation data | |
| Wu et al. | Clustering of several typical behavioral characteristics of commercial vehicle drivers based on GPS data mining: Case study of highways in China | |
| Wu et al. | Abnormal driving behavior detection for bus based on the Bayesian classifier | |
| CN114119256A (en) | UBI dangerous chemical vehicle driving behavior acquisition and analysis system and premium discount method | |
| CN115965235A (en) | Public transport city bus accident risk factor analysis method | |
| CN114912689A (en) | Map grid index and XGBOST-based over-limit vehicle destination prediction method and system | |
| CN114822044B (en) | Driving safety early warning method and device based on tunnel | |
| CN113761715B (en) | Method for establishing personalized vehicle following model based on Gaussian mixture and hidden Markov | |
| CN115205086A (en) | Abnormal track detection method and system driven by big data of moving track and electronic equipment | |
| CN115688003A (en) | Driver identification method and device, computer equipment and readable storage medium | |
| CN114331181A (en) | Vehicle driving behavior risk analysis method based on big data | |
| CN113505955A (en) | User driving behavior scoring method based on TSP system | |
| CN112036709B (en) | Random forest based rainfall weather expressway secondary accident cause analysis method | |
| CN111775948B (en) | Driving behavior analysis method and device | |
| CN111798091B (en) | Expressway lane change scoring model building method based on lane change duration | |
| CN113147781A (en) | Intelligent driving information display system for automobile | |
| CN116383678B (en) | Method for identifying abnormal speed change behavior frequent road sections of operating passenger car |
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 |