CN116665342B - New energy automobile driving behavior analysis method, system and equipment - Google Patents
New energy automobile driving behavior analysis method, system and equipment Download PDFInfo
- Publication number
- CN116665342B CN116665342B CN202310954193.3A CN202310954193A CN116665342B CN 116665342 B CN116665342 B CN 116665342B CN 202310954193 A CN202310954193 A CN 202310954193A CN 116665342 B CN116665342 B CN 116665342B
- Authority
- CN
- China
- Prior art keywords
- data
- position data
- vehicle
- driving
- risk assessment
- 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.)
- Active
Links
- 238000004458 analytical method Methods 0.000 title claims description 22
- 230000006399 behavior Effects 0.000 claims abstract description 91
- 238000012502 risk assessment Methods 0.000 claims abstract description 32
- 238000000034 method Methods 0.000 claims abstract description 27
- 238000004140 cleaning Methods 0.000 claims abstract description 21
- 238000007619 statistical method Methods 0.000 claims abstract description 8
- 238000011156 evaluation Methods 0.000 claims description 33
- 238000000605 extraction Methods 0.000 claims description 19
- 230000001133 acceleration Effects 0.000 claims description 15
- 230000008859 change Effects 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 5
- 230000003542 behavioural effect Effects 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 230000002354 daily effect Effects 0.000 description 24
- 238000010606 normalization Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 230000000875 corresponding effect Effects 0.000 description 7
- 206010039203 Road traffic accident Diseases 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000008713 feedback mechanism Effects 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- 239000000295 fuel oil Substances 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000004579 scanning voltage microscopy Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0808—Diagnosing performance data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
- G06F18/15—Statistical pre-processing, e.g. techniques for normalisation or restoring missing data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
Abstract
The application provides a method, a system and equipment for analyzing driving behaviors of a new energy automobile. The method comprises the following steps: obtaining geographic position data formed by running of the vehicle, and cleaning the obtained geographic position data; extracting feature data taking a feature factor as a unit from the geographical position data through statistical analysis of the cleaned geographical position data; performing risk assessment on driving behaviors corresponding to the geographic position data in a clustering mode on the characteristic data; and displaying and feeding back the risk assessment result. The method, the system and the equipment for analyzing the driving behavior of the new energy automobile can improve the accuracy of the driving behavior safety assessment of the new energy automobile owner, and promote drivers to improve the driving behavior so as to reduce the accident rate.
Description
Technical Field
The application relates to the technical field of new energy automobiles, in particular to a method, a system and equipment for analyzing driving behaviors of a new energy automobile.
Background
In recent years, with the rapid development of new energy automobile industry in China, the conservation amount is continuously improved. However, because the intelligent driving degree of the new energy vehicle is high, the driver is unfamiliar with the driving characteristics of the new energy vehicle, and the new energy vehicle has high accident rate and high accident cost due to factors such as high maintenance price of parts of the new energy vehicle. Correspondingly, compared with the traditional fuel oil vehicle, the insurance cost of the new energy vehicle is obviously improved, and the use cost of the new energy vehicle is increased. In order to solve the problem, the automobile insurance industry is highly demanded to introduce driving behavior habit factors of drivers in the process of pricing automobile insurance, so that the insurance pricing structure is reasonably optimized to scientifically price the 'from person' factors. For example, for drivers who are fatigued, violated and driven violently, the insurance premium of the vehicles is increased; for the driver with good driving behavior habit, the insurance premium of the vehicle is reduced. By the method, the service of the new energy automobile insurance industry is optimized; meanwhile, the feedback to the vehicle owner and the marketized price factor of insurance are evaluated through driving behaviors, so that the improvement of driving behaviors of the driver can be promoted, the occurrence probability of accidents is reduced, and the overall safety level of social traffic is improved. Since drivers are often the main cause of traffic accidents, driving behavior evaluation systems become an effective means to solve such problems. There are some driving behavior analysis technologies of internet of vehicles data in China, such as patent CN105730450B, CN111739194A, CN115027485a and the like. However, some of the analysis applications based on the traditional fuel vehicle data or based on the private protocol data of the host factory fail to meet the new energy vehicle data standard and the analysis of the behavior characteristics of the driver.
Disclosure of Invention
The application provides a method, a system and equipment for analyzing driving behaviors of a new energy automobile, which can improve the accuracy of the safety evaluation of the driving behaviors of the new energy automobile owners, and promote drivers to improve the driving behaviors so as to reduce accident rate.
In view of the above problems, the application provides a method, a system and equipment for analyzing driving behaviors of a new energy automobile.
In a first aspect, the present application provides a method for analyzing driving behavior of a new energy automobile, where the method includes:
obtaining geographic position data formed by running of the vehicle, and cleaning the obtained geographic position data;
extracting feature data taking a feature factor as a unit from the geographical position data through statistical analysis of the cleaned geographical position data;
performing risk assessment on driving behaviors corresponding to the geographic position data in a clustering mode on the characteristic data;
and displaying and feeding back the risk assessment result.
In a second aspect, the present application provides a new energy automobile driving behavior analysis system, the system comprising:
the acquisition module is used for acquiring geographic position data formed by running of the vehicle and cleaning the acquired geographic position data;
the extraction module is used for extracting the characteristic data taking the characteristic factors as units from the geographical position data through statistical analysis of the cleaned geographical position data;
the evaluation module is used for performing risk evaluation on driving behaviors corresponding to the geographic position data in a clustering mode on the characteristic data;
and the display module is used for displaying and feeding back the risk assessment result.
In a third aspect, the present application provides an electronic device, including:
a memory for storing executable instructions;
and the processor is used for realizing the new energy automobile driving behavior analysis method when executing the executable instructions stored in the memory.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
1. the basic data of the driving behavior evaluation uses the national unified data standard GB/T32960;
2. no dependence is caused on vehicle accident or risk data in the evaluation process, so that the acquisition difficulty of a data sample is reduced;
3. the evaluation factors and the process are visible, and the driver can understand the evaluation factors and the process, so that the driver can conveniently improve the behavior of the driver;
4. after feature extraction, the effective factors are screened out by using a K-means clustering method, so that the calculation difficulty and the data acquisition difficulty are reduced.
Drawings
FIG. 1 is a flow chart of a new energy automobile driving behavior analysis method provided by an embodiment of the application;
FIG. 2 is a flow chart of data cleansing according to an embodiment of the present application;
FIG. 3 is a flow chart of feature extraction provided by an embodiment of the present application;
FIG. 4 is a flow chart of risk assessment provided by an embodiment of the present application;
FIG. 5 is a flow chart of the result presentation and feedback provided by the embodiment of the application;
FIG. 6 is a block diagram of a new energy automobile driving behavior analysis system provided by an embodiment of the application;
fig. 7 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The application aims to realize a general driving behavior evaluation system based on new energy vehicle data so as to better improve the accuracy of driving behavior safety evaluation of new energy vehicle owners, promote drivers to improve driving behaviors and further reduce accident rate. Meanwhile, according to the implementation of GB/T32960 standard, networking data basic fields generated by new energy automobiles are unified and standardized, and a basis is provided for more widely analyzing the behaviors of new energy automobiles and drivers.
The application relates to a driving behavior analysis and evaluation system based on national standard GB/T32960 basic data standard, which aims to comprehensively analyze and evaluate the driving behavior characteristics of a new energy automobile. The system not only provides safety scores and driving advice. Through the driving behavior analysis of the driver, the insurance company can use the system to classify the driving behavior of the driver for targeted insurance products and pricing strategies, and provide more competitive premium and more operation service modes conforming to the habit of vehicles for the driver. As new energy automobiles have become an important travel mode, it has become very important to improve the safe driving level of new energy automobiles. The technical scheme has wide practical application prospect and can promote the rapid development of the intelligent travel field.
Meanwhile, in the conventional driving behavior evaluation model building process, the inventor often needs to match the driving behavior of the vehicle with the vehicle accident data, extract driving behavior characteristics through SVM, machine learning, neural network and other methods, and fit the possibility of accident of the vehicle, so that the risk of driving behavior is evaluated. However, it is difficult to obtain a large amount of vehicle accident and internet of vehicles data and match the vehicle accident and the internet of vehicles data in the actual process, and problems of over-fitting of a model, weak generalization capability of the model and the like often occur due to a small sample size. Meanwhile, because the machine learning process is that the black box finally affects a large factor on driving behavior, the interpretability is poor, and clear guidance can not be provided for the behavior improvement of a driver.
Fig. 1 shows a flowchart of a new energy automobile driving behavior analysis method provided by an embodiment of the application. Referring to fig. 1, the new energy automobile driving behavior analysis method includes:
s11, geographic position data formed by running of the vehicle are obtained, and the obtained geographic position data are cleaned.
Fig. 2 shows a flow chart of data cleansing provided by an embodiment of the present application. Referring to fig. 2, in an embodiment of the present application, the data cleansing operation includes the following operation steps:
s21, cleaning the repeated geographic position data.
In the internet of vehicles data transmission process, in order to ensure that vehicle data can be completely uploaded to a platform, a mode of uploading the same vehicle-end data for multiple times is often adopted. This results in the original data containing a complete repetition of the data content. The system needs to accurately filter the content of the data item according to the data uploading time, group the same data points and delete the data points in any group.
S22, cleaning the geographical position data with abnormal positioning.
First it is determined which data points are outliers. Screening the data according to the value range of each data item in the satellite positioning data items, wherein the screening content of the data items comprises:
positioning time ranges should be within 00:00:00-23:59:59;
longitude ranges between-180 degrees and 180 degrees;
the latitude range should generally be between-90 degrees and 90 degrees;
the speed should range between 0 and 200;
the data beyond the data range should be the single point data exception of the positioning data and should be removed.
S23, cleaning the geographic position data of the positioning drift.
Because the positioning data quality is affected by the satellite communication signal quality, weather, environment and other factors, there is a possibility that the positioning point deviates far from the actual vehicle position in a short time. The system needs to determine whether the data points have drifted by detecting their timing characteristics (e.g., time stamps or position coordinates, etc.). A data point is considered a drift point if its timing characteristics change significantly from other data points. Next, these drift points need to be deleted from the dataset. Some interpolation algorithms, such as linear interpolation or cubic spline interpolation, may be used to estimate the correct position of the drift point and replace it with the data point corresponding to the estimated position.
S12, extracting feature data taking a feature factor as a unit from the geographical position data through statistical analysis of the cleaned geographical position data.
FIG. 3 shows a flow chart of feature extraction provided by an embodiment of the present application. Referring to fig. 3, in an embodiment of the present application, feature extraction includes the following operation steps:
and S31, extracting daily behavior characteristics of the driving behavior of the vehicle.
After the data are cleaned, the system can conduct behavior feature extraction on the data collected by each vehicle every day. Daily behavior characteristics represent the behavior characteristics of a driver on each day, and the system can evaluate the safety of the behavior and the habit of the vehicle according to the variation trend of the behavior of the driver in a period of time.
The content of the daily data features includes:
average mileage per day: by calculating the total mileage of the vehicle in a certain period of time and then dividing by the number of days in that period of time. The GPS positioning data may be used to calculate the mileage of the vehicle.
Daily time period mileage: conventional driving behavior feature extraction is often described by using features such as night driving, but in practice, the factors of unified Beijing time, large territorial area and large longitude span are considered. The same time is different for the driver's knowledge. Therefore, in order to more objectively describe the driving intensity of a driver in different time periods every day, the driving mileage in each time period is counted by taking 2 hours as a sampling frequency.
Average daily travel time: obtained by calculating the time of travel of the vehicle over a period of time and then dividing by the number of days of the period of time. The GPS positioning data may be used to calculate the travel time of the vehicle.
Parking duration: based on the vehicle GPS data, a time interval for each stop is calculated. If the time interval exceeds a certain threshold, a new trip is determined. Therefore, the rest time of the driver can be calculated by counting the parking time period of each parking.
Daily acceleration rate of change: by calculating the acceleration change rate of the vehicle in a certain time, the acceleration performance and driving behavior of the vehicle can be known. The GPS positioning data may be used to calculate the acceleration rate of the vehicle.
Daily maximum vehicle speed: by calculating the maximum speed of the vehicle in a certain time, the running capability and driving behavior of the vehicle can be known. The GPS positioning data may be used to calculate the maximum speed of the vehicle.
Number of sudden braking: the time interval for each braking event is subtracted from the speed of the vehicle before braking and if the interval is less than a certain threshold, it is considered a sudden braking event. This characteristic value is calculated by counting the number of sudden braking events per day.
Number of rapid acceleration: similar to the calculation of the number of sudden braking events, but the characteristic value is calculated by counting the number of acceleration events.
Number of sharp turns: from the vehicle sensor data and the GPS data, the radius and the turning angle of each turn are calculated and correlated with the vehicle speed and acceleration to identify a sharp turn event. This feature value is calculated by counting the number of sharp turn events per day.
Daily driving route: by recording the travel route of the vehicle in a certain time, the travel track and the use habit of the vehicle can be known. The GPS positioning data may be used to calculate a travel route of the vehicle.
Travel route and road type: from the vehicle GPS data, the route and the road type traversed (e.g., expressway, urban road, rural road, etc.) for each trip are identified and compared to local traffic rules and accident data to assess risk levels for different routes and road types. This characteristic value can be used to set different risk factors for different routes and road types.
S32, extracting overall behavior characteristics of the driving behavior of the vehicle.
Analysis and evaluation of driving behavior requires the behavior characteristics of one driver to be described from the whole. The overall behavior characteristics are considered to describe the behavior characteristics and behavior change trend of the driver, and specifically comprise the following large classifications:
driving strength characteristics:
days of observation: last active date-first network entry date, which is used to describe the length of time the system observes the vehicle, the longer the observation time is, the better the system knows about the behavior of the vehicle. The observation day indicates the credibility of the evaluation of the driving behavior.
Days of activity: and in the observation date, the number of days when the system collects the data of the new energy automobile is larger, and the number of days is larger, so that the number of days when the vehicle runs is larger.
Activity rate: liveness = liveness days/days observed100%。
Accumulated mileage: and the sum of daily driving mileage in observation days.
Accumulating time length: the sum of the daily driving time periods in the observation days.
Accumulating the parking time period: and the sum of the daily parking time periods in the observation days.
Aging mileage: cumulative mileage/days of observation365。
Time length of year: cumulative duration/days of observation365。
Average daily parking duration: the parking duration/active days are accumulated.
Time sharing mileage: the sum of daily timeshare mileage over days of observation.
Driving safety factor:
emergency braking times per hundred kilometers: sudden braking times/accumulated mileage in current period 100。
Number of rapid acceleration per hundred kilometers: in the current period, the number of rapid acceleration/accumulated mileage100。
Number of sharp turns per hundred kilometers: in the current period, the number of sharp turns/accumulated mileage100。
Main driving route: and in the current period, the ratio of all the driving routes of the vehicle exceeds a certain threshold value. Typically the threshold is set to 5%. The vehicle transit times, vehicle travel speeds, and the like of these routes are counted simultaneously, the routes and the road types (e.g., expressways, urban roads, rural roads, etc.) traversed each time are identified, and compared with local traffic rules and accident data to evaluate the risk levels of the different routes and road types. This characteristic value can be used to set different risk factors for different routes and road types.
In order to consider the overall driving behavior and the recent trend of behavior change of the driver, the sampling range of the data used for extracting the driving intensity and the driving safety factor features is set to be multiple, and the sampling range includes all data, the last 180 days and the last 30 days.
And S13, performing risk assessment on driving behaviors corresponding to the geographic position data by clustering the characteristic data.
FIG. 4 shows a flow chart of risk assessment provided by an embodiment of the present application. Referring to fig. 4, in an embodiment of the present application, risk assessment includes the following operation steps:
s41, extracting core factors through clustering.
In order to further extract core factors when evaluating the driving behavior of new energy sources, a K-means clustering algorithm is used. K-means is a common method of unsupervised learning that can divide a set of data into clusters such that the distance between points within each cluster is minimized. In the clustering process, the data of all factors are combined into a matrix X, and the matrix X is subjected to standardization processing. Then, the normalized matrix X is input into a K-means algorithm to obtain a clustering result y. Finally, we choose the first k clusters as core factors to better characterize driving behavior.
Firstly, setting a driving behavior characteristic factor of a trolley to be provided with n sample points and k cluster centers, wherein the driving behavior characteristic vector of each sample point is as followsWhere p is the number of features. The distance of each sample point from the center of the cluster to which it belongs can be expressed as:
wherein, the liquid crystal display device comprises a liquid crystal display device,，/>，/>three characteristic values respectively representing the ith sample point, and +.>，/>，/>Three eigenvalues at the center of the j-th cluster are respectively represented. In the K-means algorithm we need to select a super parameter K to specify how many clusters the data is to be divided into. Then we randomly select k sample points as the initial cluster center and calculate the distance matrix D between them. Next, we repeat the following steps until convergence:
1. each sample point is assigned to the cluster at which the cluster center closest to it is located.
2. The position of the center of each cluster is updated to be the average of all sample points in the cluster.
3. The distance matrix D is recalculated and steps 1 and 2 are repeated until convergence.
Finally, the cluster center of each sample point is the optimal clustering result of the sample point.
S42, performing effective factor risk assessment according to the extracted core factors.
In actual operation, we find that 3 variables of time-sharing mileage, annual mileage and vehicle road complexity can be used as effective factors of driving behaviors for driving behavior evaluation.
3.2.1 Time-sharing mileage factor risk assessment
Traffic accident statistics annual reports issued by the public security department traffic authorities at regular intervals contain the number of traffic accidents in different time periods. Because the number of traffic accidents is related to the number of vehicles running on the road at the time, the number of traffic accidents can be divided by the number of active vehicles to obtain the accident density of the corresponding time period. Wherein the number of active vehicles may be calculated using the full amount of internet of vehicles data. The vehicle's mileage over a period of time indicates its risk exposure over that period of time. Combining the risk exposure intensities of all the time periods, the time-sharing mileage risk score of the vehicle can be obtained. The specific scoring process is as follows:
1. calculating the accident density of time-sharing period
Definition is respectively unified every 2 hours for a period of timeThe number of the counted traffic accidents is as follows:the number of active vehicles is:dividing the data of the same time period to obtain the time period accident density:wherein->。
2. Calculating a time-sharing mileage risk score
In the scoring stage, the time-sharing mileage characteristic of one trolley isThe time-sharing mileage risk score is:
3.2.2 Mileage factor Risk assessment
1. Mileage factor data normalization
When normalization of mileage factors is required, normalization methods based on 5% quantiles and 95% quantiles may be used. The basic idea of this approach is to divide the raw data into two parts: data less than or equal to 5% quantiles and data greater than or equal to 95% quantiles. Then, for the two parts of data, different normalization formulas are used for processing respectively.
Specifically, let us assume that we have a set of data x1, x2,..xn, where the minimum value is min_x and the maximum value is max_x. First, we need to calculate the 5% quantile p5 and the 95% quantile p95.
Next, we need to split the data into two parts: data less than or equal to p5 and data greater than or equal to p95. For these two parts of data, normalization processing is performed using the following formulas, respectively:
normalized_value = (original_value - min_x) / (max_x - min_x) scale_factor
for data less than or equal to p5, we can set scale_factor to a small value (e.g., 0.01) to ensure that the normalized value of these data is between 0 and 1. For data greater than or equal to p95, we can set scale_factor to a larger value (e.g., 100) to ensure that the normalized value of these data is between 0 and 1.
Finally, we can add the normalized values of the two parts of data to get the normalized value of the whole data. The advantage of this approach is that outliers can be handled efficiently and that most of the information in the original data can be preserved. The disadvantage is that the calculation of the values of p5 and p95 is required, which may increase the amount of calculation.
2. Mileage difference rate
For each trolley, its ranking in the whole is calculated (e.g. ordered by normalized mileage values). The vehicle data may then be divided into different clusters using a clustering algorithm.
For vehicles in each cluster, we can calculate the rate of difference between them. The difference rate refers to a measure of the distance or similarity between two vehicles. For example, we can use Euclidean distance or cosine similarity to calculate the difference rate. Wherein cosine similarity: cosine_similarity (x, y) =dot_product (x, y)/(norm (x)norm(y))。
Finally normalizing the difference rate: in order for the difference rate to be comparable, it needs to be normalized to between 0 and 1. This may be achieved by dividing the difference rate by the maximum difference rate between all vehicles in the cluster.
Assuming that there are n vehicles in the cluster, where the difference rate between the i-th vehicle and the other vehicles is di (i), the normalized maximum difference rate is max_diff=max (di (1), di (2), di (n)). Thus, the normalized difference rate is norm_diff (i) =di (i)/max_diff.
3. Mileage factor risk assessment
The normalized difference rate may be used as a scoring index. The higher the score, the more similar the vehicle; the lower the score, the more different the vehicle. An appropriate scoring threshold Sv may be selected as needed to determine which vehicles belong to the same category. Calculating the mileage risk score as Sm according to the deviation degree
3.2.3 Travel route road complexity risk assessment
First, the conventional road class is classified into: expressways, national roads, provincial roads, county roads, rural roads, county-rural internal roads, main streets, urban expressways, main roads, secondary roads, and common roads 11. The higher the road grade, the better the road condition and the lower the road complexity. The road grades of all tracks of the vehicle are acquired through a map service provider, so that mileage of the vehicle on different road grades can be counted.
Assume that the total mileage of a trolley isThen->。
Taking risk coefficients of different road grades asThe travel route road complexity risk is divided into Sr.
S43, carrying out overall risk assessment according to the extracted core factors.
For the time-sharing mileage risk score St, mileage factor risk score Sm, travel route road complexity risk score Sr that have been calculated, we use a hierarchical analysis method through group decision to obtain the weight distribution of these 3 evaluation indexes.
M decision makers evaluate the weights of the 3 evaluation indexes, wherein the evaluation vector of the importance of the ith decision maker on the 3 evaluation indexes is as follows:
by usingTo represent the proximity of the s decision maker and the t decision maker, and
thereby, the approach degree of the kth decision maker and the rest decision maker is obtained asAnd is also provided with
Further, the decision weight of the kth decision maker is obtainedAnd (2) and
finally, the weight distribution of the 3 evaluation indexes is obtained to be w, and
then S= (St, sm, sr) · (W1, W2, W3)
For a trolley, since the overall behavior characteristics of the vehicle which we count have all data, the last 180 days and the mostApproximately 30 days. So based on the above-mentioned risk assessment of the effective factors, we can obtain a risk score of the vehicle in 3 time dimensions。
And S14, displaying and feeding back the risk assessment result.
And visually displaying the scoring result, and designing a result feedback mechanism to provide personalized service for the user.
When our risk assessment for all vehicles is complete, we will get 3 sets of scores per trolley:for easier understanding and subsequent application, we first normalize the scores, here the z-score method is chosen. z-score is a commonly used normalization method that converts a set of data into a normal distribution with a mean of 0 and standard deviation of 1.
Referring to fig. 5, the following is a specific step of z-score normalization of the vehicle risk score:
s51, calculating the average value and standard deviation of each vehicle risk score. We have obtained a list of all vehicle risk scores including a risk score St, a mileage factor risk score Sm, and a travel route road complexity risk score Sr for each vehicle.
S52. For each vehicle risk score, the average is subtracted and then divided by the standard deviation. This results in a z-score value for the vehicle risk score. The specific formula is as follows:
where X represents the risk score for each vehicle, μ represents the average of all vehicle risk scores, σ represents the standard deviation of all vehicle risk scores.
Vehicles were then divided into A, B, C, D, E total 5 groups according to normalized score Z, with the highest a-score and lowest E-score.
For a long-term observed vehicle, the overall score and grouping thereof, the last 180 day score and grouping, the last 30 day score and grouping, can be obtained.
Thus, the user can conveniently see the overall style and behavior change of the driver.
Fig. 6 is a block diagram of a new energy automobile driving behavior analysis system provided by an embodiment of the present application. Referring to fig. 6, the new energy automobile driving behavior analysis system includes: an acquisition module 61, an extraction module 62, an evaluation module 63, and a presentation module 64.
The acquiring module 61 is configured to acquire geographic position data formed by driving of the vehicle, and clean the acquired geographic position data.
The extraction module 62 is configured to extract feature data in units of feature factors from the geographical location data through statistical analysis of the cleaned geographical location data.
The evaluation module 63 is configured to perform risk evaluation on driving behavior corresponding to the geographic location data by clustering the feature data.
The display module 64 is configured to display and feed back a result of the risk assessment.
In some embodiments, the acquisition module 61 comprises: and a repeated cleaning unit, an abnormal cleaning unit and a drifting cleaning unit.
The repeated cleaning unit is used for cleaning the repeated geographic position data.
The abnormality cleaning unit is used for cleaning the geographic position data for locating abnormality.
The drift cleaning unit is used for cleaning the geographic position data of the positioning drift.
In some implementations, the extraction module 62 includes: daily extraction unit, whole extraction unit.
The daily extraction unit is used for extracting daily behavior characteristics of the driving behavior of the vehicle.
The overall extraction unit is used for extracting overall behavior characteristics of the driving behavior of the vehicle.
In some embodiments, the daily behavioral characteristics include: daily average travel mileage, daily time-sharing mileage, daily average travel time, parking time, daily acceleration change rate, daily maximum vehicle speed, sudden braking frequency, sudden acceleration frequency, sudden turning frequency, daily travel route, and travel route and road type.
In some implementations, the overall behavioral characteristics include: driving intensity characteristics, driving safety factor;
the driving strength characteristics include: observation days, activity rate, accumulated mileage, accumulated time length, accumulated parking time length, annual history, annual time length, average daily parking time length and time-sharing mileage;
the driving safety factor includes: emergency braking times per hundred kilometers, emergency acceleration times per hundred kilometers, emergency turning times per hundred kilometers and main driving routes.
In some embodiments, evaluation module 63 includes: factor extraction unit, effective evaluation unit, and overall evaluation unit.
The factor extraction unit is used for extracting core factors through clustering.
The effective evaluation unit is used for performing effective factor risk evaluation according to the extracted core factors.
The overall risk assessment unit is used for carrying out overall risk assessment according to the extracted core factors.
In some embodiments, presentation module 64 includes: and (5) standardizing the unit.
The normalization unit is used for normalizing the result score of the risk assessment.
In some embodiments, the normalization unit is specifically configured to: calculating the average value and standard deviation of each vehicle risk score; for each vehicle risk score, the average value is subtracted and then divided by the standard deviation to obtain the z-score value of the risk score.
Fig. 7 is a schematic structural diagram of an electronic device provided by an embodiment of the present application, and shows a block diagram of an exemplary electronic device suitable for implementing an embodiment of the present application. The electronic device shown in fig. 7 is only an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present application. As shown in fig. 7, the electronic apparatus includes a processor 71, a memory 72, an input device 73, and an output device 74; the number of processors 71 in the electronic device may be one or more, in fig. 7, one processor 71 is taken as an example, and the processors 71, the memory 72, the input device 73 and the output device 74 in the electronic device may be connected by a bus or other means, in fig. 6, by a bus connection is taken as an example.
The memory 72 is used as a computer readable storage medium for storing software programs, computer executable programs and modules, such as program instructions/modules corresponding to a new energy automobile driving behavior analysis method in the embodiment of the present application. The processor 71 executes various functional applications of the computer device and data processing by running software programs, instructions and modules stored in the memory 72, i.e., implements a new energy automobile driving behavior analysis method as described above.
Note that the above is only a preferred embodiment of the present application and the technical principle applied. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, while the application has been described in connection with the above embodiments, the application is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the application, which is set forth in the following claims.
Claims (9)
1. The new energy automobile driving behavior analysis method is characterized by comprising the following steps of:
obtaining geographic position data formed by running of the vehicle, and cleaning the obtained geographic position data;
extracting feature data taking a feature factor as a unit from the geographical position data through statistical analysis of the cleaned geographical position data;
performing risk assessment on driving behaviors corresponding to the geographic position data in a clustering mode on the characteristic data; comprising the following steps:
extracting core factors through clustering;
according to the extracted core factors, performing effective factor risk assessment;
carrying out overall risk assessment according to the extracted core factors;
and displaying and feeding back the risk assessment result.
2. The method of claim 1, wherein the step of cleaning the obtained geographic location data comprises:
cleaning the repeated geographic position data;
cleaning the geographic position data with abnormal positioning;
and cleaning the geographic position data of the positioning drift.
3. The method of claim 1, wherein extracting feature data in units of feature factors from the geographical location data by statistical analysis of the cleaned geographical location data comprises:
extracting daily behavior characteristics of driving behaviors of the vehicle;
and extracting the overall behavior characteristics of the driving behavior of the vehicle.
4. A method according to claim 3, wherein the daily behavioral characteristics include: daily average travel mileage, daily time-sharing mileage, daily average travel time, parking time, daily acceleration change rate, daily maximum vehicle speed, sudden braking frequency, sudden acceleration frequency, sudden turning frequency, daily travel route, and travel route and road type.
5. A method according to claim 3, wherein the overall behavioral characteristics include: driving intensity characteristics, driving safety factor;
the driving strength characteristics include: observation days, activity rate, accumulated mileage, accumulated time length, accumulated parking time length, annual history, annual time length, average daily parking time length and time-sharing mileage;
the driving safety factor includes: emergency braking times per hundred kilometers, emergency acceleration times per hundred kilometers, emergency turning times per hundred kilometers and main driving routes.
6. The method of claim 1, wherein presenting and feeding back results of the risk assessment comprises:
and (5) normalizing the result score of the risk assessment.
7. The method of claim 6, wherein normalizing the outcome score of the risk assessment comprises:
calculating the average value and standard deviation of each vehicle risk score;
for each vehicle risk score, the average value is subtracted and then divided by the standard deviation to obtain the z-score value of the risk score.
8. A new energy automobile driving behavior analysis system, characterized by comprising:
the acquisition module is used for acquiring geographic position data formed by running of the vehicle and cleaning the acquired geographic position data;
the extraction module is used for extracting the characteristic data taking the characteristic factors as units from the geographical position data through statistical analysis of the cleaned geographical position data;
the evaluation module is used for performing risk evaluation on driving behaviors corresponding to the geographic position data in a clustering mode on the characteristic data;
the display module is used for displaying and feeding back the risk assessment result;
wherein the evaluation module comprises: factor extraction unit, effective evaluation unit, and overall evaluation unit:
the factor extraction unit is used for extracting core factors through clustering;
the effective evaluation unit is used for performing effective factor risk evaluation according to the extracted core factors;
the overall risk assessment unit is used for carrying out overall risk assessment according to the extracted core factors.
9. An electronic device, the electronic device comprising:
a memory for storing executable instructions;
a processor for implementing a new energy automobile driving behavior analysis method according to any one of claims 1 to 7 when executing the executable instructions stored in the memory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310954193.3A CN116665342B (en) | 2023-08-01 | 2023-08-01 | New energy automobile driving behavior analysis method, system and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310954193.3A CN116665342B (en) | 2023-08-01 | 2023-08-01 | New energy automobile driving behavior analysis method, system and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116665342A CN116665342A (en) | 2023-08-29 |
| CN116665342B true CN116665342B (en) | 2023-10-03 |
Family
ID=87717483
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310954193.3A Active CN116665342B (en) | 2023-08-01 | 2023-08-01 | New energy automobile driving behavior analysis method, system and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116665342B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20170115831A (en) * | 2016-04-08 | 2017-10-18 | 삼성에스디에스 주식회사 | Method and apparatus for guiding automobile insurance using driver recognizing |
| CN110866677A (en) * | 2019-10-25 | 2020-03-06 | 东南大学 | Driver relative risk evaluation method based on benchmark analysis |
| CN111461185A (en) * | 2020-03-19 | 2020-07-28 | 哈尔滨工程大学 | Driving behavior analysis method based on improved K-means |
| CN111739194A (en) * | 2020-06-22 | 2020-10-02 | 舒伟伟 | New energy automobile driving behavior analysis system and method |
| CN112766373A (en) * | 2021-01-19 | 2021-05-07 | 汉纳森（厦门）数据股份有限公司 | Driving behavior analysis method based on Internet of vehicles |
| CN114037015A (en) * | 2021-11-15 | 2022-02-11 | 重庆大学 | Driving behavior safety analysis method based on driving mode primitive |
| CN114627325A (en) * | 2020-12-14 | 2022-06-14 | 百度时代网络技术（北京）有限公司 | Scene classification method, device, equipment and storage medium |
| CN115871682A (en) * | 2023-01-06 | 2023-03-31 | 环球车享汽车租赁有限公司 | Method and system for monitoring, evaluating and early warning safe driving behavior of driver |
-
2023
- 2023-08-01 CN CN202310954193.3A patent/CN116665342B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20170115831A (en) * | 2016-04-08 | 2017-10-18 | 삼성에스디에스 주식회사 | Method and apparatus for guiding automobile insurance using driver recognizing |
| CN110866677A (en) * | 2019-10-25 | 2020-03-06 | 东南大学 | Driver relative risk evaluation method based on benchmark analysis |
| CN111461185A (en) * | 2020-03-19 | 2020-07-28 | 哈尔滨工程大学 | Driving behavior analysis method based on improved K-means |
| CN111739194A (en) * | 2020-06-22 | 2020-10-02 | 舒伟伟 | New energy automobile driving behavior analysis system and method |
| CN114627325A (en) * | 2020-12-14 | 2022-06-14 | 百度时代网络技术（北京）有限公司 | Scene classification method, device, equipment and storage medium |
| CN112766373A (en) * | 2021-01-19 | 2021-05-07 | 汉纳森（厦门）数据股份有限公司 | Driving behavior analysis method based on Internet of vehicles |
| CN114037015A (en) * | 2021-11-15 | 2022-02-11 | 重庆大学 | Driving behavior safety analysis method based on driving mode primitive |
| CN115871682A (en) * | 2023-01-06 | 2023-03-31 | 环球车享汽车租赁有限公司 | Method and system for monitoring, evaluating and early warning safe driving behavior of driver |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116665342A (en) | 2023-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200134735A1 (en) | Enhancement using analytics based on vehicle kinematic data | |
| KR101617349B1 (en) | Diagnostic system and method for the analysis of driving behavior | |
| Zhou et al. | Analysis of commercial truck drivers’ potentially dangerous driving behaviors based on 11-month digital tachograph data and multilevel modeling approach | |
| CN110866677B (en) | Driver relative risk evaluation method based on benchmark analysis | |
| US9082072B1 (en) | Method for applying usage based data | |
| US10262530B2 (en) | Determining customized safe speeds for vehicles | |
| CN107918826B (en) | Driver evaluation and scheduling method for driving environment perception | |
| Meng et al. | Expressway crash prediction based on traffic big data | |
| Gao et al. | What can we learn from telematics car driving data: A survey | |
| Pan et al. | Using OBD-II data to explore driving behavior model | |
| Meng et al. | Improving automobile insurance claims frequency prediction with telematics car driving data | |
| Stankevich et al. | Usage-based vehicle insurance: Driving style factors of accident probability and severity | |
| Abdel-Aty et al. | Assessing crash occurrence on urban freeways by applying a system of interrelated equations | |
| CN116665342B (en) | New energy automobile driving behavior analysis method, system and equipment | |
| CN116596307A (en) | Method for constructing driver security portrait model based on public transport operation security data | |
| CN114999181B (en) | Highway vehicle speed abnormality identification method based on ETC system data | |
| Lu et al. | A bi-level distribution mixture framework for unsupervised driving performance evaluation from naturalistic truck driving data | |
| CN116245365A (en) | Vehicle risk assessment method and device, storage medium and electronic equipment | |
| Zhao et al. | Individual driver crash risk classification based on iov data and offline consumer behavior data | |
| Dahl et al. | Understanding association between logged vehicle data and vehicle marketing parameters: Using clustering and rule-based machine learning | |
| Zhao et al. | Using traffic data to identify land-use characteristics based on ensemble learning approaches | |
| Korishchenko et al. | Usage-based vehicle insurance: Driving style factors of accident probability and severity | |
| Abeyratne et al. | Applying big data analytics on motor vehicle collision predictions in New York City | |
| CN112733891B (en) | Method for identifying bus IC card passengers to get off station points during travel chain breakage | |
| CN117315943B (en) | Monitoring analysis and early warning method and system for overrun transportation violations |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |