TWI660276B - System and method for applying user profile model to score - Google Patents

Abstract

A system applying a user profile model for scoring includes a storage device and a processing device. The storage device is used to store environmental risk historical data and user profile historical data. The processing device includes an environmental data collection module for collecting environmental risk factors corresponding to the environment, a factor data collection module for collecting user contour factors corresponding to the user, and a method for determining the relationship between environmental risk factors and historical environmental risk data. Environmental risk analysis module for calculating environmental risk scores, user contour analysis module for calculating difference scores based on differences between user contour factors and user contour history data, and environmental risk analysis The score and difference score are used to calculate an evaluation score calculation module.

Description

System for applying user profile models for scoring And methods

The present disclosure relates to a system and method, and more particularly to a system and method for applying a user profile model for scoring.

Many of them are used to evaluate users by establishing a general model, but for situations that often vary from person to person, the general model is often not applicable. Especially in the evaluation of driving fatigue, which often differs in personal behavior, the existing products that detect driving fatigue on the market are mostly judged by images and physiological information. Most of them are experimental, and their practicality and universality are still insufficient. ; Existing products that use on-board diagnostics (OBD) data to analyze driving behaviors are still mostly reviewed for abnormal driving behaviors after the end of the trip, without real-time warnings, leading to immediate driving risks. Failure to improve effectively.

An embodiment of the present disclosure discloses a system for scoring a user profile model to calculate a user's assessment score on the environment. The system includes an environmental data collection module, a factor data collection module, and a storage device. And processing equipment. The environmental data collection module is used to detect environmental risk factors. The factor data collection module is used to detect user contour factors. The storage device is used to store environmental risk historical data and user profile historical data. The processing device is configured to calculate an environmental risk score according to the difference between the environmental risk factor and the historical environmental risk data, calculate a difference score according to the difference between the user contour factor and the user contour historical data, and combine the environmental risk score with the The difference score calculates an evaluation score.

An embodiment of this disclosure document discloses a method for scoring by applying a user profile model, which is implemented by a local device to calculate a user's evaluation score on the environment. The method includes the following steps. The environmental risk score is calculated based on the differences between historical environmental risk data; the processing device calculates a difference score based on the difference between the user profile factor and the user profile historical data; and the processing device calculates based on the environmental risk score and the difference score Evaluate the score.

100‧‧‧ System for applying user contour model for scoring

121‧‧‧Environmental Data Collection Module

122‧‧‧factor data collection module

110‧‧‧Storage device

120‧‧‧Processing device

123‧‧‧Environmental Risk Analysis Module

124‧‧‧User Profile Analysis Module

125‧‧‧Evaluation score calculation module

130‧‧‧communication circuit

OBD‧‧‧Onboard diagnostic device

126‧‧‧Driving Behavior Score Calculation Module

150‧‧‧ warning device

200‧‧‧ Method for applying user contour model for scoring

D10‧‧‧ Historical Environmental Risk Data

D20‧‧‧Driving contour history data

DB1‧‧‧ Weather Database

DB2‧‧‧Traffic Database

F10‧‧‧Environmental risk factor

F20‧‧‧Driving Contour Factor

IN‧‧‧Internet

SC10‧‧‧Environmental Risk Score

SC20‧‧‧ Difference Score

SC30‧‧‧ evaluation score

SC40‧‧‧Current driving behavior score

S110 ~ S130‧‧‧step

In order to make the above and other objects, features, advantages, and embodiments of the present disclosure more comprehensible, the description of the drawings is as follows: FIG. 1A is an outline of an application user according to an embodiment of the present disclosure Functional block diagram of a model to score the system.

FIG. 1B is a functional block diagram of a system for applying a user profile model for scoring according to an embodiment of the present disclosure to collect environmental risk factors and driving profile factors through a communication circuit.

FIG. 2 is a flowchart of a method for applying a user profile model for scoring according to an embodiment of the present disclosure.

The following is a detailed description with examples and the attached drawings to better understand the aspect of the case, but the examples provided are not intended to limit the scope covered by the case, and the description of the structural operation is not used to limit it In the order of execution, any structure that reassembles the components and produces a device with equal efficacy is within the scope of this case.

Please refer to Fig. 1A, Fig. 1B, and Fig. 2. FIG. 1A is a functional block diagram of a system 100 for applying a user profile model for scoring according to an embodiment of the present disclosure. FIG. 1B is a functional block diagram of a system 100 for applying a user profile model for scoring according to an embodiment of the present disclosure to collect an environmental risk factor F10 and a driving profile factor F20 through a communication circuit 130. FIG. 2 is a flowchart of a method 200 for applying a user profile model for scoring according to an embodiment of the present disclosure. FIG. 2 is applicable to the system 100 of FIGS. 1A and 1B for implementation. In this embodiment, an example is applied to the score of driving fatigue, but it is not limited thereto.

The system 100 is used to calculate a user's environmental assessment score SC30. In this embodiment, the user may be driving, the environment may be the environment of the road section where the vehicle is driven, and the evaluation score SC30 is used to evaluate the degree of fatigue of driving. That is, the system 100 can be used to calculate the degree of fatigue of the environment driving on the road section of the car.

A system 100 for applying a user profile model for scoring includes a storage device 110 and a processing device 120, wherein the processing device 120 includes an environmental data collection module 121, a factor data collection module 122, an environmental risk analysis module 123, and a user profile analysis Module 124 and evaluation score calculation module 125. In the embodiment of FIG. 1B, the system 100 further includes a communication circuit 130.

In one embodiment, the storage device 110 may be a hard disk drive (HDD), a solid state disk (SSD), or a redundant array of independent disks (RAID). The storage device 110 is not limited to this, and any device that can be used to store data belongs to this creative category.

In one embodiment, the processing device 120 may include integrated circuits such as a micro controller, a microprocessor, a digital signal processor, and an application specific integrated circuit. ASIC) or logic circuits, respectively, or they can be implemented by commercially available processors, computers, servers, or electronic devices with computing capabilities.

The environmental data collection module 121 is used to correspond to the environmental detection environmental risk factor F10. Specifically, the environmental data collection module 121 can be connected to the Internet IN through the communication circuit 130 to communicate with the weather condition database DB1 (such as the Meteorological Bureau) and the road condition database DB2 (such as the Highway Bureau), and through the communication circuit 130 communicates with the on-board diagnostic device OBD, and retrieves the environmental risk factor F10 corresponding to the driving position to the weather condition database DB1, the road condition database DB2, and the on-board diagnostic device OBD. Environmental risk factor F10 may include weather conditions Factors, road condition factors, and vehicle condition factors, where weather condition factors can be rainfall data, wind speed data, and UV index data; road condition factors can be current location data, current vehicle speed data, and special event data, among which special events such as road construction Incident or car accident; vehicle condition factors can be data such as vehicle age data, date data, model data, and tire pressure data.

In other words, the environmental data collection module 121 can obtain the driving position through the positioning data of the positioning device, and connect to the external weather condition database DB1, the road condition database DB2, and the on-board diagnostic device OBD according to the driving position. Retrieve the risk factor corresponding to the driving position.

The factor data collection module 122 is used to correspond to the driving detection driving contour factor F20. Specifically, the factor data collection module 122 can be connected to the storage device 110, and can be communicatively connected to the on-board diagnostic device OBD through the communication circuit 130, and retrieve the driving contour factor F20 corresponding to the driving behavior to the on-board diagnostic device OBD. The driving contour factor F20 may include a physical condition factor, a driving condition factor, and a driving record factor. The physical condition factor may be data such as gender data, age data, and physical condition data. The driving status factors can be data such as rapid acceleration / deceleration change data, rapid acceleration / deceleration frequency data, sharp turning data, shift data, idle time data, and idle frequency data, among which the shift data includes engine speed, vehicle speed, and correct gear. Position and clutch and throttle interval. The driving record factor can be data such as driving intensive degree data, driving hours data, and current driving mode data.

In other words, the factor data collection module 122 can obtain the driving contour factor F20 corresponding to the driving behavior through the on-board diagnostic device.

The storage device 110 is used to store the environmental risk history of the corresponding environment Data D10 and driving contour history data D20 corresponding to driving.

The environmental risk historical data D10 may include weather condition historical data, road condition historical data, and vehicle condition historical data. The historical data of weather conditions can be rainfall reference data, wind speed reference data, and UV index reference data; road condition historical data can be historical accident location data and average vehicle speed data; historical condition data can be vehicle age reference data and scheduled inspection dates. Data, vehicle reference and tire pressure reference.

The driving contour history data D20 may include historical data of physical conditions, historical data of driving conditions, and historical data of driving records. The historical data of physical conditions may be gender reference data, age reference data, and physical condition reference data. The driving current historical data can be rapid acceleration / deceleration change amount reference, rapid acceleration / deceleration frequency reference, historical sharp turn reference, shift reference, idle time reference and idle frequency reference. The historical data of driving records can be reference information of driving intensity, driving hours reference and driving mode reference.

It should be noted that the aforementioned environmental risk factor F10, driving contour factor F20, environmental risk historical data D10, and driving contour historical data D20 are just examples, and are not limited thereto.

The environmental risk analysis module 123 of the processing device 120 can calculate the environmental risk score SC10 according to the difference between the environmental risk factor F10 and the environmental risk historical data D10 according to step S110. The user profile analysis module 124 of the processing device 120 may calculate a difference score SC20 according to the difference between the driving profile factor F20 and the driving profile history data D20 according to step S120. The evaluation score calculation module 125 of the processing device 120 may perform environmental risk assessment according to step S130. The score SC10 and the difference score are used to calculate an evaluation score SC30.

In one embodiment, the environmental risk analysis module 123 of the processing device 120 may perform a KL (Kullback-Leibler) divergence calculation based on the probability distribution represented by the current environmental risk factor F10 and historical risk data D10 to calculate the environmental risk. Score SC10. The user profile analysis module 124 of the processing device 120 performs a KL divergence calculation based on the current user profile factor F20 and the probability distribution represented by the user profile history data D20 to calculate a difference score SC20. The processing device 120 adds the environmental risk score SC10 multiplied by a first weight and the difference score SC20 multiplied by a second weight to calculate the evaluation score SC30.

The KL divergence calculation is used to evaluate the asymmetry of the two probability distributions. When the asymmetry is larger, the value of the KL divergence is larger. In this embodiment, when the environmental risk factor F10 and the historical environmental risk data The larger the D10 difference, the higher the KL divergence will be, and the calculated environmental risk score SC10 will be increased accordingly. Similarly, when the difference between the environmental risk factor F10 and the historical environmental risk data D10 is smaller, the obtained KL divergence is correspondingly reduced, so that the calculated environmental risk score SC10 is correspondingly reduced. Similarly, the calculation of the user contour factor F20 and the user contour history data D20 and the difference score SC20 have similar changes.

Further, the environmental risk factor F10 includes a weather condition factor, a road condition factor, and a vehicle condition factor. The environmental risk analysis module 123 can calculate a weather condition score according to a difference between the weather condition factor and historical weather condition data, and according to the road condition factor and the road condition history. The difference between the data is used to calculate the road condition score, and the vehicle condition score is calculated based on the difference between the vehicle condition factor and the historical condition data. ring The environmental risk analysis module 123 then calculates the environmental risk score SC10 based on the weather condition score, road condition score, and vehicle condition score.

The driving contour factor F20 includes a physical condition factor, a driving condition factor, and a driving record factor. The user contour analysis module 124 can calculate a physical condition score based on a difference between the physical condition factor and historical condition data, and according to the driving condition factor and history The difference between the driving status data is used to calculate the driving status score, and the driving record score is calculated based on the difference between the driving record factor and the historical driving record data. The user profile analysis module 124 then calculates the difference score SC20 according to the physical condition score, the driving status score, and the driving record score.

In the following, the driving of a road section as an example will be used to explain the calculation of the driving environmental risk score SC10. For the convenience of explanation, the weather condition factor only uses the rainfall data as an example, the weather condition historical data only uses the rainfall reference data as an example; the road condition factor only uses the current position data as an example, and the road condition historical data only uses the historical accident position data as an example; Take the vehicle age data as an example, and the vehicle condition historical data is only based on the vehicle age reference data as an example.

First, the environmental data collection module 121 can be connected to the storage device 110, and can be connected to the weather condition database DB1, the on-board diagnostic device OBD (including positioning data) and the storage device 110 through the communication circuit 130, and respectively to the weather condition database DB1 retrieves the rainfall data of this section, and retrieves the rainfall reference data to the storage device 110. The rainfall reference data can be the data about the rainfall and weather conditions scores. The rainfall reference data defines the rainfall of 1 hour and 40 mm as a benchmark. Line, and the weather condition score corresponding to rainfall of 1 hour and 40 mm is defined as 60 points.

In this example, assuming that the rainfall in this section of rainfall data is 1 hour and 50 mm, the environmental risk analysis module 123 can perform the difference between the rainfall of 1 hour and 50 mm and the rainfall of 1 hour and 40 mm in the rainfall data. For comparison, for example, by calculating the KL divergence, the rainfall data in this section is greater than the baseline, and the difference is 1 hour and 10 millimeters. Based on the higher rainfall, the score is higher, so the weather condition score is 65 points. This is because the danger level is higher than that of a section with less rainfall.

The environmental data collection module 121 is communicatively connected to the positioning device, the road condition database, and the storage device 110, and respectively retrieves the current position data from the road condition database, and the historical accident position data from the storage device 110. The historical accident position data may be For the historical accident location and road condition score data, the historical accident location data defines a circle with the historical accident location as the center and a distance of five kilometers from the center as a radius, and defines a road condition score corresponding to a distance of five kilometers as 60 Minute.

In this example, the environmental risk analysis module 123 can compare the difference between the current position data and the historical accident position data, for example, by calculating the KL divergence, assuming that the current position of the driver is one kilometer away from the historical accident position. The current position of the driver is already within the range of the historical accident position. Based on the closer the historical accident position, the higher the score, the road condition score can be calculated as 70 points, because the closer to the historical accident position, the higher the probability of an accident.

The environmental data collection module 121 is communicatively connected to the vehicle condition database and the storage device 110, and respectively retrieves the vehicle age data from the vehicle condition database and the storage device 110 to retrieve the vehicle age reference data, where the vehicle age reference data can be about the vehicle Age and vehicle condition score data, the vehicle age reference material defines a ten-year vehicle age as A baseline, and the vehicle condition score corresponding to ten years of age is defined as 60 points.

In this example, assuming the age of the car being driven is 20 years, the environmental risk analysis module 123 can compare the 20-year car age in the car age data with the ten-year car defined by the car age reference data. Age comparison, for example, calculated by the KL divergence, because the vehicle age has exceeded the baseline and differed by ten years, the more the vehicle age is different from the baseline, the higher the score, so the vehicle condition score is calculated as 80 points because the vehicle age A larger number indicates a higher probability of an accident.

After the environmental risk analysis module 123 calculates the weather condition score, road condition score, and vehicle condition score, the environmental risk analysis module 123 multiplies the weather condition score, road condition score, and vehicle condition score by weights and then adds up to obtain the environment. Risk score SC10.

For example, the weights corresponding to the weather condition score, road condition score, and vehicle condition score can be set to 0.25, 0.5, and 0.25 respectively, so the following formula can be obtained: Environmental risk score SC10 = 0.25 * (weather condition score) + 0.5 * (road condition Score) + 0.25 * (vehicle condition score). Further, the weather condition score, road condition score, and vehicle condition score are substituted into the above formula by 60 points, 65 points, and 70 points, respectively. Therefore, the environmental risk analysis module 123 can calculate the environmental risk score SC10 to be 65 points.

Next, take driving on this road section as an example to illustrate the calculation of the driving difference score. For the convenience of explanation, the physical condition factor uses gender data and age data as examples, and the physical condition historical data uses gender reference data and age reference data as examples; the driving status factor only uses rapid acceleration / deceleration change amount data as an example, driving status historical data Taking only rapid acceleration / deceleration change amount reference data as an example, first, the factor data collection module 122 is communicatively connected to the on-board diagnostic device and the storage device, and separately retrieves the driving gender data and age data from the on-board diagnostic device. And retrieve the gender reference and the age reference from the storage device, wherein the gender reference and the age reference can be data on the gender and the age and physical condition score, the gender reference defines the male and female, and the age reference defines the age Interval values, and the probability of driving risk behavior corresponding to gender and age is defined based on reference data.

In this example, the user profile analysis module 124 can compare the differences between the gender and age data and the gender and age reference data. For example, by calculating the KL divergence, it is assumed that the driving gender is male and the age is 31- At the age of 40, based on past references, a driving risk occurrence score can be obtained under the same conditions, and the physical condition score corresponding to this driving risk occurrence score is 50 points, and the lower the score, the lower the risk of driving risk.

In another embodiment, the physical condition score is also adjusted according to a special physical condition (such as a cold or fever). The adjustment method of the score, for example, may increase 5 points for a cold or fever. For example, if the physical condition score is 50 points, when driving a cold or fever, the physical condition score will increase by 5 points, so the adjusted physical condition score is 55 points. It should be noted that the calculation method of the adjusted score is only an example, but it is not limited thereto.

The factor data collection module 122 is communicatively connected to the positioning device, the on-board diagnostic device, and the storage device, and separately acquires the rapid acceleration / deceleration change data of the car driven by the on-board diagnostic device, and retrieves the emergency data from the storage device. Acceleration / deceleration change amount reference material, among which the rapid acceleration / deceleration change amount reference material can be information about the rapid acceleration / deceleration change amount and the driving status score, and the rapid acceleration / deceleration change amount reference material defines 20 (kph / s) The acceleration / deceleration change amount is used as a reference line, and driving corresponding to the acceleration / deceleration change amount of 20 (kph / s) The status score is defined as 60 points.

In this example, assuming that the acceleration / deceleration change of the car being driven is 30 (kph / s), the user profile analysis module 124 can compare the rapid acceleration / deceleration change with the rapid acceleration / deceleration change reference data. Compare the differences between them. For example, calculated by the KL divergence, because the rapid acceleration / deceleration change is larger than the baseline, and the difference is 10 (kph / s), the larger the rapid acceleration / deceleration change based on the same section, the higher the score. Therefore, the driving current score is calculated to be 70 points, because the representative may experience unusual acceleration / deceleration changes due to fatigue when driving on this road section.

In another embodiment, the calculation method of the running status score can also be calculated by considering the rapid acceleration / deceleration frequency change data, sharp turn data, and shift data at the same time. For example, when the acceleration / deceleration change is more than 12 (kph / s), it is 1 point, and the others are 0.2 points; when the sharp turn is more than 10 ° within one second, the azimuth angle is changed to 1 point. ; When the shift data is compared with the shift history data (which can be referred to as shift accuracy), such as comparing engine speed, vehicle speed, correct gear, clutch and throttle interval time, when any item deviates from the normal range, The score for this item is 1. Therefore, if the amount of rapid acceleration / deceleration change> 12 (kph / s), the number of sharp turns, and the number of shifts outside the normal range occur too much in a certain period, such as the amount of rapid acceleration / deceleration change> The number of 12 (kph / s) five times increases the score by 5 points, the number of sharp turns three times increases the score by 3 points, and the number of shifts out of the normal range twice increases the score by 2 . The driving status score will also increase accordingly, because the representative may experience unusual acceleration / deceleration changes, the number of sharp turns, and gear shifts due to fatigue when driving on this road. It should be noted that the above-mentioned increased scores can be directly added, multiplied by weights and then added or calculated in other ways.

The factor data collection module 122 is communicatively connected to the on-board diagnostic device and the storage device, and respectively retrieves the driving intensity data of the driving vehicle from the on-board diagnostic device, and retrieves the driving density reference data to the storage device, among which The driving intensity reference can be information about driving intensity and driving record scores. The driving intensity reference defines the number of trips on a Friday as a baseline, and the corresponding number of trips on a Friday The driving record score is defined as 65 points.

In this example, assuming that the number of driving trips is seven times a week, the user profile analysis module 124 can compare the difference between the driving intensity and the driving intensity reference, for example, by calculating the KL divergence. The number of trips this time is larger than the baseline, and the difference is twice a week. Based on the higher the number of trips, the higher the score, the higher the record score is 80 points, because the higher the number of trips driving indicates the rest time of driving. Being compressed while having doubts about fatigue driving increases the risk of accidents.

In another embodiment, the driving record score will also be adjusted according to the continuous driving hours of the driving. The adjustment method of the score can be, for example, an increase of 0.1 points every four hours. For example, if the driving record score is 80 points, when the average continuous driving time is six hours, the driving Himeji score will increase by 0.2 points, so the adjusted driving record score is 80.2 points. It should be noted that the calculation method of the adjusted score is only an example, but it is not limited thereto.

After the user profile analysis module 124 calculates the physical condition score, driving status score, and driving record score, the user profile analysis module 124 multiplies the physical condition score, driving status score, and driving record score by weights and adds up To get the difference score.

For example, the weights corresponding to the physical condition score, driving condition score, and driving record score can be set to 0.2, 0.7, and 0.1 respectively, so the following formula can be obtained: difference score = 0.2 * (physical condition score) + 0.7 * (driving Status score) + 0.1 * (driving record score). Further, the physical condition score, driving status score, and driving record score are substituted into the above formula by 50 points, 70 points, and 80 points, respectively, so the user profile analysis module 124 can calculate a difference score of 67 points.

The evaluation score calculation module 125 is configured to calculate an evaluation score SC30 from the environmental risk score SC10 and the difference score.

In detail, after the environmental risk score SC10 and the difference score SC20 are calculated by the environmental risk analysis module 123 and the user profile analysis module 124, respectively, the evaluation score calculation module 125 further separates the environmental risk score SC10 and the difference score SC20, respectively. Multiply the first weight and the second weight and add up to get the evaluation score SC30.

For example, the first and second weights corresponding to the environmental risk score SC10 and the difference score SC20 can be set to 0.3 and 0.7, respectively, so the following formula can be obtained: Evaluation score SC30 = 0.3 * (Environmental risk score SC10) + 0.7 * (different score SC20). Next, the environmental risk score SC10 and the difference score SC20 were respectively substituted into the above formula with 65 points and 67 points, so the evaluation score SC30 was 66.4 points.

In addition, the evaluation score calculation module 125 is further configured to classify the calculated evaluation score SC30 according to the fatigue level classification data, so as to determine the fatigue level corresponding to the evaluation score SC30.

Fatigue degree classification data can be data about the assessment score SC30 and fatigue degree. The fatigue degree classification data defines four levels of fatigue degree, namely non-fatigue degree, mild fatigue degree, moderate fatigue degree, and severe fatigue degree. The assessment score SC30 for non-fatigue is below 70, the assessment score SC30 for mild fatigue is between 70 and 85, and the assessment score SC30 for moderate fatigue is between 86 and 97, corresponding The severe fatigue assessment score SC30 is between 98 and 100.

Therefore, the evaluation score calculation module 125 of the processing device 120 can classify the evaluation score SC30 of 66.4 points into a non-fatigue degree according to the fatigue degree classification data. In other examples, if the evaluation score SC30 is 75 points, the evaluation score calculation module 125 can classify the 75-point evaluation score SC30 into a mild fatigue level based on the fatigue classification data; if the evaluation score SC30 is 90 points, then The evaluation score calculation module 125 can classify the 90-point evaluation score SC30 into moderate fatigue according to the fatigue level classification data; if the evaluation is 99 points, the evaluation score calculation module 125 can classify 99 points based on the fatigue level classification data The SC30 score is classified as severe fatigue.

It should be noted that the foregoing definitions and calculation methods of various scores are merely examples, and are not limited thereto.

In one embodiment, the system 100 further includes a warning device 150 for determining that when the evaluation score SC30 is greater than a threshold value, a warning is issued. For example, suppose the threshold is based on the non-fatigue evaluation score SC30 as the baseline, that is, the threshold is set to 70 points. If the evaluation score SC30 is greater than 70 points, it indicates that the degree of fatigue of driving is a light fatigue. The warning device 150 will issue a warning to inform the driver to avoid a traffic accident.

Thereby, the system 100 can judge the driving condition of the driver in real time according to the evaluation score to give corresponding instructions. For example, if the assessment score SC10 is a non-fatigue degree, no warning is given; if the assessment score SC10 is a severe fatigue degree, then Issue a warning.

In an embodiment, the processing device 120 of the system 100 further includes a driving behavior score calculation module 126 for calculating the current driving behavior score SC40 according to the evaluation score SC30 and the historical driving behavior score, wherein the historical driving behavior score refers to driving past SC30 is a statistical score of the evaluation score of each trip. After the current driving behavior score SC40 is calculated, the current driving behavior score SC40 will be recorded and replace the old historical driving behavior score to become the new historical driving behavior score.

In addition, the method of counting the scores will be determined corresponding to the new assessment score SC30. For example, if the evaluation score SC30 of this trip is less than 70 (non-fatigue), the driving behavior score calculation module 126 averages the evaluation score SC30 of this trip and the historical driving behavior score to calculate the current driving behavior score. SC40, for example, the evaluation score SC30 of this trip is 68 points, and the historical driving behavior score is 70 points, so the current driving behavior score SC40 is the average of two points, that is, 69 points; if the evaluation score SC30 of this trip is 70 to 97 Between points (mild fatigue or moderate fatigue), the driving behavior score calculation module 126 averages the evaluation score SC30 of this trip with the historical driving behavior score, and then adds 5 points to calculate the current driving behavior. Score SC40, for example, the evaluation score SC30 of this trip is 90 points, and the historical driving behavior score is 69 points, so the current driving behavior score SC40 is the average of the two scores plus 5 points, which is 79.5 points; if the evaluation of this trip The score SC30 is between 98 and 100. The driving behavior score calculation module 126 adds the average of the evaluation score SC30 and the historical driving behavior score to 10 points. For example, the evaluation score SC30 of this trip is 99 points, the historical driving behavior score is 79.5, so the current driving behavior score SC40 is the average of the two scores plus 10 points, which is 99.25 Minute.

In this way, the system 100 can judge the driving situation of the driver according to the current driving behavior score SC40 to give a corresponding instruction. For example, if the current driving behavior score SC40 belongs to a non-fatigue level, an instruction is given to the driver to continue driving; if the current driving behavior score SC40 belongs to a severe fatigue level, the driving is instructed not to continue driving.

Further, the driving behavior score calculation module 126 is further configured to establish a driving behavior model to calculate the current driving behavior score SC40 according to the driving behavior model. The driving behavior model is based on a neural network, machine learning, deep learning, or matrix decomposition. set up.

In addition, the evaluation score SC30 calculated by the evaluation score calculation module 125 can be further stored in the storage device 110 for reference.

In summary, the system and method for applying a user profile model for scoring in this disclosure document includes an environmental data collection module, a factor data collection module, a storage device, an environmental risk analysis module, and a user profile analysis module. And evaluation score calculation module to judge the driving condition of the driver based on long-term driving behavior, and also to judge the driving condition of the driver in real time and give corresponding instructions.

Although this case has been disclosed as above with examples, it is not intended to limit this case. Any person with ordinary knowledge in the technical field can make some modifications and retouching without departing from the spirit and scope of this case. Therefore, the protection of this case The scope shall be determined by the scope of the attached patent application.

Claims (16)

A system for applying a user profile model for scoring, which is used to calculate an evaluation score of a user in an environment, includes: a storage device for storing historical environmental risk data corresponding to the environment and corresponding user's A user profile history data; and a processing device including: an environmental data collection module to collect an environmental risk factor corresponding to the environment; a factor data collection module to collect a use corresponding to the user Profile factor; an environmental risk analysis module to calculate an environmental risk score based on the difference between the environmental risk factor and the historical environmental risk data; a user profile analysis module to use the user profile A difference between the factor and the user profile history data to calculate a difference score; and an evaluation score calculation module for calculating the environmental risk score and the difference score to calculate the evaluation score. The system according to claim 1, wherein the processing device calculates the environmental risk score according to the divergence calculation based on the current environmental risk factor and the probability distribution represented by the historical environmental risk data; the processing device is based on the current The user contour factor and the probability distribution represented by the user contour history data are used to calculate a divergence to calculate the difference score; the processing device multiplies the environmental risk score by a first weight and the second risk by a second weight. The difference score is added to calculate the evaluation score. The system according to claim 1, wherein the user is a driver, the user contour history data is a driving contour history data, the user contour factor is a driving contour factor, and the evaluation score is used To assess the fatigue level of the driving. The system according to claim 3, wherein the environmental risk factor includes at least one of a weather condition factor, a road condition factor, and a vehicle condition factor, and the environmental risk analysis module is configured to use the weather condition factor, the road condition factor, and The difference between the at least one of the vehicle condition factors and the historical environmental risk data respectively calculates at least one of a weather condition score, a road condition score, and a vehicle condition score. The environmental risk analysis module is used to determine the weather condition. The environmental risk score is calculated from at least one of the score, the road condition score, and the vehicle condition score. The system according to claim 3, wherein the driving contour factor includes at least one of a physical condition factor, a driving condition factor, and a line-of-vehicle record factor, and the user contour analysis module is configured to At least one of the driving condition factor and the driving record factor, and the difference between the driving profile history data and the driving contour history data are respectively calculated to calculate at least one of a physical condition score, a driving condition score, and a driving record score, and the user contour analysis The module is used to calculate the difference score by at least one of the physical condition score, the driving status score and the driving record score. The system described in claim 3 further includes a warning device for determining that when the evaluation score is greater than a threshold value, a warning is issued. The system according to claim 3, wherein the processing device further includes a driving behavior score calculation module for calculating a current driving behavior score based on the evaluation score and a historical driving behavior score, and recording the current driving behavior score. The system according to claim 7, wherein the driving behavior score calculation module is used to establish a driving behavior model, and the driving behavior score calculation module calculates the current driving behavior score according to the driving behavior model, wherein the driving behavior model Built on a type of neural network, a machine learning, a deep learning, or a matrix factorization. A method for applying a user profile model for scoring is implemented by a processing device to calculate an evaluation score of a user in an environment, including the following steps: the processing device is based on an environmental risk factor and an environmental risk history An environmental risk score is calculated based on the difference between the data; the processing device calculates a difference score based on the difference between a user profile factor and a user profile historical data; and the processing device is based on the environmental risk score and The difference score calculates the evaluation score. The method according to claim 9, wherein the processing device calculates the environmental risk score according to the divergence calculation based on the current environmental risk factor and the probability distribution represented by the historical environmental risk data; the processing device calculates the environmental risk score according to the current The user contour factor and the probability distribution represented by the user contour history data are used to calculate a divergence to calculate the difference score; the processing device multiplies the environmental risk score by a first weight and the second risk by a second weight. The difference score is added to calculate the evaluation score. The method according to claim 9, wherein the user is a driver, the user contour history data is a driving contour history data, the user contour factor is at least one of a driving contour factor, and the The evaluation score is used to evaluate the fatigue level of the driving. The method according to claim 11, wherein the environmental risk factor includes at least one of a weather condition factor, a road condition factor, and a vehicle condition factor, and the step of calculating the environmental risk score is based on the weather condition factor and the road condition factor And at least one of the vehicle condition factors, and the difference between the historical environmental risk data and at least one of a weather condition score, a road condition score, and a vehicle condition score, respectively, the environmental score is based on the weather condition score, the road condition At least one of the score and the vehicle condition score is calculated. The method according to claim 11, wherein the driving contour factor includes at least one of a physical condition factor, a driving condition factor, and a line of vehicle record factor, and the step of calculating the difference score is based on the physical condition factor, the driving At least one of the current condition factor and the driving record factor, and the driving profile history data are respectively calculated to calculate at least one of a physical condition score, a driving condition score, and a driving record score, and the difference score is based on the physical At least one of the status score, the current driving status score, and the driving record score is calculated. The method according to claim 11, further comprising the step of determining that when the evaluation score is greater than a threshold value, a warning is issued. The method according to claim 11, further comprising the steps of: calculating a current driving behavior score according to the evaluation score and a past driving behavior score, and recording the current driving behavior score. The method according to claim 15, wherein the driving behavior score is calculated according to a driving behavior model, wherein the driving behavior model is established according to a type of neural network, a machine learning, a deep learning, or a matrix decomposition .