TWI727485B - Smart scheduling analysis system with instantly collecting driving behavior data and physiological information - Google Patents

Abstract

The invention discloses a smart scheduling analysis system with instantly collecting driving behavior data and physiological information, which comprises at least one of a mobile device sensor module and a wearable device sensor module, a machine learning automatic analysis module and a smart scheduling module. The mobile device sensor module or the wearable device sensor module instantly collects driving behavior data and physiological information of a driver in a transport carrier. The machine learning automatic analysis module performs machine learning on both the driving behavior data and the physiological information of the driver to automatically analyze correlation result or the associated model of the both. The smart scheduling module performs smart scheduling based on the association result or associated model of the driving behavior data and physiological information of the driver to generate a corresponding schedule associated with the driver.

Description

Intelligent scheduling analysis system for real-time collection of driving behavior data and physiological information

The present invention relates to a smart scheduling analysis technology, in particular to a smart scheduling analysis system for real-time collection of driving behavior data and physiological information.

Passengers, tourist buses, buses, taxis, Taiwan Railways, high-speed rail, MRT and other transportation vehicles are the main means of transportation in the transportation system that passengers take, and the schedules of transportation vehicles and their drivers are mostly compiled by humans. Adopt a fixed schedule method. However, drivers may have poor physical conditions or conditions that affect driving safety, which may affect the driving safety of transportation vehicles and passengers.

In a prior art, an automatic vehicle scheduling device is proposed. The scheduling module first retrieves the vehicle, personnel, and route data registered in the data module for scheduling, and the scheduling module confirms the scheduling data to dispatch the vehicle. Then, the receiving module receives the real-time operating status and information of the vehicle, and the operating module monitors each vehicle to track the driver and vehicle situation. However, this prior art cannot collect the driving behavior data and physiological information of the driver in real time, so that it is impossible to perform intelligent scheduling analysis based on the driving behavior data and physiological information of the driver, thereby affecting the driving safety of transportation vehicles and passengers.

Therefore, how to provide a novel or innovative intelligent scheduling analysis technology for real-time collection of driving behavior data and physiological information has become a major research topic for those skilled in the art.

The present invention provides a novel or innovative intelligent scheduling analysis system for real-time collection of driving behavior data and physiological information, which can collect driver's driving behavior data and physiological information in real time to perform smart scheduling and improve driving safety.

The intelligent scheduling analysis system for real-time collection of driving behavior data and physiological information in the present invention includes: at least one of a mobile device sensor module and a wearable device sensor module, which collects or senses transportation loads in real time Driver’s driving behavior data and physiological information; a machine learning automatic analysis module, which is a driving that is collected or sensed in real time by at least one of the mobile device sensor module and the wearable device sensor module Machine learning is performed on human driving behavior data and physiological information to automatically analyze the correlation result or correlation model between the driver’s driving behavior data and physiological information; and a smart scheduling module based on machine learning automatic analysis module The correlation result or correlation model of the analyzed driving behavior data and physiological information of the driver is used for intelligent scheduling to generate a schedule related to the driver.

In order to make the above-mentioned features and advantages of the present invention more comprehensible, embodiments are specifically described below in conjunction with the accompanying drawings. In the following description, additional features and advantages of the present invention will be described, and these features and advantages will be partly known from the description, or can be learned by practicing the present invention. The features and advantages of the present invention are realized and achieved by means of the elements and combinations specifically pointed out in the scope of the patent application. It should be understood that the foregoing general description and the following detailed description are both illustrative and explanatory, and are not intended to limit the scope of the present invention.

1‧‧‧Smart scheduling analysis system

10‧‧‧Mobile device sensor module

20‧‧‧Wearable device sensor module

30‧‧‧Signal processing module

40‧‧‧Signal Correlation Evaluation Module

50‧‧‧Machine learning automatic analysis module

60‧‧‧Feature Grouping Module

70‧‧‧Smart scheduling module

A1‧‧‧Operation Database

A2‧‧‧Driving Database

A3‧‧‧Physiological Database

B1‧‧‧Operational risk

B2‧‧‧Driving risk

B3‧‧‧Medical Physiological Risk

D‧‧‧Estimated value of work performance indicators

E‧‧‧Schedule

Figure 1 is a schematic diagram of the structure of the intelligent scheduling analysis system for real-time collection of driving behavior data and physiological information in the present invention;

Figure 2 is a schematic diagram of the machine learning automatic analysis module of the present invention;

Figure 3 is a schematic diagram of the calculation of operational risks, driving risks and medical physiological risks in the present invention;

Figure 4 is a schematic diagram of the driving route matching analysis of the driver in the present invention;

Figure 5 is another schematic diagram of the driver's driving route matching analysis in the present invention; and

Figure 6 is a schematic diagram of the smart scheduling module to optimize the driver's schedule in the present invention.

The following describes the implementation of the present invention with specific specific embodiments. Those familiar with this technology can understand the other advantages and effects of the present invention from the contents disclosed in this specification, and can also implement other different specific equivalent embodiments. Or apply.

Figure 1 is a schematic diagram of the structure of the intelligent scheduling analysis system 1 for real-time collection of driving behavior data and physiological information in the present invention. As shown in the figure, the smart scheduling analysis system 1 may include a mobile device sensor module 10, a wearable device sensor module 20, a signal processing module 30, and a signal correlation that are connected to or communicate with each other. The evaluation module 40, a machine learning automatic analysis module 50, a feature grouping module 60, and a smart scheduling module 70.

The mobile device sensor module 10 can be a hardware sensor or a software sensing program (such as an APP), and is installed in a mobile device (not shown) or connected/communicated with the mobile device. The wearable device sensor module 20 can be a hardware sensor or a software sensing program (such as an APP), which is installed in a wearable device (not shown) or is connected/communicated with the wearable device. The mobile device can be a mobile phone, a smart phone, a tablet computer, or a personal digital assistant (PDA), and the wearable device can be a watch, a smart watch, or a medical watch, etc. The signal processing module 30 can be a hardware signal processor or a software signal processing program, the signal correlation evaluation module 40 can be a software signal correlation evaluation program, and the machine learning automatic analysis module 50 can be a software machine learning Automatic analysis program, the feature grouping module 60 can be a software feature grouping program, and the smart scheduling module 70 can be a software smart scheduling program. The following transportation vehicles can be passenger transport, tour buses, buses, taxis, Taiwan Railways, high-speed rail, MRT and other vehicles, and the servers can be back-end servers, web servers, or cloud servers, etc. However, the present invention is not limited to this.

At least one of the mobile device sensor module 10 and the wearable device sensor module 20 can collect or sense the driving behavior data and physiological information of the driver of the transportation vehicle (not shown) in real time. The machine learning automatic analysis module 50 can perform machine learning on the driver’s driving behavior data and physiological information collected or sensed by at least one of the mobile device sensor module 10 and the wearable device sensor module 20 in real time. , To automatically analyze the correlation result or correlation model between the driver's driving behavior data and physiological information. The smart scheduling module 70 can perform smart scheduling based on the correlation result or correlation model of the driver's driving behavior data and physiological information analyzed by the feature grouping module 60 and the driver's driving behavior data and the physiological information analyzed by the machine learning automatic analysis module 50. Related to the driver's schedule E (see Figure 6).

Specifically, the mobile device sensor module 10 and/or the wearable device sensor The module 20 can collect the driving behavior data of the driver, the physiological information of the driver, the driving history data of the transportation vehicle, and so on. For example, the mobile device sensor module 10 can collect or sense the driving behavior data of the driver in real time, and the wearable device sensor module 20 can collect or sense the physiological information of the driver in real time. At the same time, the present invention can transmit the driving behavior data and physiological information of the driver collected by the mobile device sensor module 10 and/or the wearable device sensor module 20 to the server (not shown), and then according to the driving The characteristics of human driving behavior data and physiological information are matched or scored to the driving situation to automatically generate a high safety schedule E.

The signal processing module 30 can use the mobile device sensor module 10 and/or the wearable device sensor module 20 to collect the driving behavior data of the driver, the physiological information of the driver, the driving history data of the transportation vehicle, etc. , For re-sampling, signal discretization, signal standardization, filter analysis, event detection or data pre-processing, etc. The signal correlation evaluation module 40 can perform time-frequency analysis or correlation analysis on the driving behavior data of the driver, the physiological information of the driver, and the driving history data of the transportation vehicle processed by the signal processing module 30.

The machine learning automatic analysis module 50 can perform machine learning based on the driving behavior data of the driver, the physiological information of the driver and the driving history data of the transportation vehicle analyzed by the signal correlation evaluation module 40 to automatically analyze, filter or evaluate The result of the correlation or the correlation model (such as the best model) of the driving behavior data, the physiological information of the driver, and the driving history data of the transportation vehicle. The smart scheduling module 70 can use the clustering results of the feature clustering module 60 and the analysis results of the machine learning automatic analysis module 50 to perform smart scheduling, so as to automatically generate a high-safety scheduling table E. Therefore, the smart scheduling module 70 of the present invention can incorporate the driver's individual physiological differences and actual driving conditions to provide matching recommendations suitable for the driver and the route.

In detail, the signal correlation evaluation module 40 can simultaneously analyze the driver’s driving The relationship between at least two signals (such as orientation or coherence) between behavioral data and physiological information (physiological signals). Driving behavior data can be rapid acceleration and deceleration, sharp turns, number of speeding events, constant speed holding ratio, driving time ratio, etc., and physiological signals can be heart rate measurement index, blood oxygen saturation monitoring index, stress index, heart rate variability, etc. For example, Coh (heart rate variability, the score of the constant speed keeping ratio) = 0.9, and Coh represents coherence. Accordingly, the present invention can find highly correlated data pairs, and provide scientific evidence supporting the coherence of collaborative behavior and personal physiological information.

The time-frequency analysis of the signal correlation evaluation module 40 can use Fourier analysis or autoregressive model coefficients to obtain the correlation coefficient of the two signals related to the driver's driving behavior data and physiological information (physiological signal). Moreover, the correlation analysis of the signal correlation evaluation module 40 can calculate the converted driving behavior data of the driver and the physiological information (physiological signal) from the wearable device sensor module 20 to analyze the driving behavior data and the physiological The correlation between the two information (physiological signals), for example, the correlation between the driving behavior data and the physiological information (physiological signals) is a positive relationship, a reverse relationship, or the magnitude of the two. In this way, it is analyzed whether the coherence between the driving behavior data and the physiological information (physiological signal) has reached a significant statistical result through the statistical verification method.

Figure 2 is a schematic diagram of the machine learning automatic analysis module 50 of the present invention. As shown in the figure, the field data used by the machine learning automatic analysis module 50 is related to factors such as the driver’s driving behavior data, physiological information, and driving history data, and may include [1] heart rate measurement index, blood oxygen saturation Monitoring index, stress index, heart rate variability; [2] number of sudden accelerations and decelerations, sharp turns, speeding events, constant speed keeping ratio, driving time ratio; [3] known disease characteristics, questionnaire results; [4] orientation, Homology; or [5] route travel time, route travel distance, single-day route, number of round trips, total number of traffic jams, etc.

In addition, the machine learning automatic analysis module 50 can be based on the aforementioned use field data (because Sub), that is, factors related to the driver's driving behavior data, physiological information, and driving history data. The advanced factors are automatically generated through basic calculation combinations of factors, bagging algorithm grouping, sensitivity analysis, or decision tree rule analysis, etc. According to the aforementioned advanced factors, the important factors are selected through stepwise regression, random forest, Information Value (IV value), genetic algorithm, resampling algorithm, or Elastic Net, etc., and based on the aforementioned important factors through Luo Guice regression, random forest, decision tree, SVM (support vector machine)/SVR (support vector regression), deep learning, K-means classification or Bayesian network and other methods to generate multiple model combinations, Then according to the aforementioned multiple model combination, through confusion matrix, ROC curve (receiver operating characteristic curve; receiver operating characteristic curve), F value (F-Measure) or mean square error, etc. to screen or evaluate the correlation result or correlation model (such as the best model) ).

Figure 3 is a schematic diagram of the calculation of the operational risk B1, the driving risk B2, and the medical physiological risk B3 of the present invention. As shown in the figure, the signal correlation evaluation module 40, the machine learning automatic analysis module 50, and the feature grouping module 60 can obtain the operational database A1 of the transportation vehicle and the driver, the driving database A2, and the physiological data of the driver. Database A3, to further calculate factors related to operational risk B1, driving risk B2 (driving behavior data), and medical physiological risk B3 (physiological information). For example, the data of operational risk B1 includes factors such as route travel time, route travel distance, number of round trips in a single day or total traffic congestion in a single day, and data of driving risk B2 (driving behavior data) includes constant speed keeping ratio, time period, road Factors such as grade, rapid acceleration and deceleration, sharp turns, medical physiological risk B3 (physiological information) data include questionnaire results, sleep index (such as sleep efficiency (Eff)), blood oxygen saturation monitoring index (such as blood oxygen saturation) Factors such as Oxygen desaturation index (ODI), heart rate variability, heart rate measurement index, or the number of abnormal physiological information.

Figure 4 is a schematic diagram of the driver's driving route matching analysis in the present invention. As shown in the figure, the wearable device sensor module 20 (such as a medical watch) in Figure 1 collects the physiological information (medical physiological information) of the driver, and the smart scheduling module 70 uses the data of the operational risk B1 Match the driving route of the driver with the corresponding driving risk B2 (driving behavior data) or medical physiological risk B3 (physiological information) data. For example, the data of operational risk B1 includes route travel time, route travel distance, number of round trips per day, total number of traffic jams on a single-day route, or driving age, etc. Data of driving risk B2 (driving behavior data) includes constant speed keeping ratio, rapid increase The number of decelerations, sharp turns or speeding events, etc. The data of medical physiological risk B3 (physiological information) includes sleep index, blood oxygen saturation monitoring index, heart rate variability or heart rate measurement index, etc. Therefore, the present invention will produce different driving types for different operating risks B1, driving risks B2 (driving behavior data) or medical physiological risks B3 (physiological information), and different driving types require different driving characteristics, which will also affect driving. The pairing of human driving routes has different effects.

Figure 5 is another schematic diagram of the driving route matching analysis of the present invention. As shown in Figure 5, the signal correlation evaluation module 40 in Figure 1 analyzes that the factors of operational risk B1 are mainly the total travel time of the route, the number of round trips in a single day, and the total number of traffic jams in a single day, and analyzes the medical physiology The factors of risk B3 (physiological information) are mainly sleep index and heart rate measurement index.

Then, the machine learning automatic analysis module 50 can perform machine learning on the operational risk B1, the driving risk B2, and the medical physiological risk B3 (physiological information) analyzed by the signal correlation evaluation module 40 to automatically analyze the work performance index estimation Value D. For example, the machine learning automatic analysis module 50 uses Logis regression to analyze: work performance index = a ₁ * x ₁ + a ₂ * x ₂ + a ₃ * x ₃ +…+ a _n-1 * x _{n -1} + a _n * x _n . The preceding a respective _I (as a ₁ to a _n of any one of) any of the explanatory variables influence coefficient, and x _i (x ₁ as according to any one of x _n) to the actual value of the variable to interpretation.

Fig. 6 is a schematic diagram of the smart scheduling module 70 of Fig. 1 to optimize the driver’s schedule E in the present invention. As shown in the figure, the machine learning automatic analysis module 50 in Figure 5 analyzes: work performance index=f(x ₁ , x ₂ , x ₃ ,..., x _n-1 , x _n ). The aforementioned work performance indicators can be internal self-determined performance indicators of the transportation company, such as transportation performance achievement rate, personal output value, personal work quality achievement rate or accident rate, etc. The company's objectively recognized operating goals, and the f(.) function is used The combination of multiple models shown in the machine learning automatic analysis module 50 can be generalized to a multiple combination function of any explanatory variable. _{The data source of the explanatory variable x 1~n} is operational risk B1, driving risk B2 (driving behavior data) medical and physiological risk B3 (physiological information), and through the model evaluation methods to identify the best model of combination coefficient (a ₁ ~ a _n), whereby the driver can calculate the unknown risk estimates of job performance indicators D.

The feature clustering module 60 can use an unsupervised learning method to perform cluster analysis using operational risk B1, driving risk B2 (driving behavior data) and medical physiological risk B3 (physiological information), such as K-means classification or analytic hierarchy process, to target driving Different characteristics of the driver (such as the result of the correlation between the driving behavior data and the physiological information of the driver) classify or group the drivers, such as the ethnic group of long/short-circuit drivers, the ethnic group of day/night departure drivers, or various age groups The race of drivers.

The smart scheduling module 70 can calculate the estimated value of f(.) for each driver in each group defined by the characteristic grouping module 60 according to the input driving list. After all calculations are completed, the smart scheduling module 70 can then Find the best combination that meets the maximum value of F(.) and the required quotas for each group (the same driver can only be scheduled to a single group), where F(.) refers to the forecast of each driver f(.) in all groups Total valuation. Then, the F(.) value is used to determine the priority candidates for each group to output the best overall Shift table E.

The smart scheduling module 70 performs group matching and sorting of all drivers through a server (such as a back-end server) to find the best matching score C1 for the long/short-circuit group or the day/night departure group, and consider the continuity Compliance conditions such as the number of days or working hours are then output for a period of time (such as the current week or the current month), or when there is a manpower shortage, the smart scheduling module 70 finds the best agent driver.

In summary, the intelligent scheduling analysis system for real-time collection of driving behavior data and physiological information in the present invention can at least have the following characteristics, advantages or technical effects.

1. The present invention can collect the driving behavior data of the driver, the physiological information of the driver and the driving history data of the transportation vehicle through the mobile device sensor module and/or the wearable device sensor module, and then according to the driver The characteristics of the driving behavior data and physiological information are matched or scored to the driving situation, so as to automatically generate a high safety schedule.

2. The smart scheduling module of the present invention can incorporate the driver's individual physiological differences and actual driving conditions, so as to provide matching recommendations suitable for the driver and the route to improve driving safety.

3. The feature grouping module of the present invention can optimize the classification or grouping of drivers, including the classification or grouping of drivers for different characteristics such as driving behavior data and physiological information of drivers, so as to improve the transportation vehicle and its driver. Driving safety.

4. Different from the fixed scheduling method of traditional transportation vehicles, the present invention energizes the actual driving behavior data and physiological information of the driver, and scores the route and time period according to the difference of different drivers to optimize the matching of the driver and the route , In order to improve the safety of transportation vehicles, and also provide more operational information on the management side of transportation vehicles.

The above embodiments only illustrate the principles, features and effects of the present invention. It is not intended to limit the applicable scope of the present invention. Anyone familiar with the art can modify and change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Any equivalent changes and modifications made using the content disclosed in the present invention should still be covered by the scope of the patent application. Therefore, the protection scope of the present invention should be as listed in the scope of the patent application.

1‧‧‧Smart scheduling analysis system

10‧‧‧Mobile device sensor module

20‧‧‧Wearable device sensor module

30‧‧‧Signal processing module

40‧‧‧Signal Correlation Evaluation Module

50‧‧‧Machine learning automatic analysis module

60‧‧‧Feature Grouping Module

70‧‧‧Smart scheduling module

A1‧‧‧Operation Database

A2‧‧‧Driving Database

A3‧‧‧Physiological Database

B1‧‧‧Operational risk

B2‧‧‧Driving risk

B3‧‧‧Medical Physiological Risk

Claims (11)

A smart scheduling analysis system for real-time collection of driving behavior data and physiological information, comprising: at least one of a mobile device sensor module and a wearable device sensor module, which collects or senses transportation vehicles in real time The driving behavior data and physiological information of the driver; a machine learning automatic analysis module, which is collected or sensed in real time by at least one of the mobile device sensor module and the wearable device sensor module The driver’s driving behavior data and physiological information are machine-learned, and the machine-learning automatic analysis module uses confusion matrix, receiver operating characteristic (ROC) curve, F-Measure, or mean square error to filter or evaluate it The correlation result or correlation model between the driver’s driving behavior data and physiological information; a smart scheduling module is based on the machine learning automatic analysis module through the confusion matrix, the receiver’s operating characteristic curve, the F value or the average The correlation result or correlation model between the driver’s driving behavior data and physiological information filtered or evaluated by the variance will schedule the driver to generate the driver’s schedule; and a signal correlation evaluation module, It is to analyze the directivity or coherence of at least two signals related to both the driving behavior data and the physiological information of the driver. For example, the intelligent scheduling analysis system described in the first item of the scope of patent application, wherein at least one of the mobile device sensor module and the wearable device sensor module further collects driving history data of the transportation vehicle , And the machine learning automatic analysis module performs machine learning on the driver’s driving behavior data and physiological information and the driving history data of the transportation vehicle to automatically analyze the driver’s driving behavior data and physiological information and the transportation The correlation result or correlation model of the vehicle's driving history data. For example, the intelligent scheduling analysis system described in the first item of the patent application further includes a signal processing module, which is collected by at least one of the mobile device sensor module and the wearable device sensor module The driving behavior data and physiological information of the driver are resampled, signal discretized, signal standardization, filter analysis, event detection or data pre-processing. For example, the intelligent scheduling analysis system described in item 3 of the scope of patent application, wherein the signal correlation evaluation module performs time-frequency analysis or time-frequency analysis of the driver’s driving behavior data and physiological information processed by the signal processing module Correlation analysis. For example, the intelligent scheduling analysis system described in item 4 of the scope of patent application, wherein the time-frequency analysis of the signal correlation evaluation module uses Fourier analysis or autoregressive model coefficients to obtain the driving behavior data and physiological information of the driver The correlation coefficient of the two signals related to the two, and the correlation analysis of the signal correlation evaluation module is to analyze whether the correlation between the driving behavior data of the driver and the physiological information is a positive relationship, a reverse relationship, or both The magnitude of the value. A smart scheduling analysis system for real-time collection of driving behavior data and physiological information, comprising: at least one of a mobile device sensor module and a wearable device sensor module, which collects or senses transportation vehicles in real time The driving behavior data and physiological information of the driver; a machine learning automatic analysis module, which is collected or sensed in real time by at least one of the mobile device sensor module and the wearable device sensor module The driver’s driving behavior data and physiological information are machine-learned, and the machine-learning automatic analysis module uses confusion matrix, receiver operating characteristic (ROC) curve, F-Measure, or mean square error to filter or evaluate it The correlation result or correlation model of the driver's driving behavior data and physiological information; A smart scheduling module is based on the combination of the driver’s driving behavior data and physiological information filtered or evaluated by the machine learning automatic analysis module through the confusion matrix, the receiver’s operating characteristic curve, the F value or the mean square error The correlation result or correlation model schedules the driver to generate the driver's schedule; and a signal correlation evaluation module, the time-frequency analysis of the signal correlation evaluation module uses Fourier analysis or autoregressive model The coefficient obtains the correlation coefficient of the two signals related to the driver's driving behavior data and physiological information, and the correlation analysis of the signal correlation evaluation module is to analyze both the driver's driving behavior data and physiological information The correlation is the value of the positive relationship, the reverse relationship, or both. For example, the intelligent scheduling analysis system described in item 1 or 6 of the scope of patent application, wherein the machine learning automatic analysis module is based on the driver’s driving behavior data and the factors of physiological information through factor basic calculation combinations and bagging algorithms The method of grouping, sensitivity analysis or decision tree rule analysis automatically produces advanced factors. A smart scheduling analysis system for real-time collection of driving behavior data and physiological information, comprising: at least one of a mobile device sensor module and a wearable device sensor module, which collects or senses transportation vehicles in real time The driving behavior data and physiological information of the driver; a machine learning automatic analysis module, which is collected or sensed in real time by at least one of the mobile device sensor module and the wearable device sensor module The driver’s driving behavior data and physiological information are machine-learned, and the machine-learning automatic analysis module uses confusion matrix, receiver operating characteristic (ROC) curve, F-Measure, or mean square error to filter or evaluate it The correlation result or correlation model of the driver's driving behavior data and physiological information; and A smart scheduling module is based on the combination of the driver’s driving behavior data and physiological information filtered or evaluated by the machine learning automatic analysis module through the confusion matrix, the receiver’s operating characteristic curve, the F value or the mean square error The correlation result or correlation model is used to schedule the driver to generate the driver's schedule. The machine learning automatic analysis module further uses stepwise regression, random forest, information value, genetic algorithm, resampling algorithm or flexibility The important factors of the driver’s driving behavior data and physiological information are screened out through the Internet, and then logistic regression, random forest, decision tree, SVM (support vector machine)/SVR (support vector regression) are used according to the important factors , Deep learning, K-means classification or Bayesian network methods to generate multiple model combinations. For example, the intelligent scheduling analysis system described in item 1, 6 or 8 of the scope of patent application, wherein the machine learning automatic analysis module also evaluates the obtained factors of operational risk, driving risk and medical physiological risk. Analyze, so that the smart scheduling module generates the best matching score based on the three factors of the operational risk, driving risk, and medical physiological risk, and then generates a scheduling table based on the best matching score. For example, the intelligent scheduling analysis system described in item 1, 6 or 8 of the scope of patent application further includes a feature grouping module based on the driving behavior data and physiological information of the driver analyzed by the machine learning automatic analysis module The correlation results of the two are used to classify the driver, so that the intelligent scheduling module will analyze the driving behavior data and physiological information of the driver according to the grouping result of the feature grouping module and the machine learning automatic analysis module The optimized intelligent scheduling of the two generates a scheduling table related to the driver. For example, the intelligent scheduling analysis system described in item 10 of the scope of patent application, wherein the feature grouping module adopts an unsupervised learning method to perform the driving based on the correlation result of the driver's driving behavior data and physiological information Groups of people.

Images