Driving safety monitoring method, device, electronic equipment and readable storage medium

Abstract

The present invention provides a driving safety monitoring method, device, electronic device and readable storage medium, which relate to the field of vehicle networking. By using the driving data regularly reported by the target vehicle, the driving risk level is continuously predicted at each current monitoring node during the driving process of the target vehicle based on the data of the driver's state change, driving warning situation and vehicle speed change in multiple consecutive preset cycles, and the driver is prompted in time when the driving risk level is medium risk or high risk, so as to effectively ensure the driving safety of the driver in medium and long-distance driving and avoid accidents.

Description

Technical Field

[0001] The present invention relates to the field of vehicle networking, and in particular to a driving safety monitoring method, device, electronic equipment and readable storage medium. Background Art

[0002] In the express logistics industry, there are usually air transport, road transport, pipeline transport and other modes of cargo transportation. For road transport, medium and long-distance transportation is generally carried out by medium and large trucks. The transportation journey of such trucks can be as short as a few hours or as long as one or two days, which is a great test for freight drivers.

[0003] During the driving process, freight drivers are bound to become fatigued or occasionally drive aggressively (speeding at high speeds, sudden acceleration, etc.), which poses a risk of causing traffic accidents. On the one hand, it threatens the lives of freight drivers and drivers of other vehicles affected. On the other hand, if an accident occurs, the goods may be delivered late at the very least, or even completely lost.

[0004] Therefore, how to supervise the driving safety of freight drivers during transportation journeys to ensure driving safety is an urgent problem that needs to be solved. Summary of the invention

[0005] The purpose of the present invention is to provide a driving safety monitoring method, device, electronic equipment and readable storage medium to improve the problems existing in the prior art.

[0006] The embodiments of the present invention can be implemented as follows:

[0007] In a first aspect, the present invention provides a driving safety monitoring method, comprising:

[0008] Receive driving data reported by the target vehicle at preset intervals; the driving data reflects changes in driver status, driving warning conditions, and vehicle position and speed changes within the preset period;

[0009] At the current monitoring node receiving the driving data, based on the driving data in a plurality of consecutive preset periods before the current monitoring node, statistical data of driving events within a first preset time length and a second preset time length before the current monitoring node are respectively counted;

[0010] Standardizing the traffic event statistics data within the first preset time period and the second preset time period respectively to obtain the traffic event statistics standardized data within the first preset time period and the second preset time period;

[0011] Using a pre-trained driving risk prediction model to process the driving event statistical standardized data within the first preset time period and the second preset time period to determine the driving risk level of the target vehicle at the current monitoring node;

[0012] If the driving risk level is medium risk or high risk, a driving safety reminder is issued to the driver of the target vehicle based on the driving event statistical data.

[0013] In an optional embodiment, the driving data includes driver status data, driving warning data and driving track speed information within the preset period; the driving event statistical data includes driver status statistical data, driving warning statistical data and vehicle speed warning statistical data; the second preset time length is equal to the plurality of consecutive preset periods; the second preset time length is equal to N times the first preset time length;

[0014] The step of respectively counting the traffic event statistics data within a first preset time length and a second preset time length before the current monitoring node based on the traffic data within a plurality of consecutive preset periods before the current monitoring node comprises:

[0015] Acquire driver status data, driving warning data, and driving track speed information for a plurality of consecutive preset periods before the current monitoring node;

[0016] Based on the driver status data of the plurality of consecutive preset periods, respectively counting driver status statistical data within the first preset time length and the second preset time length before the current monitoring node;

[0017] Based on the traffic warning data of the plurality of consecutive preset periods, respectively counting the traffic warning statistical data within the first preset time length and the second preset time length before the current monitoring node;

[0018] Based on the vehicle track speed information of the plurality of consecutive preset periods, the vehicle speed warning statistical data within the first preset time length and the second preset time length before the current monitoring node are respectively counted.

[0019] In an optional implementation, the driving risk prediction model is trained in the following manner:

[0020] Constructing a training set; the training set includes a plurality of training samples and a dangerous driving hazard label corresponding to each training sample, and each training sample includes a reference driver's driving event statistical data within a first preset time and a second preset time before a historical monitoring node;

[0021] Input the training set into the called automatic machine learning library for automatic optimization to obtain the optimal classification model algorithm and model parameter file;

[0022] The driving risk prediction model is established by using the classification model algorithm and the model parameter file.

[0023] In an optional implementation, the second preset duration is equal to N times the first preset duration; and the step of constructing a training set includes:

[0024] Obtaining full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information corresponding to multiple historical trips of a number of reference drivers;

[0025] For any historical trip of each of the reference drivers, based on the full-trip driver status data, full-trip driving warning data and full-trip driving track speed information corresponding to the historical trip, the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information corresponding to each historical trip subsection corresponding to the historical trip are obtained; wherein the historical trip is equal to a plurality of consecutive historical trip subsections, and the total duration of the historical trip subsections is equal to the first preset duration; and the historical monitoring nodes are determined at intervals of the second preset duration;

[0026] For any of the historical monitoring nodes of the historical trip, based on the short-distance driver status data, short-distance driving warning data, and short-distance driving track speed information of N consecutive historical trip sections before the historical monitoring node, driving event statistics within the first preset time length and the second preset time length before the historical monitoring node are respectively counted;

[0027] Traversing each of the historical monitoring nodes of the historical trip, obtaining driving event statistics within the first preset time length and the second preset time length before each of the historical monitoring nodes in the historical trip;

[0028] Traversing each of the historical trips of each of the reference drivers, obtaining driving event statistics within the first preset time length and the second preset time length before each of the historical monitoring nodes in each of the historical trips of each of the reference drivers;

[0029] Taking the driving event statistics data within the first preset time period and the second preset time period before each historical monitoring node in each historical trip of each reference driver as an original sample, to obtain a plurality of original samples;

[0030] Performing standardization on each of the original samples to obtain a plurality of standardized samples;

[0031] Each of the standardized samples is labeled with the dangerous driving hazard label to obtain the training set.

[0032] In an optional embodiment, the step of respectively collecting driving event statistics within the first preset time length and the second preset time length before the historical monitoring node based on the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information of N consecutive historical travel sections before the historical monitoring node includes:

[0033] Based on the short-distance driver status data of the N consecutive historical travel sections before the historical monitoring node, respectively counting the driver status statistical data within the first preset time length and the second preset time length before the historical monitoring node;

[0034] Based on the short-distance driving warning data of the N consecutive historical travel sections before the historical monitoring node, respectively counting driving warning statistical data within the first preset time length and the second preset time length before the historical monitoring node;

[0035] Based on the short-distance vehicle track speed information of the N consecutive historical travel sections before the historical monitoring node, respectively counting the vehicle speed warning statistical data within the first preset time length and the second preset time length before the historical monitoring node;

[0036] Among them, the driving event statistics within the first preset time length and the second preset time length before the historical monitoring node include: driver status statistics, driving warning statistics and vehicle speed warning statistics within the first preset time length and the second preset time length before the historical monitoring node.

[0037] In an optional implementation manner, before the step of receiving the driving data reported by the target vehicle at every preset period, the method further includes:

[0038] Determining risk probability ranges of multiple risk levels based on the training set and the driving risk prediction model;

[0039] The step of using the pre-trained driving risk prediction model to process the driving event statistical standardized data within the first preset time period and the second preset time period to determine the driving risk level of the target vehicle at the current monitoring node includes:

[0040] Inputting the statistical standardized data of driving events within the first preset time period and the second preset time period into the driving risk prediction model to obtain the target driving risk probability of the target vehicle at the current monitoring node;

[0041] The target driving risk probability is compared with the risk probability ranges of the multiple risk levels to determine the driving risk level of the target vehicle at the current monitoring node.

[0042] In an optional implementation, the risk level includes four types: no risk, low risk, medium risk, and high risk;

[0043] The step of determining the risk probability ranges of multiple risk levels based on the training set and the driving risk prediction model includes:

[0044] The training set is input into the driving risk prediction model to obtain a predicted driving risk probability corresponding to each training sample; the predicted driving risk probability represents the driving risk of the reference driver in the corresponding training sample at the historical monitoring node;

[0045] Clustering is performed on the predicted driving risk probabilities corresponding to all the training samples to obtain four classification groups; the four classification groups correspond to no risk, low risk, medium risk, and high risk, respectively;

[0046] Based on the minimum value and the maximum value in each of the classification groups, a risk probability range of the risk level corresponding to each of the classification groups is determined.

[0047] In a second aspect, the present invention provides a driving safety monitoring device, comprising:

[0048] A data receiving module is used to receive driving data reported by the target vehicle at preset intervals; the driving data reflects the driver's state changes, driving warning conditions, and vehicle speed changes within the preset period;

[0049] A statistics module is used for, at a current monitoring node receiving the driving data, respectively counting driving event statistics within a first preset time length and a second preset time length before the current monitoring node based on driving data within a plurality of consecutive preset periods before the current monitoring node;

[0050] A standardization module, used to perform standardization processing on the statistical data of traffic events within the first preset time length and the second preset time length, respectively, to obtain the statistical standardized data of traffic events within the first preset time length and the second preset time length;

[0051] A prediction module, configured to process the statistical standardized data of driving events within the first preset time period and the second preset time period using a pre-trained driving risk prediction model to determine the driving risk level of the target vehicle at the current monitoring node;

[0052] The reminder module is used to issue a driving safety reminder to the driver of the target vehicle based on the driving event statistical data when the driving risk level is medium risk or high risk.

[0053] In a third aspect, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory stores a software program, and when the electronic device is running, the processor executes the software program to implement the vehicle safety monitoring method as described in the first aspect above.

[0054] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the vehicle driving safety monitoring method described in the first aspect is implemented.

[0055] Compared with the prior art, the embodiments of the present invention provide a driving safety monitoring method, device, electronic device and readable storage medium, which can receive driving data reported regularly by the target vehicle, and at the current monitoring node where the driving data is received, based on the driving data in a plurality of consecutive preset periods before the current monitoring node, respectively count the driving event statistics within the first preset time length and the second preset time length before the current monitoring node. Since the driving data reflects the driver's state changes, driving warning conditions and vehicle speed changes within the preset period; therefore, the pre-trained driving risk prediction model can be used to process the standardized data of driving event statistics within the first preset time length and the second preset time length to determine the driving risk level of the target vehicle at the current monitoring node. Moreover, when the driving risk level is medium risk or high risk, a driving safety reminder is issued to the driver of the target vehicle based on the driving event statistics. In this way, by utilizing the driving data reported regularly by the target vehicle, the driving risk level can be continuously predicted at each current monitoring node during the target vehicle's driving process through three aspects of data: driver status changes, driving warning conditions, and vehicle speed changes. When the driving risk level is medium or high risk, the driver can be promptly reminded, thereby effectively ensuring the driver's driving safety during medium and long-distance driving and avoiding accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0057] Figure 1 One of the flow charts of a driving safety monitoring method provided by an embodiment of the present invention.

[0058] Figure 2 The second flowchart of a driving safety monitoring method provided by an embodiment of the present invention.

[0059] Figure 3The third flowchart of a driving safety monitoring method provided by an embodiment of the present invention.

[0060] Figure 4 A velocity scatter plot of all standardized samples after clustering provided by an embodiment of the present invention.

[0061] Figure 5 A schematic structural diagram of a vehicle driving safety monitoring device provided by an embodiment of the present invention.

[0062] Figure 6 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0064] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0066] In addition, the terms “first”, “second”, etc., if used, are merely used to distinguish between the descriptions and should not be understood as indicating or implying relative importance.

[0067] It should be noted that, in the absence of conflict, the features in the embodiments of the present invention may be combined with each other.

[0068] The driving safety monitoring method provided in the embodiment of the present invention can be applied to an electronic device, which can be a smart phone, a cloud server, etc. The following embodiment is described by taking a server as an example.

[0069] Please refer to Figure 1 , Figure 1 A flow chart of a driving safety monitoring method provided by an embodiment of the present invention, the method comprises the following steps S101-S105:

[0070] S101. Receive driving data reported by a target vehicle at preset intervals.

[0071] In this embodiment, the driving data may reflect changes in the driver's state, driving warning conditions, and vehicle position speed (position speed refers to position and speed) within a preset period.

[0072] S102. At the current monitoring node that receives the driving data, based on the driving data in a plurality of consecutive preset periods before the current monitoring node, statistical data of driving events within a first preset time period and a second preset time period before the current monitoring node are respectively counted.

[0073] S103 , standardize the statistical data of traffic events within the first preset time length and the second preset time length respectively to obtain the standardized statistical data of traffic events within the first preset time length and the second preset time length.

[0074] In this embodiment, the standardized data of driving event statistics may have inconsistent dimensions. The standardization process is to perform dimensionless processing on the standardized data of driving event statistics so that the data are comparable.

[0075] S104: Use a pre-trained driving risk prediction model to process the statistical standardized data of driving events within the first preset time period and the second preset time period to determine the driving risk level of the target vehicle at the current monitoring node.

[0076] S105: If the driving risk level is medium risk or high risk, a driving safety reminder is issued to the driver of the target vehicle based on the driving event statistical data.

[0077] In this embodiment, if the driving risk level is medium risk or high risk, it means that the target vehicle has a driving safety hazard at the current monitoring node, and an accident may easily occur if the driver is not careful, so it is necessary to remind the driver in time.

[0078] The driving safety monitoring method provided by the embodiment of the present invention uses the driving data regularly reported by the target vehicle to continuously predict the driving risk level at each current monitoring node during the driving process of the target vehicle based on the driver's state changes, driving warning conditions and vehicle speed changes in multiple consecutive preset cycles. When the driving risk level is medium risk or high risk, the driver is promptly prompted, thereby effectively ensuring the driver's driving safety in medium and long-distance driving and avoiding accidents.

[0079] In an optional implementation, the driving data reported by the target vehicle each time may include driver status data, driving warning data, and driving track speed information within a preset period. The following introduces these three types of data respectively:

[0080] Optionally, the target vehicle may be equipped with a camera aimed at the driver, and the control core of the target vehicle may perform behavior recognition on the surveillance video within a preset period collected in real time by the camera to obtain driver status data. The driver status data may include at least one driver status mark, and a driver status mark may include the time when the behavior occurred and the corresponding driver status type. In the optional example, the driver status type may include but is not limited to: fatigue, continuous eye closure (for example, eye closure time exceeds 1s), frequent blinking, yawning, answering or making phone calls, distraction, obstruction of the monitoring field of view, leaving the monitoring, frequent lowering of the head, etc.

[0081] Optionally, the radar carried by the target vehicle can detect the front, back, left, and right sides of the vehicle during driving, and based on the data collected by the radar in real time, a driving warning can be issued in time. Therefore, the driving warning data may include at least one driving warning, and the types of driving warnings may include but are not limited to: vehicle distance too close, front collision warning, left lane departure, pedestrian collision warning, rear blind spot warning, left blind spot warning, right blind spot warning, etc.

[0082] Optionally, the onboard GPS module of the target vehicle can perform real-time positioning and record the driving trajectory, while the speed measurement module of the target vehicle can measure the driving speed of the vehicle in real time. Therefore, the driving trajectory speed information can include the driving trajectory data and driving speed data within a preset period. In an optional example, the vehicle speed warning type determined based on the driving trajectory speed information can include but is not limited to: sudden acceleration, sudden deceleration, long-term idling, engine-off coasting, etc.

[0083] It can be understood that the time point of sudden acceleration or sudden deceleration can be found from the driving speed data, and combined with the driving speed and vehicle location at a time point, it can be determined whether the driving speed meets the speed requirements at the location of the vehicle. For example, if the vehicle is located in the leftmost lane of a three-lane highway, the speed limit of the lane is: 60km / h~120km / h, but the vehicle speed is lower than 60km / h for several seconds, it can be determined that the vehicle is idling continuously. This example is only an example and is not limited here.

[0084] Optionally, the second preset time length may be equal to a plurality of consecutive preset periods, and the second preset time length is equal to N first preset time lengths.

[0085] In a possible example, the sizes of the preset period, the first preset duration, and the second preset duration can be flexibly set. For example, the preset period can be 5s or 10s, which is only an example and is not limited in the embodiment of the present invention; the first preset duration and the second preset duration can be 15 minutes and 30 minutes, respectively, or 10 minutes and 20 minutes, respectively. This example is only an example and is not limited in the embodiment of the present invention.

[0086] Therefore, the driving event statistics may include driver status statistics, driving warning statistics, and vehicle speed warning statistics. Correspondingly, the process of "based on the driving data in a plurality of consecutive preset periods before the current monitoring node, respectively counting the driving event statistics within the first preset time length and the second preset time length before the current monitoring node" in the above step S102 may include the following sub-steps S1021 to S1024:

[0087] S1021. Obtain driver status data, driving warning data, and driving track speed information for multiple consecutive preset periods before the current monitoring node.

[0088] In the optional example, if the preset period is 1 minute, the first preset duration and the second preset duration are 10 minutes and 20 minutes respectively, and the current monitoring node is 12:00, then it is necessary to obtain the driver status data, driving warning data, and driving track speed information for 20 consecutive preset periods before 12:00 (i.e., 11:40~12:00). This example is only an example and is not limited here.

[0089] S1022. Based on the driver status data of a plurality of consecutive preset periods, statistical data of the driver status within a first preset time length and a second preset time length before the current monitoring node are respectively counted.

[0090] In this embodiment, the driver status statistics may include: the number of marks of various driver status types occurring within a corresponding preset time period.

[0091] It can be understood that there may be multiple driver status types in the multiple driver status marks in the driver status data of multiple consecutive preset periods. Therefore, the number of marks of each driver status type that appear within the first preset time length before the current monitoring node and the number of marks of each driver status type that appear within the second preset time length can be counted separately.

[0092] Continuing with the above example, in the driver status data of 20 consecutive preset cycles, if there are three types of driver status: continuous eye closure, frequent blinking, and yawning, the driver status statistical data within the first preset time (the first 10 minutes) and the second preset time (the first 20 minutes) can be shown in the following table:

[0093]

[0094] This example is merely an example and is not intended to be limiting.

[0095] S1023. Based on the traffic warning data of a plurality of consecutive preset periods, traffic warning statistical data within a first preset time length and a second preset time length before the current monitoring node are respectively counted.

[0096] In this embodiment, the traffic warning statistical data may include: the number of marks of various types of traffic warnings that appear within a corresponding preset time period.

[0097] It is understandable that there may be multiple types of traffic warnings in the traffic warning data of multiple consecutive preset periods. Therefore, the number of marks of each type of traffic warning that appears in the first preset time period before the current monitoring node and the number of marks of each type of traffic warning that appears in the second preset time period can be counted respectively.

[0098] Continuing with the above example, in the driving warning data of 20 consecutive preset cycles, if there are three types of driving warnings, namely, too close distance, front collision warning, and left lane departure, the driving warning statistics within the first preset time (the first 10 minutes) and the second preset time (the first 20 minutes) can be shown in the following table:

[0099]

[0100] This example is merely an example and is not intended to be limiting.

[0101] S1024: Based on the vehicle track speed information of a plurality of consecutive preset periods, the vehicle speed warning statistical data within a first preset time length and a second preset time length before the current monitoring node are respectively counted.

[0102] In this embodiment, the vehicle track speed information may include: the number of various types of vehicle speed warnings that appear within a corresponding preset time period. It can be understood that based on the vehicle track speed information of multiple consecutive preset periods, multiple vehicle speed warnings can be counted.

[0103] Continuing with the above example, in the driving warning data of 20 consecutive preset cycles, if there are two types of vehicle speed warnings, namely, rapid acceleration and rapid deceleration, the vehicle speed warning statistics within the first preset time (the first 10 minutes) and the second preset time (the first 20 minutes) can be shown in the following table:

[0104]

[0105] This example is merely an example and is not intended to be limiting.

[0106] In the above explanation of the lower steps S1021 to S1024 of step S102, the process of calculating the driving event statistics within the first preset time period before a current monitoring node and the driving event statistics within the second preset time period is described in detail. The driving event statistics include: the number of marks of various driver status types, the number of marks of various driving warning types, and the number of various vehicle speed warning types.

[0107] Optionally, the driving event statistics may also include driving statistics corresponding to the preset duration period, and the driving statistics may be determined based on the driving track speed information to reflect the mileage and speed distribution of the preset duration period. The driving statistics may include: mileage, minimum speed, maximum speed, average speed, speed standard deviation, speed difference standard deviation, etc., so step S103 needs to be dimensionless.

[0108] Next, the training process of the driving risk prediction model is introduced. It should be noted that the sample data content during the training of the driving risk prediction model is consistent with the data content obtained in the above step S102. Some descriptions are brief, and the details can be referred to in conjunction with the explanation of the above content.

[0109] For alternative implementations, see Figure 2 The training process of the driving risk prediction model may include the following steps S201 to S203:

[0110] S201. Build a training set.

[0111] In this embodiment, the training set may include several training samples and dangerous driving hazard labels corresponding to each training sample, and each training sample includes a reference driver's driving event statistics within a first preset time period and a second preset time period before a historical monitoring node.

[0112] Optionally, the construction of the training set can collect the full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information corresponding to the historical trip of the reference driver for splitting and statistics. Figure 2 Based on Figure 3 , the sub-steps of step S201 may include S2011~S2018.

[0113] S2011. Obtain full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information corresponding to multiple historical trips of several reference drivers.

[0114] Taking the medium and long-distance express transportation scenario under the express delivery company as an example, a freight driver is regarded as a reference driver, and his historical trip can be as short as several hours or as long as one or two days. Assume that the start time of a historical trip of a reference driver Xiao Xin is 06:00:00 on November 1, 2023, and the end time is 17:00:00 on November 1, 2023. The total duration of the entire historical trip is 11 hours. Therefore, the full-trip driver status data of the historical trip includes all driver status marks generated within these 11 hours, the full-trip driving warning data of the historical trip includes all driving warnings generated within these 11 hours, and the full-trip driving track speed information of the historical trip includes the vehicle's driving trajectory data and driving speed data for these 11 hours.

[0115] This example is merely an example and is not intended to be limiting.

[0116] S2012. For any historical trip of each reference driver, based on the full-trip driver status data, full-trip driving warning data and full-trip driving track speed information corresponding to the historical trip, obtain the short-trip driver status data, short-trip driving warning data and short-trip driving track speed information corresponding to each historical trip subsection corresponding to the historical trip.

[0117] In this embodiment, the historical journey is equal to a plurality of consecutive historical journey sections, and the total duration of the historical journey sections may be equal to the first preset duration; the historical monitoring nodes may be determined at intervals of the second preset duration.

[0118] Optionally, since the total duration of a historical trip is too long, segmentation processing may be performed: the historical trip is segmented into a number of historical trip sections that are equal to the first preset duration.

[0119] Continuing with the above example, for the historical trip starting from 06:00:00 on 2023.11.01 and ending at 17:00:00 on 2023.11.01, the historical trip with a total duration of 11 hours can be divided into 44 continuous historical trip sections according to the time axis, using the first preset duration of 15 minutes as the cutting granularity. In this way, the data of the entire trip is then split to obtain the short-distance driver status data, short-distance driving warning data, and short-distance driving track speed information corresponding to each historical trip section of the historical trip. This example is only an example and is not limited here.

[0120] S2013. For any historical monitoring node of the historical trip, based on the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information of N consecutive historical trip sections before the historical monitoring node, statistical data of driving events within the first preset time length and the second preset time length before the historical monitoring node are respectively counted.

[0121] In this embodiment, every second preset time period in a historical trip is a historical monitoring node. Therefore, for each historical monitoring node of the historical trip, it is necessary to perform data statistics based on the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information of N consecutive historical trip sections before the historical monitoring node, and respectively calculate the driving event statistical data within the first preset time period and the second preset time period before the historical monitoring node.

[0122] Optionally, the process of "calculating driving event statistics within a first preset time period and a second preset time period before a historical monitoring node based on the short-distance driver status data, short-distance driving warning data, and short-distance driving track speed information of N consecutive historical travel sections before the historical monitoring node" in step S2013 may include the following sub-steps S20131 to S20133:

[0123] S20131. Based on the short-distance driver status data of N consecutive historical travel sections before the historical monitoring node, respectively count the driver status statistical data within a first preset time length and a second preset time length before the historical monitoring node;

[0124] S20132, based on the short-distance driving warning data of N consecutive historical travel sections before the historical monitoring node, respectively counting the driving warning statistical data within a first preset time length and a second preset time length before the historical monitoring node;

[0125] S20133, based on the short-distance vehicle track speed information of N consecutive historical travel sections before the historical monitoring node, respectively counting the vehicle speed warning statistical data within a first preset time length and a second preset time length before the historical monitoring node;

[0126] In this embodiment, the driving event statistics within the first preset time length and the second preset time length before the historical monitoring node include: driver status statistics, driving warning statistics and vehicle speed warning statistics within the first preset time length and the second preset time length before the historical monitoring node.

[0127] S2014. Traverse each historical monitoring node of the historical journey to obtain driving event statistics within a first preset time length and a second preset time length before each historical monitoring node in the historical journey.

[0128] S2015. Traverse each historical trip of each reference driver to obtain driving event statistics within a first preset time length and a second preset time length before each historical monitoring node in each historical trip of each reference driver.

[0129] S2016. Statistical data of driving events within a first preset time period and a second preset time period before each historical monitoring node in each historical trip of each reference driver are taken as an original sample to obtain a plurality of original samples.

[0130] Continuing with the above example, for the historical trip of the reference driver Xiaoxin starting from 06:00:00 on 2023.11.01 and ending at 17:00:00 on 2023.11.01, assuming that the second preset time is 30 minutes, then based on the full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information of Xiaoxin's 11-hour historical trip, 22 original samples can be obtained. This example is only an example and is not limited here.

[0131] S2017. Standardize each original sample to obtain several standardized samples.

[0132] In this embodiment, the driving event statistics in the original sample may also include driving statistics, which may include data of different dimensions such as mileage, minimum speed, maximum speed, average speed, standard deviation of speed, standard deviation of speed difference, etc., so the original sample needs to be dimensionlessly processed.

[0133] S2018. Label each standardized sample with a dangerous driving hazard label to obtain a training set.

[0134] In this embodiment, the dangerous driving hazard label of the training sample can be 0 or 1, 0 represents the existence of dangerous driving hazard, and 1 represents the absence of dangerous driving hazard.

[0135] Optionally, each standardized sample can be labeled with a dangerous driving hazard by clustering. In the optional example, all standardized samples can be clustered using three clustering methods: Kmeans clustering, DBscan clustering, and hierarchical clustering. In this way, each clustering method can divide all standardized samples into two cluster groups. Then, a speed scatter plot can be constructed in a visual way to evaluate each clustering method.

[0136] For example, for Kmeans clustering, the velocity scatter plots of the two cluster groups K1 and K2 are as follows: Figure 4 As shown in the figure, the horizontal axis and the vertical axis are the average vehicle speed and the standard deviation of the vehicle speed in the original sample corresponding to the standardized sample. Figure 4In the figure, a light gray scattered point represents a standardized sample in cluster group K1, and a black scattered point represents a standardized sample in cluster group K2. It can be seen that all scattered points in cluster group K2 present a clustered distribution, and all scattered points in cluster group K2 also present a clustered distribution. Therefore, the clustering effect of clustering all standardized samples by Kmeans clustering is better. The dangerous driving hazard label of each standardized sample in cluster group K1 is 0, and the dangerous driving hazard label of each standardized sample in cluster group K2 is 1. This example is only an example and is not limited here.

[0137] S202, input the training set into the called automatic machine learning library for automatic optimization, and obtain the optimal classification model algorithm and model parameter file.

[0138] In this embodiment, the training set is input into the automatic machine learning library, which is an automated machine learning library that can use genetic algorithms to perform automated feature selection and model selection, and finally output the optimal classification model algorithm and model parameter file.

[0139] It should be noted that the called automatic machine learning library can be set according to the actual application requirements. For example, the automatic machine learning library can be selected but not limited to: TPOT, H20 AutoML, ELI5, PyCaret, etc. This example is only an example and is not limited here.

[0140] S203: Establish a driving risk prediction model using the classification model algorithm and model parameter file.

[0141] In this embodiment, based on the classification model algorithm and model parameter file output by the automatic machine learning library, a driving risk prediction model can be constructed. The driving risk prediction model can be used to predict the probability of driving risk.

[0142] In an optional implementation, multiple risk levels may be preset, and then the risk probability range of each risk level needs to be determined. After step S203, step S204 may also be included:

[0143] S204: Determine risk probability ranges for multiple risk levels based on the training set and the driving risk prediction model.

[0144] Optionally, the risk level may include four types: no risk, low risk, medium risk, and high risk. The sub-steps of step S204 may include S2041 to S2043:

[0145] S2041. Input the training set into the driving risk prediction model to obtain the predicted driving risk probability corresponding to each training sample.

[0146] In this embodiment, the predicted driving risk probability may represent the driving danger of the reference driver in the corresponding training sample at the historical monitoring node.

[0147] S2042: Cluster the predicted driving risk probabilities corresponding to all training samples to obtain four classification groups.

[0148] In this embodiment, the four classification groups may correspond to no risk, low risk, medium risk, and high risk;

[0149] S2043. Based on the minimum value and the maximum value in each classification group, determine the risk probability range of the risk level corresponding to each classification group.

[0150] In an optional implementation, during the application process, the driving risk prediction model outputs a probability value, and the risk probability range of each risk level needs to be used to determine the range to which the probability value belongs to determine the actual risk level. Correspondingly, the sub-steps of the above step S104 may include S1041~ S1042:

[0151] S1041. Input the statistical standardized data of driving events within the first preset time period and the second preset time period into a driving risk prediction model to obtain a target driving risk probability of the target vehicle at the current monitoring node.

[0152] S1042: Compare the target driving risk probability with the risk probability ranges of multiple risk levels to determine the driving risk level of the target vehicle at the current monitoring node.

[0153] In the optional example, assume that the risk probability ranges corresponding to the four risk levels are as follows:

[0154] 1. No risk: (0, 25%];

[0155] 2. Low risk: (25%, 40%];

[0156] 3. Medium risk: (40%, 75%];

[0157] 4. High risk: (75%, 100%];

[0158] Assuming that the target driving risk probability predicted in step S1041 is 80%, it can be determined that the driving risk level of the target vehicle at the current monitoring node is high risk. This example is only an example and is not limited here.

[0159] In an optional implementation, in the above step 105, the driving safety reminder to the driver of the target vehicle may be issued by voice, phone, text message, combination of sound and light signals, etc. The specific reminder content may be determined based on the driving event statistics within the first preset time length and the second preset time length before the current monitoring node.

[0160] For example, if the number of signs such as fatigue, continuous eye closure (for example, eye closure time exceeds 1s), frequent blinking, yawning, etc. exceeds a certain preset value, it means that the driver is driving fatigued, and the prompt content may be "Do not drive while tired, pay attention to rest"; if the number of signs such as close distance, front collision warning, left lane departure or pedestrian collision warning exceeds a certain preset value, it means that the driver is driving aggressively, and the prompt content may be "Please pay attention to the distance between vehicles and drive steadily"; if the average speed exceeds the speed limit value at the positioning point, it means that the driver is speeding, and the prompt content may be "Your current speed is s km / h! Please reduce the speed and do not exceed the speed limit." This example is only an example and is not limited here.

[0161] It should be noted that the execution order of each step in the above method embodiment is not limited to that shown in the drawings, and the execution order of each step shall be based on the actual application situation.

[0162] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0163] First, the present invention collects full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information corresponding to multiple historical trips of several reference drivers, and then calculates driving event statistics within the first preset time length and the second preset time length before each historical monitoring node in each historical trip of each reference driver. The driving event statistics include driver status statistics, driving warning statistics, vehicle speed warning statistics, and driving statistics data. Then, the two sets of driving event statistics corresponding to a historical monitoring node are used as a sample, standardized and labeled to construct a training set, and then the training set is used to train a driving risk prediction model. The driving risk prediction model can accurately predict the driving risk probability at a monitoring node; and the training set and the driving risk prediction model are used to determine the risk probability range of multiple risk levels, so that the driving risk probability output by the model can be converted into a corresponding risk level in the application stage;

[0164] Secondly, during the driving process of the target vehicle, the driving data of the target vehicle is collected regularly, so that at each current monitoring node, the driving event statistics within the first preset time and the second preset time before the current monitoring node are counted, and then the driving event statistics within the first preset time and the second preset time before the current monitoring node are input into the driving risk prediction model, so that the target driving risk probability of the target vehicle at the current monitoring node can be predicted in real time, and then the target driving risk probability is compared with the risk probability range of each risk level to determine the driving risk level of the target vehicle at the current monitoring node. When the driving risk level is medium risk or high risk, the driver is prompted in time, so as to effectively ensure the driving safety of the driver throughout the medium and long-distance driving process and avoid accidents.

[0165] In order to execute the corresponding steps in the above method embodiment and each possible implementation mode, a method for implementing a vehicle safety monitoring device is provided below.

[0166] See also Figure 5 , Figure 5 The schematic diagram of the structure of the vehicle safety monitoring device provided by the embodiment of the present invention is shown. The vehicle safety monitoring device 200 comprises: a data receiving module 210 , a statistical module 220 , a standardization module 230 , a prediction module 240 and a reminder module 250 .

[0167] The data receiving module 210 is used to receive the driving data reported by the target vehicle at every preset period; the driving data reflects the driver's state changes, driving warning conditions and vehicle speed changes within the preset period;

[0168] The statistics module 220 is used to, at the current monitoring node receiving the driving data, respectively count the driving event statistics within a first preset time length and a second preset time length before the current monitoring node based on the driving data within a plurality of consecutive preset periods before the current monitoring node;

[0169] The standardization module 230 is used to perform standardization processing on the traffic event statistical data within the first preset time length and the second preset time length, respectively, to obtain the traffic event statistical standardized data within the first preset time length and the second preset time length;

[0170] Prediction module 240, used to process the statistical standardized data of driving events within the first preset time period and the second preset time period using a pre-trained driving risk prediction model to determine the driving risk level of the target vehicle at the current monitoring node;

[0171] The reminder module 250 is used to issue a driving safety reminder to the driver of the target vehicle based on the driving event statistical data when the driving risk level is medium risk or high risk.

[0172] Optionally, the vehicle safety monitoring device 200 may further include a model training module 260, which may be used to implement the above steps S201 to S204 and their respective sub-steps.

[0173] Those skilled in the art can clearly understand that the data receiving module 210 can be used to implement the above step S101, the statistical module 220 can be used to implement the above step S102 and its sub-steps, the standardization module 230 can be used to implement the above step S103 and its sub-steps, the prediction module 240 can be used to implement the above step S104 and its sub-steps, and the reminder module 250 can be used to implement the above step S105. For the convenience and simplicity of description, the specific working process of the driving safety monitoring device 200 described above can refer to the corresponding process in the above method embodiment, and will not be repeated here.

[0174] See also Figure 6 , Figure 6 The electronic device 300 includes a processor 310 , a memory 320 , and a bus 330 , wherein the processor 310 is connected to the memory 320 via the bus 330 .

[0175] The memory 320 may be used to store software programs, for example, software programs corresponding to the vehicle safety monitoring device 200 provided in the embodiment of the present invention. The processor 310 executes various functional applications and data processing to implement the vehicle safety monitoring method provided in the embodiment of the present invention by running the software programs stored in the memory 320.

[0176] Among them, the memory 320 can be but is not limited to: RAM (Random Access Memory), ROM (Read Only Memory), FLASH (Flash Memory), PROM (Programmable Read-Only Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electric Erasable Programmable Read-Only Memory), etc.

[0177] The processor 310 may be an integrated circuit chip with signal processing capability. The processor 310 may be a general-purpose processor, including: CPU (Central Processing Unit), NP (Network Processor), etc.; it may also be: DSP (Digital Signal Processing), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0178] Understandably, Figure 6 The structure shown is for illustration only. The electronic device 300 may also include Figure 6 More or fewer components as shown, or with Figure 6 Different configurations are shown. Figure 6 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0179] The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the driving safety monitoring method disclosed in the above embodiment is implemented. The computer-readable storage medium can be, but is not limited to, various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a PROM, an EPROM, an EEPROM, a FLASH disk, or an optical disk.

[0180] In summary, the embodiments of the present invention provide a driving safety monitoring method, device, electronic device and readable storage medium, which can receive driving data reported regularly by the target vehicle; and at the current monitoring node where the driving data is received, based on the driving data in multiple consecutive preset periods before the current monitoring node, statistical data of driving events within the first preset time length and the second preset time length before the current monitoring node are respectively counted. Since the driving data reflects the changes in the driver's state, driving warning conditions and vehicle position speed changes within the preset period; therefore, the pre-trained driving risk prediction model can be used to process the standardized data of driving event statistics within the first preset time length and the second preset time length to determine the driving risk level of the target vehicle at the current monitoring node. Moreover, when the driving risk level is medium risk or high risk, a driving safety reminder is issued to the driver of the target vehicle based on the driving event statistical data. In this way, by utilizing the driving data reported regularly by the target vehicle, the driving risk level can be continuously predicted at each current monitoring node during the target vehicle's driving process through three aspects of data: driver status changes, driving warning conditions, and vehicle speed changes. When the driving risk level is medium or high risk, the driver can be promptly reminded, thereby effectively ensuring the driver's driving safety during medium and long-distance driving and avoiding accidents.

[0181] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A driving safety monitoring method, characterized in that: include: Receive driving data reported by the target vehicle at preset intervals; The driving data reflects the driver's state changes, driving warning conditions and vehicle speed changes within the preset period; At the current monitoring node receiving the driving data, based on the driving data in a plurality of consecutive preset periods before the current monitoring node, statistical data of driving events within a first preset time length and a second preset time length before the current monitoring node are respectively counted; the second preset time length is equal to N times the first preset time length; Standardizing the traffic event statistics data within the first preset time period and the second preset time period respectively to obtain the traffic event statistics standardized data within the first preset time period and the second preset time period; Using a pre-trained driving risk prediction model to process the driving event statistical standardized data within the first preset time period and the second preset time period to determine the driving risk level of the target vehicle at the current monitoring node; If the driving risk level is medium risk or high risk, a driving safety reminder is issued to the driver of the target vehicle based on the driving event statistical data; The driving risk prediction model is trained in the following way: Constructing a training set; the training set includes a plurality of training samples and a dangerous driving hazard label corresponding to each training sample, and each training sample includes a reference driver's driving event statistical data within a first preset time and a second preset time before a historical monitoring node; inputting the training set into the called automatic machine learning library for automatic optimization to obtain the optimal classification model algorithm and model parameter file; using the classification model algorithm and the model parameter file, establishing the driving risk prediction model; The steps of constructing a training set include: Obtaining full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information corresponding to multiple historical trips of a number of reference drivers; For any historical trip of each of the reference drivers, based on the full-trip driver status data, full-trip driving warning data and full-trip driving track speed information corresponding to the historical trip, the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information corresponding to each historical trip subsection corresponding to the historical trip are obtained; wherein the historical trip is equal to a plurality of consecutive historical trip subsections, and the total duration of the historical trip subsections is equal to the first preset duration; and the historical monitoring nodes are determined at intervals of the second preset duration; For any of the historical monitoring nodes of the historical trip, based on the short-distance driver status data, short-distance driving warning data, and short-distance driving track speed information of N consecutive historical trip sections before the historical monitoring node, driving event statistics within the first preset time length and the second preset time length before the historical monitoring node are respectively counted; Traversing each of the historical monitoring nodes of the historical trip, obtaining driving event statistics within the first preset time length and the second preset time length before each of the historical monitoring nodes in the historical trip; Traversing each of the historical trips of each of the reference drivers, obtaining driving event statistics within the first preset time length and the second preset time length before each of the historical monitoring nodes in each of the historical trips of each of the reference drivers; Taking the driving event statistics data within the first preset time period and the second preset time period before each historical monitoring node in each historical trip of each reference driver as an original sample, to obtain a plurality of original samples; Performing standardization on each of the original samples to obtain a plurality of standardized samples; Each of the standardized samples is labeled with the dangerous driving hazard label to obtain the training set.

2. The method according to claim 1, characterized in that The driving data includes driver status data, driving warning data and driving track speed information within the preset period; the driving event statistical data includes driver status statistical data, driving warning statistical data and vehicle speed warning statistical data; The second preset duration is equal to the plurality of consecutive preset periods; The step of respectively counting the traffic event statistics within a first preset time length and a second preset time length before the current monitoring node based on the traffic data within a plurality of consecutive preset periods before the current monitoring node comprises: Acquire driver status data, driving warning data, and driving track speed information for a plurality of consecutive preset periods before the current monitoring node; Based on the driver status data of the plurality of consecutive preset periods, respectively counting driver status statistical data within the first preset time length and the second preset time length before the current monitoring node; Based on the traffic warning data of the plurality of consecutive preset periods, respectively counting the traffic warning statistical data within the first preset time length and the second preset time length before the current monitoring node; Based on the vehicle track speed information of the plurality of consecutive preset periods, the vehicle speed warning statistical data within the first preset time length and the second preset time length before the current monitoring node are respectively counted.

3. The method according to claim 1, characterized in that The step of respectively counting driving event statistics within the first preset time length and the second preset time length before the historical monitoring node based on the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information of N consecutive historical travel sections before the historical monitoring node comprises: Based on the short-distance driver status data of the N consecutive historical travel sections before the historical monitoring node, respectively counting the driver status statistical data within the first preset time length and the second preset time length before the historical monitoring node; Based on the short-distance driving warning data of the N consecutive historical travel sections before the historical monitoring node, respectively counting driving warning statistical data within the first preset time length and the second preset time length before the historical monitoring node; Based on the short-distance vehicle track speed information of the N consecutive historical travel sections before the historical monitoring node, respectively counting the vehicle speed warning statistical data within the first preset time length and the second preset time length before the historical monitoring node; Among them, the driving event statistics within the first preset time length and the second preset time length before the historical monitoring node include: driver status statistics, driving warning statistics and vehicle speed warning statistics within the first preset time length and the second preset time length before the historical monitoring node.

4. The method according to claim 1, characterized in that: Before the step of receiving the driving data reported by the target vehicle at every preset period, the method further includes: Determining risk probability ranges of multiple risk levels based on the training set and the driving risk prediction model; The step of using the pre-trained driving risk prediction model to process the statistical standardized data of driving events within the first preset time period and the second preset time period to determine the driving risk level of the target vehicle at the current monitoring node includes: Inputting the statistical standardized data of driving events within the first preset time period and the second preset time period into the driving risk prediction model to obtain the target driving risk probability of the target vehicle at the current monitoring node; The target driving risk probability is compared with the risk probability ranges of the multiple risk levels to determine the driving risk level of the target vehicle at the current monitoring node.

5. The method according to claim 4, characterized in that The risk levels include four types: no risk, low risk, medium risk, and high risk; The step of determining the risk probability ranges of multiple risk levels based on the training set and the driving risk prediction model comprises: The training set is input into the driving risk prediction model to obtain a predicted driving risk probability corresponding to each training sample; the predicted driving risk probability represents the driving risk of the reference driver in the corresponding training sample at the historical monitoring node; Clustering is performed on the predicted driving risk probabilities corresponding to all the training samples to obtain four classification groups; the four classification groups correspond to no risk, low risk, medium risk, and high risk, respectively; Based on the minimum value and the maximum value in each of the classification groups, a risk probability range of the risk level corresponding to each of the classification groups is determined.

6. A driving safety monitoring device, characterized in that: include: A data receiving module, used to receive driving data reported by the target vehicle at preset intervals; The driving data reflects the driver's state changes, driving warning conditions and vehicle speed changes within the preset period; A statistics module is used for, at a current monitoring node receiving the driving data, respectively counting driving event statistics within a first preset time length and a second preset time length before the current monitoring node based on driving data within a plurality of consecutive preset periods before the current monitoring node; The second preset duration is equal to N times the first preset duration; A standardization module, used to perform standardization processing on the statistical data of traffic events within the first preset time length and the second preset time length, respectively, to obtain the statistical standardized data of traffic events within the first preset time length and the second preset time length; A prediction module, configured to process the statistical standardized data of driving events within the first preset time period and the second preset time period using a pre-trained driving risk prediction model to determine the driving risk level of the target vehicle at the current monitoring node; A reminder module, configured to issue a driving safety reminder to the driver of the target vehicle based on the driving event statistical data when the driving risk level is medium risk or high risk; The driving risk prediction model is trained in the following way: Constructing a training set; the training set includes a plurality of training samples and a dangerous driving hazard label corresponding to each training sample, and each training sample includes a reference driver's driving event statistical data within a first preset time and a second preset time before a historical monitoring node; inputting the training set into the called automatic machine learning library for automatic optimization to obtain the optimal classification model algorithm and model parameter file; using the classification model algorithm and the model parameter file, establishing the driving risk prediction model; The steps of constructing a training set include: Obtaining full-trip driver status data, full-trip driving warning data, and full-trip driving track speed information corresponding to multiple historical trips of a number of reference drivers; For any historical trip of each of the reference drivers, based on the full-trip driver status data, full-trip driving warning data and full-trip driving track speed information corresponding to the historical trip, the short-distance driver status data, short-distance driving warning data and short-distance driving track speed information corresponding to each historical trip subsection corresponding to the historical trip are obtained; wherein the historical trip is equal to a plurality of consecutive historical trip subsections, and the total duration of the historical trip subsections is equal to the first preset duration; and the historical monitoring nodes are determined at intervals of the second preset duration; For any of the historical monitoring nodes of the historical trip, based on the short-distance driver status data, short-distance driving warning data, and short-distance driving track speed information of N consecutive historical trip sections before the historical monitoring node, driving event statistics within the first preset time length and the second preset time length before the historical monitoring node are respectively counted; Traversing each of the historical monitoring nodes of the historical trip, obtaining driving event statistics within the first preset time length and the second preset time length before each of the historical monitoring nodes in the historical trip; Traversing each of the historical trips of each of the reference drivers, obtaining driving event statistics within the first preset time length and the second preset time length before each of the historical monitoring nodes in each of the historical trips of each of the reference drivers; Taking the driving event statistics data within the first preset time period and the second preset time period before each historical monitoring node in each historical trip of each reference driver as an original sample, to obtain a plurality of original samples; Performing standardization on each of the original samples to obtain a plurality of standardized samples; Each of the standardized samples is labeled with the dangerous driving hazard label to obtain the training set.

7. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a software program, and when the electronic device is running, the processor executes the software program to implement the driving safety monitoring method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the driving safety monitoring method according to any one of claims 1 to 5 is implemented.

Images