Dangerous driving behavior identification method and device based on multi-dimensional data analysis

Abstract

The invention discloses a dangerous driving behavior recognition method and device based on multi-dimensional data analysis, and the method comprises the steps: obtaining multi-dimensional data information corresponding to a vehicle, the multi-dimensional data information comprising alarm information and environment information which are quantitatively updated in real time according to time slices and alarm types; analyzing the multi-dimensional data information, if the multi-dimensional data information meets a score deduction rule, triggering the score deduction rule, and deducting the current score of the corresponding vehicle according to a deduction score corresponding to the score deduction rule to obtain a deducted score; identifying the dangerous driving behavior of the corresponding vehicle according to the deducted score; therefore, the alarm information is quantified and updated in real time through the time slice, and dynamic scoring is performed according to the score deduction rule in combination with the environment information, so that the analysis dimension is comprehensive, and the recognition result is more accurate.

Description

Technical Field

[0001] The present application relates to the field of safe driving technology, and in particular to a method and device for identifying dangerous driving behaviors based on multi-dimensional data analysis. Background Art

[0002] With the development of society, the recognition of dangerous driving behaviors of vehicles based on ADAS (Advanced Driver Assistance System) and DSM (Driver State Monitor) has become popular. However, the images collected by the vehicle-mounted terminal of this method are often unrelated. For example, lane departure, excessive speed, fatigue driving, and dangerous driving behaviors of the driver (smoking, distraction, fatigue, dozing, overtime driving) are all collected by various functional cameras and then identified as dangerous driving behaviors through intelligent algorithm recognition. However, there is no connection between various dangerous driving behaviors. Although the high-risk driving behaviors represented by independent data have certain persuasiveness, they also have certain limitations. In addition, due to the limitation of computing power and risk identification model, the recognition of driving behavior by vehicle-mounted equipment is too rough. Furthermore, there is a lack of collection and analysis of external environment data. Therefore, due to the incompleteness of the analysis dimension, the result of dangerous driving behavior recognition is not accurate enough. Summary of the invention

[0003] The present application aims to solve at least one of the technical problems in the above-mentioned technology to a certain extent. To this end, one purpose of the present application is to propose a dangerous driving behavior identification method based on multi-dimensional data analysis, which updates the alarm information in real time through time slicing, and dynamically scores according to the deduction rules in combination with environmental information, so that the analysis dimension is comprehensive and the identification result is more accurate.

[0004] To achieve the above-mentioned objectives, the first aspect embodiment of the present application proposes a method for identifying dangerous driving behaviors based on multidimensional data analysis, comprising the following steps: obtaining multidimensional data information corresponding to a vehicle, wherein the multidimensional data information includes alarm information and environmental information that are updated in real time and quantitatively by time slice and alarm type; analyzing the multidimensional data information, and if the multidimensional data information satisfies a deduction rule, triggering the deduction rule, and deducting the current score of the corresponding vehicle according to the deduction score corresponding to the deduction rule to obtain a deducted score; and identifying the dangerous driving behavior of the corresponding vehicle according to the deducted score.

[0005] According to the above-mentioned technical means, the embodiment of the present application first obtains multi-dimensional data information corresponding to the vehicle, wherein the multi-dimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types; then the multi-dimensional data information is analyzed, and if the multi-dimensional data information meets the deduction rules, the deduction rules are triggered, and the current score of the corresponding vehicle is deducted according to the deduction score corresponding to the deduction rules to obtain the deducted score; finally, the dangerous driving behavior of the corresponding vehicle is identified according to the deducted score; thus, the alarm information is quantitatively updated in real time through time slices, and dynamically scored according to the deduction rules in combination with environmental information, so that the evaluation dimensions are comprehensive and the identification results are more accurate.

[0006] In addition, the dangerous driving behavior identification method based on multidimensional data analysis proposed in the above embodiment of the present application may also have the following additional technical features:

[0007] Furthermore, multi-dimensional data information corresponding to the vehicle is obtained, including: using message middleware to receive alarm information collected by vehicle-mounted equipment; dynamically storing the alarm information according to a real-time slicing method to obtain a first alarm information that is updated in real time and quantified; dynamically storing the alarm information according to a time period slicing method to obtain a second alarm information that is updated in real time and quantified; and collecting weather data in real time to obtain the environmental information.

[0008] Furthermore, the alarm information is dynamically stored according to a real-time sharding method to obtain first alarm information that is updated in real-time quantification, including: creating a message queue; using a sliding window to dynamically store the alarm information according to a preset timeout period to obtain first alarm information that is updated in real-time quantification; wherein the alarm information is stored in a key-value storage method, and the vehicle ID, the corresponding alarm type identifier and the real-time sharding time are used as the unique key for each data item in the key-value storage, and the real-time sharding time includes the current, last five minutes, last ten minutes, last thirty minutes and last hour.

[0009] Furthermore, a sliding window is used to dynamically store the alarm information according to a preset timeout period, including: each time the latest alarm information is pulled from the message middleware into the message queue, a data sliding operation is immediately performed; if the alarm time in the alarm information is less than the timeout period, the corresponding data is removed from the message queue, and the latest alarm information is stored in the message queue, the count is updated, and the corresponding real-time shard key is updated at the same time.

[0010] Furthermore, the alarm information is dynamically stored according to a time period slicing method to obtain a second alarm information that is updated in real time and quantitatively, including: creating a message queue; using a rolling window to dynamically store the alarm information according to a preset start time and end time to obtain a first alarm information that is updated in real time and quantitatively; wherein the alarm information is stored in a key-value storage method, and the vehicle ID, the corresponding alarm type identifier and the time period slicing time are used as the unique key for each data item in the key-value storage, and the time period slicing time includes year, quarter, month, week, day, hour and 15 minutes.

[0011] Furthermore, a rolling window is used to dynamically store the alarm information according to the preset start time and end time, including: each time the latest alarm information is pulled from the message middleware into the message queue, a data rolling operation is immediately performed; if the alarm time in the alarm information is between the start time and the end time, the count is accumulated, otherwise it is discarded; when the deferral time is exceeded, the rolling window execution ends, the count is updated, and it is also updated to the corresponding time period shard key, wherein the deferral time is greater than the end time.

[0012] According to the above technical means, the present application stores the alarm information in two different time slices, identifies dangerous driving behaviors from two different dimensions: time period and real time, thereby further improving the recognition accuracy.

[0013] Furthermore, weather data is collected in real time to obtain the environmental information, including: obtaining real-time GPS longitude and latitude information reported by the vehicle-mounted equipment; parsing the GPS longitude and latitude information to obtain the corresponding physical address; obtaining the corresponding administrative division code according to the physical address; calling the weather service open API corresponding to the administrative division code to obtain the corresponding real-time weather data.

[0014] Furthermore, there are multiple deduction rules, each of which includes multiple condition indicators, corresponding deduction scores and a deduction upper limit. The multiple condition indicators include time period indicators of multiple alarm types, real-time indicators of multiple alarm types and weather indicators. If the multi-dimensional data information corresponding to a vehicle simultaneously satisfies multiple condition indicators corresponding to a certain deduction rule, the corresponding deduction rule is triggered, and the deduction rule is stopped from being triggered when the deduction upper limit is reached.

[0015] Further, the dangerous driving behavior of the corresponding vehicle is identified according to the deducted score, including: if the deducted score is less than a preset threshold, identifying the corresponding vehicle as having a dangerous driving behavior.

[0016] The second aspect of the present application provides a dangerous driving behavior identification device based on multidimensional data analysis, including an acquisition module, the acquisition module is used to obtain multidimensional data information corresponding to a vehicle, wherein the multidimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types; an analysis and scoring module, the analysis and scoring module is used to analyze the multidimensional data information, if the multidimensional data information meets the deduction rule, the deduction rule is triggered, and the current score of the corresponding vehicle is deducted according to the deduction score corresponding to the deduction rule to obtain the deducted score; an identification module, the identification module is used to identify the dangerous driving behavior of the corresponding vehicle according to the deducted score.

[0017] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of a method for identifying dangerous driving behaviors based on multidimensional data analysis according to an embodiment of the present application;

[0019] Figure 2 Schematic diagram of a flow chart of a dangerous driving behavior identification method based on multi-dimensional data analysis according to another embodiment of the present application;

[0020] Figure 3 4 is a block diagram of a dangerous driving behavior identification device based on multi-dimensional data analysis according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0022] In the related technologies, ADAS (Advanced Driver Assistance System) focuses on the interaction between the vehicle and the external environment, such as lane departure and front obstacle detection alarms, while DSM focuses on monitoring and analyzing the driver's behavior and status, such as fatigue, distraction or abnormal behavior. Therefore, the in-vehicle intelligent terminal with ADAS and DSM functions has certain intelligent driving assistance capabilities, but there are also certain limitations: First, the data correlation is weak and the characterization effect is inaccurate. For example: yawning does not necessarily mean fatigue, but if the yawning behavior is associated with overtime driving, then the judgment of fatigue driving is relatively accurate. If the yawning and driving time are calculated at a certain time frequency, and the risk level is output according to a certain model, the result is more convincing. In addition, the characterization results are too rough. For example: in a continuous driving journey, the risk levels of dangerous driving behaviors (smoking, yawning, distraction) generated by driving for 1 hour and 5 hours respectively cannot be treated equally. Furthermore, there is a lack of collection and analysis of external environmental data, such as weather, visibility, and road conditions. These factors are undoubtedly important objective factors in traffic accidents, but current on-board equipment is unable to collect and analyze them.

[0023] To this end, the present application proposes a dangerous driving behavior identification method based on multidimensional data analysis, firstly, multidimensional data information corresponding to the vehicle is obtained, wherein the multidimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types; then the multidimensional data information is analyzed, and if the multidimensional data information meets the deduction rule, the deduction rule is triggered, and the current score of the corresponding vehicle is deducted according to the deduction score corresponding to the deduction rule to obtain the deducted score; finally, the dangerous driving behavior of the corresponding vehicle is identified according to the deducted score; thus, the alarm information is quantitatively updated in real time through time slicing, and dynamically scored according to the deduction rule in combination with the environmental information, so that the analysis dimension is comprehensive and the identification result is more accurate.

[0024] In order to better understand the above technical solution, exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0025] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0026] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for identifying dangerous driving behaviors based on multidimensional data analysis according to an embodiment of the present application. Figure 1 As shown, the dangerous driving behavior identification method based on multidimensional data analysis includes the following steps:

[0027] S101, obtaining multi-dimensional data information corresponding to the vehicle, wherein the multi-dimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types.

[0028] As an embodiment, obtaining multi-dimensional data information corresponding to a vehicle includes: using a message middleware to receive alarm information collected by an on-board device; dynamically storing the alarm information according to a real-time slicing method to obtain a first alarm information that is updated in real-time quantification; dynamically storing the alarm information according to a time period slicing method to obtain a second alarm information that is updated in real-time quantification; and collecting weather data in real time to obtain environmental information.

[0029] That is to say, the alarm information collected by the on-board equipment corresponding to each vehicle is received through the message middleware, and the first alarm information and the second alarm information corresponding to each vehicle after update are obtained based on the quantitative calculation of the alarm indicators based on the time slices, wherein the on-board equipment are ADAS and DSM.

[0030] It should be noted that this can be implemented using technologies such as Kafka or RocketMQ. Both Kafka and RocketMQ are powerful message queue systems.

[0031] As a specific embodiment, the alarm information is dynamically stored according to the real-time sharding method to obtain the first alarm information that is updated in real-time quantification, including: creating a message queue; using a sliding window to dynamically store the alarm information according to a preset timeout period to obtain the first alarm information that is updated in real-time quantification; wherein the alarm information is stored in a key-value storage manner, and the vehicle ID, the corresponding alarm type identifier and the real-time sharding time are used as the unique key of each data item in the key-value storage, and the real-time sharding time includes the current, last five minutes, last ten minutes, last thirty minutes and last hour.

[0032] That is to say, when performing time slicing, storage and key design, the real-time slicing includes newest, last5minutes, last10minutes, last30 minutes, and last1 hour; Nosql (such as Hbase) is used to store data. Nosql is a key-value storage method. In this storage model, each data item has a unique key (Key) and a corresponding value (Value). The key is used to quickly retrieve data, and the value is the data associated with the key.

[0033] It should be noted that in order to keep the key unique and easy to read, the key of the real-time shard is designed as follows: assuming that the primary key of a vehicle is id and the fatigue driving alarm identifier is 301, the key for storing the number of fatigue alarms in the last 5 minutes (real-time time shard) is: last5minutes_301_id.

[0034] Furthermore, a sliding window is used to dynamically store the alarm information according to a preset timeout period, including: each time the latest alarm information is pulled from the message middleware into the message queue, a data sliding operation is immediately performed; if the alarm time in the alarm information is less than the timeout period, the corresponding data is removed from the message queue, and the latest alarm information is stored in the message queue, the count is updated, and the corresponding real-time shard key is updated at the same time.

[0035] In other words, the real-time sharding quantization calculation method is a sliding window, which is implemented as follows: design a queue (first-in, first-out), set a timeout for data removal from the queue according to the shard size (5 minutes, 10 minutes, etc.), recorded as: expiretime. Every time the latest message is pulled from the message middleware into this queue, a data sliding operation is immediately performed to remove the data with an alarm time less than expiretime from the queue, and the newly added data enters the end of the queue, the count is updated, and the corresponding real-time sharding key is updated at the same time.

[0036] As a specific embodiment, the alarm information is dynamically stored according to the time period slicing method to obtain the second alarm information that is updated in real time and quantitatively, including: creating a message queue; using a rolling window to dynamically store the alarm information according to the preset start time and end time to obtain the first alarm information that is updated in real time and quantitatively; wherein the alarm information is stored in a key-value storage method, and the vehicle ID, the corresponding alarm type identifier and the time period slicing time are used as the unique key of each data item in the key-value storage, and the time period slicing time includes year, quarter, month, week, day, hour and 15 minutes.

[0037] That is to say, when performing time slicing, storage, and key design, the time period slicing includes year, quarter, month, week, day, hour, and 15 minutes; in addition, the time period slicing also uses Nosql (such as: Hbase) to store data, and Nosql is a key-value storage method.

[0038] It should be noted that in order to keep the key unique and easy to read, the key of the real-time sharding is designed as follows: Assuming that the primary key of a vehicle is id and the fatigue driving alarm indicator is 301, the key of the fatigue driving data of the vehicle in the second 15 minutes (1 hour is divided into 4 quarters) at 17:00 on August 20, 2024 (time period sharding)

[0039] It is:20240820_17_quater_2_301_id.

[0040] Furthermore, a rolling window is used to dynamically store the alarm information according to the preset start time and end time, including: each time the latest alarm information is pulled from the message middleware into the message queue, a data rolling operation is immediately performed; if the alarm time in the alarm information is between the start time and the end time, the count is accumulated, otherwise it is discarded; when the deferral time is exceeded, the rolling window execution ends, the count is updated, and it is also updated to the corresponding time period shard key, where the deferral time is greater than the end time.

[0041] That is to say, the quantitative calculation method for time period sharding is a rolling window, which is implemented as follows: the rolling window statistical method sets the start time and end time, which are recorded as: starttime and endtime. Each time the latest message is pulled from the message middleware into the queue, a judgment is immediately made to determine whether the alarm time is within this time period. If so, the count is accumulated, otherwise it is discarded. In order to solve the problem of delayed data reporting and avoid data loss, the rolling window sets a deferral time (for example, 20 minutes). When the endtime is exceeded by 20 minutes, the rolling window execution ends, the count is updated, and it is also updated to the corresponding time period sharding key.

[0042] As an embodiment, real-time weather data is collected to obtain environmental information, including: obtaining real-time GPS longitude and latitude information reported by the vehicle-mounted equipment; parsing the GPS longitude and latitude information to obtain the corresponding physical address; obtaining the corresponding administrative division code according to the physical address; calling the weather service open API corresponding to the administrative division code to obtain the corresponding real-time weather data.

[0043] That is to say, the physical address can be parsed according to the real-time GPS longitude and latitude reported by the vehicle-mounted equipment, and the administrative division code (national standard) can be obtained according to the physical address, which is recorded as: area code (such as: 110117 identifies Pinggu District, Beijing). According to this area code, the weather service open API (such as Baidu Weather) can be called to obtain the real-time weather information of the administrative division code, such as: rain and snow and the corresponding intensity and visibility distance.

[0044] S102, analyzing the multidimensional data information, if the multidimensional data information satisfies the deduction rule, triggering the deduction rule, and deducting the current score of the corresponding vehicle according to the deduction score corresponding to the deduction rule to obtain the deducted score.

[0045] That is to say, a rule engine is used to dynamically score quantitative indicators and environmental data; specifically, a rule engine tool (drools) based on the RETE algorithm is used as the core module of dynamic scoring. The biggest advantage of the rule engine is the separation of facts and rules. The adjustment of rules can immediately affect the scoring results, and the real-time scoring update can identify possible dangerous driving behaviors in real time, making the identification accurate, real-time and flexible.

[0046] As an embodiment, there are multiple deduction rules, each of which includes multiple condition indicators, corresponding deduction scores and a deduction upper limit. The multiple condition indicators include time period indicators of multiple alarm types, real-time indicators of multiple alarm types and weather indicators. If the multi-dimensional data information corresponding to a vehicle simultaneously satisfies multiple condition indicators corresponding to a certain deduction rule, the corresponding deduction rule is triggered, and the deduction rule is stopped from being triggered when the deduction upper limit is reached.

[0047] That is to say, the design rules include: a. time period indicator rules integrating multiple alarm types; b. real-time indicator rules integrating multiple alarm types; c. weather. Among them, a, b, and c are in an "and" relationship, that is, the rule is triggered when the conditions of a, b, and c are met at the same time.

[0048] Specifically, the deduction rules are as follows:

[0049] Rule 1:

[0050]

[0051] It should be noted that this rule 1 represents the vehicle with vehicle id = 123. If on that day (2024.08.20), the cumulative driving time exceeds 120 minutes (value key = 20240820_day_501_123), and the average driving speed in the last 5 minutes exceeds 80km / h (value key = last5minutes_802_123), and the weather in the current driving area is moderate rain or light snow, when the device reports the second-level fatigue driving alarm (alarmType = 302_02), 10 points will be deducted from the current score immediately; in order to avoid frequent device alarms, resulting in repeated deductions of points and distorted effects, the upper limit of deductions for this rule is set at 30 points (no more points will be deducted after 30 points are deducted).

[0052] Rule 2:

[0053]

[0054]

[0055] It should be noted that this rule 2 represents the vehicle with vehicle id = 123. If on that day (2024.08.20), the cumulative driving time exceeds 120 minutes (value key = 20240820_day_501_123), and the average driving speed in the last 5 minutes exceeds 80km / h (value key = last5minutes_802_123), and the weather in the current driving area is moderate rain or light snow, when the device reports a lane departure alarm (alarmType = 502), 20 points will be deducted from the current score immediately; in order to avoid frequent device alarms, repeated deductions of scores, and distortion of scores, the upper limit of deductions for this rule is 40 points.

[0056] From the above, we can see that the two rules use the quantitative indicators "cumulative driving time of the day" and "average speed in the last 5 minutes", the external environment (rain and snow), and the equipment reports fatigue alarms or lane departure alarms in real time. Multiple data dimensions represent a very likely dangerous driving behavior, namely: in rainy and snowy weather, the driver drives continuously for more than 2 hours, the current speed is fast, and there is fatigue driving and lane departure.

[0057] S103, identifying the dangerous driving behavior of the corresponding vehicle according to the deducted score.

[0058] As an embodiment, the dangerous driving behavior of the corresponding vehicle is identified based on the deducted score, including: if the deducted score is less than a preset threshold, then the corresponding vehicle is identified as having a dangerous driving behavior.

[0059] That is to say, each vehicle has a current score. When the score is lower than a certain threshold, it is determined that the vehicle may have dangerous driving behavior; the device reports acc on (ignition), initializes the current vehicle score (the vehicle initial score can be set according to the vehicle type: such as dangerous goods, ordinary freight trucks, commercial vehicles, buses, and driver age groups, for example, the driver of a dangerous goods vehicle is over 60 years old, the initial score is 90, and the driver of an ordinary truck is 40 years old, and the initial score is 100; it can be flexibly configured, and this application does not make specific restrictions on this), and enters the driving behavior monitoring state; acc off terminates the scoring until the next acc on restarts the scoring; during the driving behavior monitoring process, the device continuously reports various alarms, and the quantitative indicators are continuously calculated and updated until each value triggers one or more rules, and the corresponding score is deducted when the rule is triggered.

[0060] In a specific embodiment of the present application, Figure 2As shown, the dangerous driving behavior identification method based on multi-dimensional data analysis proposed in the present application first obtains the alarm information corresponding to each vehicle through the message middleware, and realizes the quantitative statistics of real-time time-slicing indicators based on the sliding window of the queue and the quantitative statistics of time period slicing indicators based on the rolling window of the queue according to the alarm information; at the same time, the administrative division code is obtained according to the geographic coordinates, and the real-time weather data is obtained according to the administrative division code; finally, the rule engine is used to calculate the score, set the dangerous driving score threshold, initial value, and deduction cap value, and comprehensively judge the real-time quantitative indicator effectiveness value, alarm, weather and time period indicator value. Multiple conditions are effective at the same time. If the rule conditions are met, the deduction calculation is triggered, and the score is updated in real time. If the score is lower than the threshold, it is identified as a dangerous driving behavior.

[0061] In order to implement the above embodiment, the embodiment of the present application proposes a dangerous driving behavior identification device based on multi-dimensional data analysis, such as Figure 3 As shown, the dangerous driving behavior identification device based on multi-dimensional data analysis proposed in the present application includes: an acquisition module 10, an analysis and scoring module 20 and an identification module 30.

[0062] Among them, the acquisition module 10 is used to obtain multi-dimensional data information corresponding to the vehicle, wherein the multi-dimensional data information includes alarm information and environmental information that are updated in real time and quantitatively according to time slices and alarm types; the analysis and scoring module 20 is used to analyze the multi-dimensional data information. If the multi-dimensional data information meets the deduction rules, the deduction rules are triggered, and the current score of the corresponding vehicle is deducted according to the deduction score corresponding to the deduction rules to obtain the deducted score; the identification module 30 is used to identify the dangerous driving behavior of the corresponding vehicle according to the deducted score.

[0063] It should be noted that the aforementioned explanation of the embodiment of the dangerous driving behavior identification method based on multi-dimensional data analysis is also applicable to the dangerous driving behavior identification device based on multi-dimensional data analysis of this embodiment, and will not be repeated here.

[0064] In summary, according to the dangerous driving behavior identification device based on multidimensional data analysis of the embodiment of the present application, the multidimensional data information corresponding to the vehicle is obtained through the acquisition module, wherein the multidimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types; the analysis and scoring module analyzes the multidimensional data information, and if the multidimensional data information meets the deduction rules, the deduction rules are triggered, and the current score of the corresponding vehicle is deducted according to the deduction score corresponding to the deduction rules to obtain the deducted score; the recognition module identifies the dangerous driving behavior of the corresponding vehicle according to the deducted score; thus, the alarm information is quantitatively updated in real time through time slices, and is dynamically scored according to the deduction rules in combination with the environmental information, so that the evaluation dimensions are comprehensive and the recognition results are more accurate.

[0065] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0066] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0067] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0068] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0069] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0070] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0071] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0072] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A dangerous driving behavior identification method based on multidimensional data analysis, characterized in that: The following steps are involved: Acquire multi-dimensional data information corresponding to the vehicle, wherein the multi-dimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types; Analyzing the multidimensional data information, if the multidimensional data information satisfies a deduction rule, triggering the deduction rule, and deducting the current score of the corresponding vehicle according to the deduction score corresponding to the deduction rule to obtain a deducted score; The dangerous driving behavior of the corresponding vehicle is identified according to the deducted score.

2. The method for identifying dangerous driving behaviors based on multidimensional data analysis according to claim 1, characterized in that: Get multi-dimensional data information corresponding to the vehicle, including: Use message middleware to receive alarm information collected by vehicle-mounted equipment; Dynamically storing the alarm information in a real-time slicing manner to obtain first alarm information that is updated in real-time quantification; Dynamically storing the alarm information according to time segmentation to obtain second alarm information that is quantitatively updated in real time; The weather data is collected in real time to obtain the environmental information.

3. The dangerous driving behavior identification method based on multidimensional data analysis according to claim 2, characterized in that: The alarm information is dynamically stored according to a real-time slicing method to obtain first alarm information that is updated in real-time quantitative manner, including: Create a message queue; Using a sliding window to dynamically store the alarm information according to a preset timeout period to obtain first alarm information that is quantitatively updated in real time; Among them, the alarm information is stored in a key-value storage manner, and the vehicle ID, the corresponding alarm type identifier and the real-time sharding time are used as the unique key of each data item in the key-value storage. The real-time sharding time includes the current, last five minutes, last ten minutes, last thirty minutes and last hour.

4. The method for identifying dangerous driving behaviors based on multidimensional data analysis according to claim 3, characterized in that: The sliding window is used to dynamically store the alarm information according to a preset timeout period, including: Every time the latest alarm information is pulled from the message middleware and entered into the message queue, a data sliding operation is immediately performed; If the alarm time in the alarm information is less than the timeout time, the corresponding data will be removed from the message queue, and the latest alarm information will be stored in the message queue, the count will be updated, and it will be updated to the corresponding real-time shard key at the same time.

5. The method for identifying dangerous driving behaviors based on multidimensional data analysis according to claim 2, characterized in that: The alarm information is dynamically stored according to the time segmentation method to obtain the second alarm information updated in real time quantitatively, including: Create a message queue; The alarm information is dynamically stored using a rolling window according to a preset start time and end time to obtain first alarm information that is quantitatively updated in real time; The alarm information is stored in a key-value storage manner, and the vehicle ID, the corresponding alarm type identifier and the time segmentation time are used as the unique key for each data item in the key-value storage. The time segmentation time includes year, quarter, month, week, day, hour and 15 minutes.

6. The method for identifying dangerous driving behaviors based on multidimensional data analysis according to claim 5, characterized in that: The alarm information is dynamically stored using a rolling window according to a preset start time and end time, including: Every time the latest alarm information is pulled from the message middleware and entered into the message queue, a data rolling operation is immediately performed; If the alarm time in the alarm information is between the start time and the end time, the count is accumulated, otherwise it is discarded; When the extension time is exceeded, the rolling window execution ends, the count is updated, and it is updated to the corresponding time period shard key at the same time, where the extension time is greater than the end time.

7. The dangerous driving behavior identification method based on multidimensional data analysis according to claim 2, characterized in that: Collect weather data in real time to obtain the environmental information, including: Get the real-time GPS latitude and longitude information reported by the vehicle-mounted equipment; Parsing the GPS longitude and latitude information to obtain the corresponding physical address; Obtaining a corresponding administrative division code according to the physical address; Call the weather service open API corresponding to the administrative division code to obtain the corresponding real-time weather data.

8. The method for identifying dangerous driving behaviors based on multidimensional data analysis according to claim 2, characterized in that: There are multiple deduction rules, each of which includes multiple condition indicators, corresponding deduction scores and a deduction upper limit. The multiple condition indicators include time period indicators of multiple alarm types, real-time indicators of multiple alarm types and weather indicators. If the multi-dimensional data information corresponding to a vehicle simultaneously satisfies multiple condition indicators corresponding to a certain deduction rule, the corresponding deduction rule is triggered, and the deduction rule is stopped from being triggered when the deduction upper limit is reached.

9. The method for identifying dangerous driving behaviors based on multidimensional data analysis according to claim 8, characterized in that: Identifying the dangerous driving behavior of the corresponding vehicle according to the deducted score includes: If the deducted score is less than a preset threshold, the corresponding vehicle is identified as having dangerous driving behavior.

10. A device for identifying dangerous driving behavior based on multidimensional data analysis, characterized in that: include: An acquisition module, the acquisition module is used to acquire multi-dimensional data information corresponding to the vehicle, wherein the multi-dimensional data information includes alarm information and environmental information that are quantitatively updated in real time according to time slices and alarm types; An analysis and scoring module, the analysis and scoring module is used to analyze the multidimensional data information, and if the multidimensional data information satisfies a deduction rule, the deduction rule is triggered, and the current score of the corresponding vehicle is deducted according to the deduction score corresponding to the deduction rule to obtain a deducted score; An identification module is used to identify the dangerous driving behavior of the corresponding vehicle according to the deducted score.

Images