CN111231971B - Automobile safety performance analysis and evaluation method and system based on big data - Google Patents
Automobile safety performance analysis and evaluation method and system based on big data Download PDFInfo
- Publication number
- CN111231971B CN111231971B CN202010134138.6A CN202010134138A CN111231971B CN 111231971 B CN111231971 B CN 111231971B CN 202010134138 A CN202010134138 A CN 202010134138A CN 111231971 B CN111231971 B CN 111231971B
- Authority
- CN
- China
- Prior art keywords
- data
- vehicle
- driving
- driver
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/08—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to drivers or passengers
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/08—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to drivers or passengers
- B60W40/09—Driving style or behaviour
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Mathematical Physics (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a big data-based automobile safety performance analysis and evaluation method, which comprises the following steps of: inputting auxiliary driving strategy data to be evaluated, corresponding trigger event data for triggering the auxiliary driving strategy, vehicle running parameter data, environment data and behavior data of a driver into an evaluation model unit; the evaluation model unit evaluates the auxiliary driving strategy to be evaluated and outputs an evaluation result; and the automobile safety performance analysis and evaluation system based on the big data comprises a data acquisition module and a data analysis module. By adopting the technical scheme, big data analysis is carried out on the vehicle driving data, so that the safety of the vehicle driving assisting strategy is evaluated or the event information which cannot be matched with the driving assisting strategy is extracted, so that reference is provided for designers, and the improvement of the vehicle driving assisting strategy or the strategy complementary formulation of the event which cannot be matched with the driving assisting strategy is facilitated.
Description
Technical Field
The invention belongs to the technical field of analysis and application of automobile big data, and particularly relates to an automobile safety performance analysis and evaluation method and system based on big data.
Background
The automotive industry today faces increasing challenges and pressures: cost pressure, industry competition, globalization trends, and market changes and fluctuations. However, the large data and data analytics have created unprecedented possibilities for automobile manufacturers to address various challenges and issues. With the improvement of the analysis capability of the written data, the predictive analysis is evolving into a powerful tool, and the prediction efficiency, operation and performance can be greatly improved. The challenge is whether the automotive manufacturer can efficiently process these massive amounts of knowledge and experience data. In short, the huge amount of data makes evaluation difficult, and it is difficult to support the establishment of strategic decisions. Data analysis can fuse this information together. Whether a "machine-readable" data set, or unstructured data such as a visual, audio, or text, the effect will be surprising as long as it is properly processed. Through analysis of the big data platform, market recycling information can be provided for automobile manufacturers, and support can be provided for research, development, design and improvement of automobiles, such as fault data of each equipment, index data of automobile body light weight, new energy automobile performance evaluation data and the like. The big data platform generally obtains structured, unstructured or semi-structured data through a data source, integrates, cleans and stores the data through a basic platform, analyzes the data or performs modeling analysis, and finally interacts an analysis result with a user.
The safe driving of the automobile is a very important index of the automobile, the auxiliary driving system can carry out auxiliary control on the automobile through a series of vehicle-mounted auxiliary equipment, to improve the safety of automobile driving, such as the anticipation of traffic lights or the prompt of overspeed behavior, and can determine an auxiliary strategy through the analysis of driving behavior, for example, when the driver fails to take a deceleration or avoidance measure within a preset time when encountering an obstacle, appropriate deceleration control is performed, or the triggering time of the accelerator opening is delayed when a driver watches the mobile phone or bows a pickup in the driving process is monitored, so as to prevent the driver from accidentally stepping on the accelerator suddenly to accelerate the automobile and cause accidents, for example, the car driving assisting method disclosed in chinese patent document CN109050536A, and the composite automatic driving assisting decision system and method disclosed in chinese patent document CN 105911989A. For the analysis of the driving behavior, the driving behavior may be obtained by collecting vehicle control data, monitoring and capturing driver actions, and capturing external driving environments including external road conditions and weather environments, for example, a driving behavior analysis method disclosed in chinese patent document CN 108009475A. The posture of the driver can be captured by a camera device and recognized and classified by a classification algorithm, a human body motion classification method based on posture recognition is disclosed in chinese patent document CN106570480B, and a computer-implemented method for calculating the posture or shape of an articulated or deformable entity is disclosed in chinese patent document CN 103093453B.
The driving assistance system is intended to improve the driving safety of the automobile, but is liable to cause an accident if the assistance strategy is not properly made. In the assisted driving system, the establishment of the auxiliary strategy and the nodal events (i.e. triggering events) triggering the initiation of the auxiliary strategy are usually based on the set of driving experiences of designers, but the driving experiences are limited after all, and all situations cannot be dealt with in a complex driving environment, for example, if the driver fails to take deceleration or avoidance measures within a preset time when encountering an obstacle, the driver performs appropriate deceleration control, how the degree of deceleration control is determined, how the time when the deceleration measures are taken is determined, and more importantly, unlike a fully automatic driving system, the assisted driving system is an auxiliary measure for the driving behavior of the driver, and how the driver should adjust the auxiliary measure after the intervention control is performed at different time nodes during the implementation of the auxiliary measure; and if the steering lamp is not turned on when the driver turns, triggering an auxiliary driving strategy for turning on the steering lamp at a preset time point after the turning is started, but the preset trigger time point of the designer is not supported by reference data, is simply set by only depending on the driving experience of the designer, and the designer cannot reasonably evaluate the preset trigger time point. Therefore, such cases need to be refined in subsequent system improvements. In addition, the designer is unable to exhaust driving conditions that may affect safety, particularly the prediction of the driver's own non-driving behavior. For example, when a driver drives on a highway, the driving speed is first reduced, and the double flashing warning light is turned on. The reason for this is that the driver turns on the double-flashing warning light while finding that the vehicle ahead reduces the driving speed, takes speed reduction measures to avoid the vehicle distance from being too small, and turns on the double-flashing warning light to warn the following vehicle behind. This may cause a hazard if the driver finds the front vehicle to decelerate later, or even does not find the front vehicle to decelerate. But for this case the designer does not necessarily know it. In urban expressway with dense vehicles, the driver can find the falling liquid in the front windshield, first reduce the driving speed and move the wiper to clear the liquid on the windshield. However, if the driver does not clear the liquid in a timely manner or does not reduce the speed, there is a risk that the driver may be visually impaired or distracted. Therefore, in the subsequent process of improving the auxiliary driving system, similar events can be extracted through the analysis of the vehicle driving data, so that the events can be conveniently researched and clarified for subsequent research, and corresponding auxiliary driving strategies are formulated for the events possibly causing driving risks for supplementation.
Disclosure of Invention
Therefore, the invention aims to provide a big data analysis method and a big data analysis system for evaluating the driving assistance performance of an automobile, which are used for assisting designers to design and perfect an automobile driving assistance system.
In order to achieve the purpose, the automobile safety performance analysis and evaluation method based on the big data comprises the following steps:
inputting auxiliary driving strategy data to be evaluated, corresponding trigger event data for triggering the auxiliary driving strategy, vehicle running parameter data, environment data and behavior data of a driver into an evaluation model unit;
the evaluation model unit evaluates the driving assistance strategy to be evaluated and outputs an evaluation result.
The establishment of the evaluation model comprises the following steps: acquiring trigger event data, auxiliary driving strategy data, vehicle driving parameter data, environment data and operation behavior data of a driver, which trigger an auxiliary driving system to make an auxiliary driving strategy, evaluating the auxiliary driving strategy triggered each time and marking an evaluation result so as to establish a data set; the data set is divided into a training set and a test set, and the evaluation model is trained through the training set and tested through the test set.
Acquiring in-vehicle video data in the driving process, and dividing the video data into basic video data and classified video data; extracting a video frame from a video of the basic video data, and converting the video frame into a picture; carrying out image identification and classification on the pictures, and marking the classified picture categories displaying the safety behaviors so as to form an exclusion set;
extracting video frames from the video of the classified video data, and converting the video frames into pictures; and carrying out image identification and classification on the pictures, and excluding the same type of classification in the exclusion set to form a classification set.
Obtaining vehicle running parameter data at least including vehicle speed at each moment in the driving process, and judging t1The vehicle speed at the moment and the preset interval time Deltat1After t2Speed reduction difference v of vehicle speed at time1Whether the difference value is larger than or equal to a preset difference value delta v or not is judged, and if yes, t is judgediThe vehicle speed at the moment and the preset interval time Deltat1After ti+1Speed reduction difference v of vehicle speed at timeiWhether or not v is greater than or equal to1Wherein i is more than or equal to 2, if so, recursion is carried out in sequence until tnThe vehicle speed at the moment and the preset interval time Deltat1After tn+1Speed reduction difference v of vehicle speed at timenLess than v1Then the time t is1Time to tnThe deceleration event at the moment is recorded as a to-be-evaluated event for t1P seconds to t before timenAnd extracting the vehicle running parameter data, the environment data and the behavior data of the driver between m seconds after the moment to form the event data to be evaluated, wherein p and m are preset values.
With the average acceleration a of the vehicle in the event to be evaluated as the x-axis, t1Establishing a coordinate system for the y axis at the moment; drawing coordinate points of each event to be evaluated in a coordinate system; and classifying the coordinate points in the coordinate system by a k-nearest neighbor method, and extracting the event data to be evaluated corresponding to all the coordinate points of each category to form a data set.
Acquiring vehicle running parameter data at least comprising vehicle control item data and driver behavior data in the driving process, and judging whether the interval time between different vehicle control items or driver behaviors which continuously occur at least twice is less than preset interval time delta t2And if so, extracting the vehicle running parameter data, the environment data and the behavior data of the driver from h seconds before the first vehicle operation item or the driver behavior to k seconds after the last vehicle operation item or the driver behavior, wherein h and k are preset values.
The invention also provides an automobile safety performance analysis and evaluation system based on the big data, which comprises a data acquisition module and a data analysis module, wherein the data acquisition module is used for acquiring analysis data at least comprising the vehicle driving parameter data; the data analysis module is used for judging t1The vehicle speed at the moment and the preset interval time Deltat1After t2Speed reduction difference v of vehicle speed at time1Whether the difference value is larger than or equal to a preset difference value delta v or not is judged, and if yes, t is judgediThe vehicle speed at the moment and the preset interval time Deltat1After ti+1Speed reduction difference v of vehicle speed at timeiWhether or not v is greater than or equal to1Wherein i is more than or equal to 2, if so, recursion is carried out in sequence until tnThe vehicle speed at the moment and the preset interval time Deltat1After tn+1Speed reduction difference v of vehicle speed at timenLess than v1Recording the deceleration event as a to-be-evaluated event and comparing t1P seconds to t before timenAnd extracting the vehicle running parameter data, the environment data and the behavior data of the driver within m seconds after the moment.
The automobile safety performance analysis and evaluation system based on the big data comprises a data acquisition module, wherein the data acquisition module comprises an outside-automobile shooting and recording device, an inside-automobile shooting and recording device, a laser radar range finder and a vehicle-mounted control system; the vehicle exterior shooting device is used for shooting images outside the vehicle; the in-vehicle camera device is used for shooting in-vehicle images; the laser radar range finder is used for at least acquiring distance data of the road surface barrier; the vehicle-mounted control system is used for at least acquiring vehicle running parameter data.
The data acquisition module comprises an automobile driving simulation virtual field, wherein the automobile driving simulation virtual field comprises a real vehicle at least provided with an auxiliary driving system, an in-automobile shooting and recording device and a surrounding screen device for displaying a virtual driving scene.
The data acquisition module is including setting up the sensor subassembly on car steering wheel, the sensor subassembly includes a plurality of sensors of evenly distributed in circumference.
By adopting the technical scheme, the big data analysis and evaluation method and the system for the automobile safety performance based on the big data analyze the automobile driving data, so that the safety of the auxiliary driving strategy of the automobile is evaluated or the event information which cannot be matched with the auxiliary driving strategy is extracted, so that the reference is provided for designers, and the improvement of the auxiliary driving strategy of the automobile or the supplement formulation strategy of the event which cannot be matched with the auxiliary driving strategy are facilitated.
Drawings
Fig. 1 is a schematic diagram of an embodiment of classifying coordinate points in a coordinate system by a k-nearest neighbor method.
FIG. 2 is a diagram illustrating an example of excluding a focus in an embodiment of the present invention.
FIG. 3 is a diagram illustrating an example of a set of pictures in a classification set according to an embodiment of the invention.
Fig. 4 is a schematic diagram of the state of the vehicle A, B, C when the vehicle a attempts to cut-in.
FIG. 5 is a schematic diagram of state two of vehicle A, B, C when vehicle A attempts to cut-in.
Fig. 6 is a state three schematic diagram of the vehicle A, B, C when the vehicle a attempts to cut-in.
FIG. 7 is a state four schematic diagram of vehicle A, B, C when vehicle A attempts to cut-in.
FIG. 8 is a schematic diagram illustrating the state of the vehicle A, B, C when an assist driving maneuver is triggered, in accordance with one embodiment of the present invention.
FIG. 9 is a schematic diagram of a state two of the vehicle A, B, C when an assist driving maneuver is triggered in accordance with an embodiment of the present invention.
Detailed Description
The automobile assistant driving system sets a corresponding assistant driving strategy for the trigger event in a mode of listing the trigger event, and the corresponding assistant driving strategy can be triggered when the trigger event occurs in the driving process. Whether a triggering event occurs is determined by capturing corresponding vehicle driving data, which typically includes vehicle driving parameter data, environmental data, and driver behavior data. The vehicle running parameter data is parameters in the running process of the vehicle, and at least comprises vehicle speed data and various vehicle control item data, the vehicle control items at least comprise acceleration control, brake control, steering control and operation control of various facilities such as lamps, wipers and air conditioners, and the vehicle running parameter data can be obtained by extracting data of a vehicle-mounted computer system; the environment data comprises obstacle position and distance data, guideboard data, traffic light data, lane change data and the like, and is obtained by analyzing images captured by an external vehicle shooting device or an internal vehicle shooting device and analyzing data acquired by a laser radar range finder; the behavior data of the driver at least comprises arm posture data or eye movement data and the like, and can be obtained by capturing the image of the driver through a camera device in the vehicle for analysis and obtaining through an eye movement instrument.
The automobile assistant driving system is different from an automatic driving system, the automatic driving system automatically adjusts various driving parameters through judging road conditions, the assistant driving system does not participate in too much automobile control, otherwise, the driver can feel out of control of the automobile, and accidents are easily caused on the contrary. By analyzing a plurality of items of vehicle driving data, a more complete driving assistance strategy can be performed, but the complete driving assistance strategy needs to analyze various events which may need driving assistance in the driving process and finally serves as a trigger event for triggering the corresponding driving assistance strategy. Therefore, the vehicle driving data are analyzed to extract the event information to be evaluated, the event information to be evaluated can be conveniently researched subsequently, and a corresponding auxiliary driving strategy is formulated for the event to be evaluated which possibly causes driving danger to supplement. The invention is described in further detail below with reference to the figures and the detailed description.
The invention provides a big data-based automobile safety performance analysis and evaluation method, which comprises the steps of firstly obtaining vehicle driving data, and judging t1The vehicle speed at the moment and the preset interval time Deltat1After t2Speed reduction difference v of vehicle speed at time1Whether the difference value is larger than or equal to a preset difference value delta v or not is judged, and if yes, t is judgediThe vehicle speed at the moment and the preset interval time Deltat1After ti+1Speed reduction difference v of vehicle speed at timeiWhether or not v is greater than or equal to1Wherein i is more than or equal to 2, if so, recursion is carried out in sequence until tnThe vehicle speed at the moment and the preset interval time Deltat1After tn+1Speed reduction difference v of vehicle speed at timenLess than v1Then the time t is1Time to tnThe deceleration event at the moment is recorded as a to-be-evaluated event for t1P seconds to t before timenAnd extracting the vehicle running parameter data, the environment data and the behavior data of the driver between m seconds after the moment to form the event data to be evaluated, wherein p and m are preset values.
For example, a predetermined time Δ t1For 1s, the preset difference value delta v is 7km/h, the preset p value is 3, the m value is 4, the vehicle speed data are analyzed, and t is obtained1The vehicle speed at the moment is 87.7km/h, t2The speed of the vehicle at the moment is 80.1km/h, the deceleration difference v1At 7.6km/h, gives v1If the speed is larger than delta v, the vehicle is judged to have rapid deceleration, and t is further obtained2T at time interval 1s3The speed of the vehicle at the moment is 70.6km/h, and the deceleration difference v2Is 9.5km/h, v2Greater than v1Then get the sum t3T at time interval 1s4The speed of the vehicle at the moment is 65.6km/h, and the deceleration difference v3Is 5km/h, v3Less than v1If the vehicle is suddenly decelerated, the time t is determined1Time to t3The deceleration event at the moment is recorded as a to-be-evaluated event for t13 seconds to t before the moment3And extracting the vehicle driving data between 4 seconds after the moment to form the event data to be evaluated.
The rapid deceleration event existing in the driving process can be obtained through analyzing the vehicle speed data and used as an event to be evaluated, and the vehicle driving data related to the rapid deceleration event is extracted and used for analyzing and evaluating the next rapid deceleration event. For example, in the analysis process of a certain sudden deceleration event, the reason that the sudden deceleration event is caused is that a maintenance roadblock is arranged on the left lane of an expressway through an automobile running video shot by an automobile outer shooting device, the distance between a driver and the roadblock is 85m during emergency braking through a distance meter, the distance between the driver and the roadblock when the road block is detected by the distance meter is 235m, and the speed of the driver during braking can be 120km/h through vehicle running parameter data.
After acquiring the event data to be evaluated, establishing a coordinate system, wherein an x axis represents the average acceleration a of the vehicle in the event to be evaluated, wherein a = (v)1+v2+……vn-1）/（n×△t1-△t1) The y-axis represents t1Vehicle speed at the time; drawing coordinate points of each event to be evaluated in a coordinate system; and classifying the coordinate points in the coordinate system by a k-nearest neighbor method, and extracting the event data to be evaluated corresponding to all the coordinate points of each category to form a data set. As shown in fig. 1, in this embodiment, four regions including a high-risk region, a low-risk region, a high-speed medium-risk region, and a low-speed medium-risk region can be determined by classifying the coordinate points, and further analysis can be performed by using the event data to be evaluated corresponding to the coordinate points in each region. Of course, in other embodiments, the classification need not be solely for different sets of coordinate pointsThe four partitions described above are generated.
In another embodiment of the present invention, the method for analyzing and evaluating the safety performance of the automobile based on the big data further includes: judging whether the interval time between different vehicle control items which continuously occur at least twice is less than the preset interval time delta t2And if so, extracting the vehicle running parameter data, the environment data and the behavior data of the driver from h seconds before the first vehicle operation item to k seconds after the last vehicle operation item, wherein h and k are preset values.
For example, the preset interval time Δ t2The preset h value is 2s, the preset h value is 3, the preset k value is 3, two different vehicle control items which continuously occur are respectively brake and touch a double-flash warning lamp, the interval time between the two vehicle control items is 1.9s, the vehicle control event is recorded as a to-be-evaluated event, and the vehicle driving data from 3s before the first vehicle control item to 3s after the second vehicle control item is extracted to form the to-be-evaluated event data. The reason that the automobile running video shot by the outer vehicle shooting device is used for knowing that the secondary sudden deceleration event is caused is that a double-flash warning lamp of the front vehicle on the highway is lightened, the front vehicle is known to decelerate through a range finder, a driver is 66m away from the front vehicle during braking, the distance of the range finder when detecting the front vehicle to decelerate is 197m, the speed of the driver during braking can be known to be 108km/h through vehicle running parameter data, therefore, designers can perfect an assistant driving system in follow-up according to the event, a specific assistant driving strategy is formulated, the double-flash warning lamp of the front vehicle is monitored to be lightened, and the preset time interval delta t is arranged3The distance between the inner and the front vehicle is reduced by a distance difference Delta S exceeding a preset value1And then the driver is reminded to notice the road condition ahead as soon as possible. For example, setting Δ t3Is 1S,. DELTA.S1And when the distance between the two vehicles is 202m when the double-flash warning lamp of the vehicle in front is monitored to be turned on and the distance between the two vehicles is 176m after 1s, the auxiliary driving strategy is started to remind the driver.
As another example, the preset interval time Δ t2Is 5s, h is preset to be 5, k is preset to be 5, haveThe different vehicle actions or driver behaviors that occur in succession many times are: the driver's visual foothold is the right side rearview mirror, the 3s rear right turn light is on, the 2s rear vehicle changes to the right lane and the vehicle accelerates, the 3s rear brake brakes and the vehicle decelerates, and the 2s rear vehicle changes to the left lane. In the face of the continuous operation, the designer does not know what driving situation is at that time, but the reason why the continuous operation event is caused is known through the automobile running video shot by the external shooting device is as follows: as shown in fig. 4, when the driver drives the vehicle a on the left lane on the highway, the driver wants to overtake the vehicle B ahead, but there is another vehicle C on the right lane and in front of the vehicle B; the driver then looks at the right side rear view mirror, changes the vehicle to the right lane and accelerates the vehicle after determining that a lane change is possible, as shown in fig. 5; however, the distance between the vehicle B and the vehicle C is too small, so that the driver can judge that the vehicle cannot overtake, as shown in fig. 6; the driver then applies the brakes to slow the vehicle and change the vehicle back to the left lane as shown in fig. 7.
Aiming at the driving situation, designers improve an auxiliary driving system, establish an auxiliary driving strategy aiming at the situation, capture the eye movement data of a driver through an eye tracker, and judge whether a visual foot-drop point of the driver falls on a rear-view mirror (left side or right side) on one side when the driver drives a vehicle A through an auxiliary driving module of a vehicle-mounted computer system, and if so, judge whether the distance from the vehicle A to a front vehicle B is smaller than a preset vehicle distance Delta S2And a preset vehicle distance Delta S on the same side of the side rearview mirror and the vehicle A on the lane3Whether another vehicle C exists in the vehicle, the distance from the vehicle A to the vehicle C is larger than the distance from the vehicle A to the vehicle B, if yes, the driver is motivated to overtake the front vehicle, and the time delta t preset at the interval is predicted3Whether the rear vehicle C is located in front of the vehicle B with a vehicle distance S therebetweenBC PrepGreater than a predetermined vehicle distance Delta S4And if not, starting an auxiliary driving strategy to remind the driver of undersize overtaking distance. Estimated at a predetermined time Δ t3Vehicle distance S between inner vehicle C and vehicle BBC PrepBy the following steps: acquisition of t4Distance S between vehicle A and vehicle B at the timeAB primerDistance S between vehicle A and vehicle CAC primaryAs shown in fig. 8, the sum t is collected4Time interval preset time delta t4T of5Distance S between vehicle A and vehicle B at the timeAB terminalDistance S between vehicle A and vehicle CAC terminalAs shown in FIG. 9, t can be calculated4Distance S from vehicle B to vehicle C at timeBC initialIs SAC primaryAnd SAB primerDifference of (d), t5Distance S from vehicle B to vehicle C at timeBC terminalIs SAC terminalAnd SAB terminalThe relative average velocity v of the vehicle B and the vehicle C4Is (S)BC initial-SBC terminal）/△t4(ii) a Then, SBC Prep=SBC initial-△t3×v4。
For example, preset Δ S2Is 200m,. DELTA.S3Is 500m,. DELTA.t3Is 8S,. DELTA.S4Is 120m,. DELTA.t4For 1s, the driver's eye movement data is captured by an eye tracker, and when the driver is driving the vehicle A and his visual foothold is on the right side rear-view mirror, the time (t) is obtained by a distance meter4Time of day) from the preceding vehicle BAB primer185m, less than Δ S2Obtaining a distance S from another vehicle C on the right lane by a distance meterAC primary356m, less than Δ S3Distance S from vehicle B to vehicle CBC initialIs 171 m; further obtaining t after 1s by a distance meter5Distance S between vehicle A and vehicle B at timeAB terminal179m, distance S between vehicle A and vehicle CAC terminal341m, the distance S from the vehicle B to the vehicle CBC terminalIs 162 m; calculated to give v4Is 9m/S, SBC Prep99m, which is smaller than Δ S4And starting an auxiliary driving strategy to remind the driver of undersize passing distance.
For the designed driving assistance strategy, the data at the time of application can be acquired through actual driving or simulated driving, so that the data can be evaluated through establishing an evaluation model, which is described in detail below through another embodiment.
In another embodiment of the present invention, the method for analyzing and evaluating the safety performance of the automobile based on the big data further includes: the data acquisition module inputs auxiliary driving strategy data to be evaluated, corresponding trigger event data for triggering the auxiliary driving strategy, vehicle running parameter data, environment data and behavior data of a driver into the evaluation model unit; the evaluation model unit evaluates the driving assistance strategy to be evaluated and outputs an evaluation result.
The evaluation model in the evaluation model unit is a decision tree model, and the establishment of the evaluation model comprises the following steps: acquiring trigger event data, auxiliary driving strategy data, vehicle driving parameter data, environment data and operation behavior data of a driver, which trigger an auxiliary driving system to make an auxiliary driving strategy, evaluating the auxiliary driving strategy triggered each time and marking an evaluation result so as to establish a data set; the data set is divided into a training set and a test set, and the evaluation model is trained through the training set and tested through the test set.
The data set may be established uniformly for a plurality of triggering events or may be established for only one triggering event, for example, table 1 shows a vehicle driving data table for an auxiliary driving strategy triggered only after the speed limit indication guideboard is identified. By acquiring the vehicle driving related data in table 1, the assistant driving strategy triggered each time is evaluated and labeled, thereby forming a training data set. For example, the data set includes 15000 triggered assisted driving strategy related data, 14000 of which are divided as a training set, and the remaining 1000 of which are divided as a test set. And training the evaluation model through a decision tree algorithm. After the model is formed, the model may then be used to evaluate an assisted driving strategy to be evaluated.
Table 1:
| serial number | Speed of vehicle at time of triggering | Distance of guideboard | Speed limit Numerical value | Driver state | Time of early warning Spacer | Driver feedback control System time interval | Driver feedback Control mode | When passing by the guideboard Vehicle speed | Evaluation results |
| 1 | 72km/h | 340m | 60 | Normal driving | 2.2s | 3.5s | Speed reduction | 58km/h | Superior food |
| 2 | 45km/h | 166 m | 15 | Telephone | 0.8s | 4.1s | Speed reduction | 27km/h | Bad quality |
| 3 | 55km/h | 237 m | 30 | Normal driving | 2.5s | 5.0s | Speed reduction | 29km/h | Superior food |
| 4 | 38km/h | 188 m | 30 | Fatigue driving | 0.8s | 6.2s | Acceleration | 55km/h | Bad quality |
| 5 | 97 km/h | 95 m | 80 | Normal driving | 0.9s | 3.8s | Speed reduction | 85km/h | Bad quality |
| 6 | 125km/h | 298 m | 120 | Normal driving | 1.5s | 4.2s | Speed reduction | 118km/h | Superior food |
| 7 | 80km/h | 169 m | 60 | Mobile phone | 0.5s | ∞ | Without feedback | 80km/h | Bad quality |
| 8 | 75km/h | 360m | 60 | Normal driving | 2.3s | 3.8s | Speed reduction | 56km/h | Superior food |
| 9 | 49km/h | 176 m | 15 | Telephone | 0.9s | 4.3s | Speed reduction | 25km/h | Bad quality |
| 10 | 51km/h | 232 m | 30 | Normal driving | 2.8s | 5.2s | Speed reduction | 27km/h | Superior food |
| 11 | 36km/h | 189 m | 30 | Fatigue driving | 0.7s | 6.1s | Acceleration | 53km/h | Bad quality |
| 12 | 99 km/h | 65 m | 80 | Normal driving | 0.8s | 3.4s | Speed reduction | 88km/h | Bad quality |
| 13 | 121km/h | 307 m | 120 | Normal driving | 1.8s | 4.6s | Speed reduction | 112km/h | Superior food |
| 14 | 81km/h | 144 m | 60 | Mobile phone | 0.6s | ∞ | Without feedback | 81km/h | Bad quality |
| …… |
In another embodiment of the present invention, the method for analyzing and evaluating the safety performance of the automobile based on the big data further includes: acquiring in-vehicle video data in the driving process, and dividing the video data into basic video data and classified video data; extracting a video frame from a video of the basic video data, and converting the video frame into a picture; carrying out image identification and classification on the pictures through a CNN convolutional neural network algorithm, and marking the classified picture categories displaying the safety behaviors so as to form an exclusion set; extracting video frames from the video of the classified video data, and converting the video frames into pictures; and carrying out image identification and classification on the pictures, and excluding the same type of classification in the exclusion set to form a classification set.
Through collecting video data in the vehicle, including the behavior of a driver and the behavior of passengers, analyzing, identifying and classifying the images, data can be provided for designers to judge which types of non-operation behaviors of the driver or which types of behaviors of the passengers belong to dangerous behaviors, so that the auxiliary driving strategy is designed accordingly.
The collected video data is first divided into base video data and classified video data. Extracting a video frame from a video of the basic video data, and converting the video frame into a picture; the images are identified and classified through a CNN convolutional neural network algorithm, the classifications can be used for a designer to evaluate in advance and mark safe normal driving behaviors to form an exclusion set, and then the images of the same type in the exclusion set can be firstly discharged in the subsequent classification. That is, video frames are extracted from a video of classified video data, and the video frames are converted into pictures; and carrying out image identification and classification on the pictures, and excluding the same type of classification in the exclusion set to form a classification set. The formed classification set can be provided for designers to refer to. Taking fig. 2 as an example, after the pictures are classified, a designer can determine that the type of the pictures belongs to safe normal driving behaviors, so as to mark the pictures in an exclusion set, when a classified video is identified, for example, the pictures of the same type as that in fig. 2 are identified, the pictures are not marked and classified, and are directly excluded, while the pictures shown in fig. 3 are classified and stored in the classification set.
The invention also provides an automobile safety performance analysis and evaluation system based on the big data, which comprises a vehicle-mounted unit and an analysis unit, wherein the vehicle-mounted unit comprises a data acquisition module and a vehicle-mounted computer system, and the analysis unit comprises a data acquisition module and a data analysis module.
The data acquisition module at least comprises an outside-vehicle shooting device, an inside-vehicle shooting device, a laser radar range finder and an eye tracker; the vehicle exterior shooting device is used for shooting images outside the vehicle, the vehicle interior shooting device is used for shooting images inside the vehicle, and the vehicle interior shooting device can respectively comprise a plurality of cameras and collect images of various angles from different directions; the laser radar range finder is used for acquiring barrier distance data around the vehicle; the eye tracker is used for collecting eye movement data of a driver.
The vehicle-mounted computer system is used for receiving various data collected by the data collection module, analyzing the data to form behavior data of a driver at least comprising arm posture data or eye movement data, and forming environment data at least comprising obstacle position and distance data, guideboard data, traffic light data and lane change data; the system comprises a data processing unit, a data processing unit and a data processing unit, wherein the data processing unit is used for recording vehicle running parameter data generated in the vehicle driving process, at least comprises vehicle speed data and various vehicle operation item data, and the vehicle operation items at least comprise acceleration control, braking control, steering control and operation control of various facilities such as vehicle lamps, wipers, air conditioners and the like; the vehicle-mounted computer system also comprises an auxiliary driving module which is used for triggering a corresponding auxiliary driving strategy according to various data collected by the data collection module and vehicle driving parameter data generated in the vehicle driving process, wherein the auxiliary driving strategy at least comprises a corresponding prompt for a driver through sound and light media.
The data acquisition module is used for acquiring vehicle running parameter data, environment data and driver behavior data, and the data analysis module is used for realizing the big data analysis method for evaluating the auxiliary driving performance of the automobile.
In one embodiment of the invention, the data acquisition module is used for acquiring analysis data at least comprising vehicle driving parameter data; the data analysis module is used for judging t1The vehicle speed at the moment and the preset interval time Deltat1After t2Speed reduction difference v of vehicle speed at time1Whether the difference value is larger than or equal to a preset difference value delta v or not is judged, and if yes, t is judgediThe vehicle speed at the moment and the preset interval time Deltat1After ti+1Speed reduction difference v of vehicle speed at timeiWhether or not v is greater than or equal to1Wherein i is more than or equal to 2, if so, recursion is carried out in sequence until tnThe vehicle speed at the moment and the preset interval time Deltat1After tn+1Speed reduction difference v of vehicle speed at timenLess than v1Recording the deceleration event as a to-be-evaluated event and comparing t1P seconds to t before timenExtracting vehicle running parameter data, environment data and driver behavior data within m seconds after the moment; but also for taking the average acceleration a of the vehicle in the event to be evaluated as the x-axis, t1Establishing a coordinate system for the y axis at the moment; drawing coordinate points of each event to be evaluated in a coordinate system; and classifying the coordinate points in the coordinate system by a k-nearest neighbor method, and extracting the event data to be evaluated corresponding to all the coordinate points of each category to form a data set.
In addition, the data analysis module is further configured to determine whether an interval time between at least two consecutively occurring different vehicle maneuver items is less than a preset interval time Δ t2If so, the first vehicle will beAnd extracting vehicle running parameter data, environment data and driver behavior data between h seconds before the vehicle control item and k seconds after the last vehicle control item, wherein h and k are preset values.
The data analysis module is also used for inputting the auxiliary driving strategy data to be evaluated, the corresponding trigger event data for triggering the auxiliary driving strategy, the vehicle running parameter data, the environment data and the behavior data of the driver into the evaluation model unit; the evaluation model unit evaluates the driving assistance strategy to be evaluated and outputs an evaluation result.
And evaluating the auxiliary driving strategy triggered each time and marking an evaluation result by a designer so as to establish a data set. The data analysis module is also used for dividing the data set into a training set and a test set, training the evaluation model through the training set and testing through the test set. Preferably, the driving assistance strategy of the same type of trigger event is evaluated and the evaluation result is marked, so as to establish a data set, such as shown in table 1, in this case, the trigger event is the identification of the speed limit indicating guideboard.
The data analysis module is also used for dividing the acquired in-vehicle video data into basic video data and classified video data; extracting a video frame from a video of the basic video data, and converting the video frame into a picture; carrying out image identification and classification on the pictures, and marking the classified picture categories displaying the safety behaviors so as to form an exclusion set; extracting video frames from the video of the classified video data, and converting the video frames into pictures; and carrying out image identification and classification on the pictures, and excluding the same type of classification in the exclusion set to form a classification set.
The automobile safety performance analysis and evaluation system based on the big data comprises a data acquisition module, wherein the data acquisition module comprises an outside-automobile shooting and recording device, an inside-automobile shooting and recording device, a laser radar range finder and a vehicle-mounted control system; the vehicle exterior shooting device is used for shooting images outside the vehicle; the in-vehicle camera device is used for shooting in-vehicle images; the laser radar range finder is used for at least acquiring distance data of the road surface barrier; the vehicle-mounted control system is used for at least acquiring vehicle running parameter data.
In addition, vehicle driving data may also be obtained by way of simulated driving. In this manner, the data acquisition module includes an automobile driving simulation virtual ground, which includes a real vehicle having at least an auxiliary driving system, an in-vehicle camera device, and a surround screen device for displaying a virtual driving scene.
As another embodiment of the invention, the data acquisition module comprises a sensor assembly arranged on a steering wheel of an automobile, the sensor assembly comprises a plurality of sensors, such as pressure sensors, temperature sensors, touch sensors or other sensors, the plurality of sensors are uniformly distributed in the circumferential direction of the steering wheel, and the distance between the plurality of sensors can be set to be 5-10 cm. The vehicle-mounted control system is used for receiving data of the sensor and judging whether the left hand or the right hand is separated from the steering wheel according to the triggering condition of the sensor. For example, the sensors are divided into a left sensor and a right sensor according to the symmetry axis of the steering wheel, and when either of the left sensors is activated, it is determined that the left hand is held on the steering wheel, and when either of the right sensors is activated, it is determined that the right hand is held on the steering wheel. The duration of holding the hand on the steering wheel can be obtained according to the triggering duration of the sensor.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (7)
1. A big data-based automobile safety performance analysis and evaluation method is characterized by comprising the following steps: obtaining vehicle running parameter data at least including vehicle speed at each moment in the driving process, and judging t1The vehicle speed at the moment and the preset interval time Deltat1After t2Speed reduction difference v of vehicle speed at time1Whether the difference value is larger than or equal to a preset difference value delta v or not is judged, and if yes, t is judgediThe vehicle speed at the moment andspaced by a predetermined time Δ t1After ti+1Speed reduction difference v of vehicle speed at timeiWhether or not v is greater than or equal to1Wherein i is more than or equal to 2, if so, recursion is carried out in sequence until tnThe vehicle speed at the moment and the preset interval time Deltat1After tn+1Speed reduction difference v of vehicle speed at timenLess than v1Then the time t is1Time to tnThe deceleration event at the moment is recorded as a to-be-evaluated event for t1P seconds to t before timenAnd extracting the vehicle running parameter data, the environment data and the behavior data of the driver between m seconds after the moment to form the event data to be evaluated, wherein p and m are preset values.
2. The big-data-based automobile safety performance analysis and evaluation method according to claim 1, wherein: with the average acceleration a of the vehicle in the event to be evaluated as the x-axis, t1Establishing a coordinate system for the y axis at the moment; drawing coordinate points of each event to be evaluated in a coordinate system; and classifying the coordinate points in the coordinate system by a k-nearest neighbor method, and extracting the event data to be evaluated corresponding to all the coordinate points of each category to form a data set.
3. The big-data-based automobile safety performance analysis and evaluation method according to claim 1, wherein: acquiring vehicle running parameter data at least comprising vehicle control item data and driver behavior data in the driving process, and judging whether the interval time between different vehicle control items or driver behaviors which continuously occur at least twice is less than preset interval time delta t2And if so, extracting the vehicle running parameter data, the environment data and the behavior data of the driver from h seconds before the first vehicle operation item or the driver behavior to k seconds after the last vehicle operation item or the driver behavior, wherein h and k are preset values.
4. The utility model provides an automobile safety performance analysis evaluation system based on big data which characterized in that: comprises a data acquisition moduleThe data acquisition module is used for at least acquiring vehicle running parameter data; the data analysis module is used for judging t1The vehicle speed at the moment and the preset interval time Deltat1After t2Speed reduction difference v of vehicle speed at time1Whether the difference value is larger than or equal to a preset difference value delta v or not is judged, and if yes, t is judgediThe vehicle speed at the moment and the preset interval time Deltat1After ti+1Speed reduction difference v of vehicle speed at timeiWhether or not v is greater than or equal to1Wherein i is more than or equal to 2, if so, recursion is carried out in sequence until tnThe vehicle speed at the moment and the preset interval time Deltat1After tn+1Speed reduction difference v of vehicle speed at timenLess than v1Recording the deceleration event as a to-be-evaluated event and comparing t1P seconds to t before timenAnd extracting the vehicle running parameter data, the environment data and the behavior data of the driver within m seconds after the moment.
5. The big-data-based automobile safety performance analysis and evaluation system according to claim 4, wherein: the automobile safety performance analysis and evaluation system based on big data comprises a data acquisition module and a vehicle-mounted computer system, wherein the data acquisition module comprises an outside-automobile shooting and recording device, an inside-automobile shooting and recording device, a laser radar distance meter and an eye tracker; the vehicle exterior shooting device is used for shooting images outside the vehicle; the in-vehicle camera device is used for shooting in-vehicle images; the laser radar range finder is used for acquiring barrier distance data around the vehicle; the eye tracker is used for collecting eye movement data of a driver.
6. The big-data-based automobile safety performance analysis and evaluation system according to claim 5, wherein: the data acquisition module comprises an automobile driving simulation virtual field, wherein the automobile driving simulation virtual field comprises a real vehicle at least provided with an auxiliary driving system, an in-automobile shooting and recording device and a surrounding screen device for displaying a virtual driving scene.
7. The big-data-based automobile safety performance analysis and evaluation system according to claim 5, wherein: the data acquisition module is including setting up the sensor subassembly on car steering wheel, the sensor subassembly includes a plurality of sensors of evenly distributed in circumference.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010134138.6A CN111231971B (en) | 2020-03-02 | 2020-03-02 | Automobile safety performance analysis and evaluation method and system based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010134138.6A CN111231971B (en) | 2020-03-02 | 2020-03-02 | Automobile safety performance analysis and evaluation method and system based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111231971A CN111231971A (en) | 2020-06-05 |
| CN111231971B true CN111231971B (en) | 2021-04-30 |
Family
ID=70868186
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010134138.6A Active CN111231971B (en) | 2020-03-02 | 2020-03-02 | Automobile safety performance analysis and evaluation method and system based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111231971B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111811835A (en) * | 2020-07-14 | 2020-10-23 | 中国第一汽车股份有限公司 | Intelligent networked automobile open road test scene fault data automatic processing method |
| CN112677983B (en) * | 2021-01-07 | 2022-04-12 | 浙江大学 | System for recognizing driving style of driver |
| CN113200048B (en) * | 2021-05-28 | 2023-02-28 | 东风汽车集团股份有限公司 | Intelligent driving-assistant vehicle monitoring method and system |
| CN114595597B (en) * | 2022-05-09 | 2022-08-09 | 天津大学 | Vehicle performance comprehensive evaluation method and system based on big data |
| CN114596713B (en) * | 2022-05-09 | 2022-08-09 | 天津大学 | Real-time remote monitoring control method and system for vehicle |
| CN115578665B (en) * | 2022-09-07 | 2023-09-29 | 台州动产质押金融服务有限公司 | Resource processing method and system based on vehicle driving data |
| CN116660650A (en) * | 2023-05-17 | 2023-08-29 | 襄阳达安汽车检测中心有限公司 | Method, device, equipment and medium for testing detection delay of low-speed driving auxiliary system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784587A (en) * | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | A kind of driving behavior evaluation system |
| CN107784251A (en) * | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | The method evaluated based on image recognition technology driving behavior |
| CN108091139A (en) * | 2018-01-23 | 2018-05-29 | 余绍志 | A kind of traffic safety automatic evaluation system based on big data |
| CN108765627A (en) * | 2018-04-12 | 2018-11-06 | 深圳市拓保软件有限公司 | A kind of method of driving data risk quantification |
| CN110263836A (en) * | 2019-06-13 | 2019-09-20 | 南京师范大学 | A kind of bad steering state identification method based on multiple features convolutional neural networks |
| CN110688729A (en) * | 2019-08-26 | 2020-01-14 | 南京航空航天大学 | LSTM-IDM (least squares inverse discrete cosine transform) -following characteristic fusion method based on adaptive Kalman filtering, storage medium and equipment |
| CN110706517A (en) * | 2019-11-07 | 2020-01-17 | 深圳鼎然信息科技有限公司 | Intelligent safety early warning method, device and system for vehicle and storage medium |
-
2020
- 2020-03-02 CN CN202010134138.6A patent/CN111231971B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784587A (en) * | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | A kind of driving behavior evaluation system |
| CN107784251A (en) * | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | The method evaluated based on image recognition technology driving behavior |
| CN108091139A (en) * | 2018-01-23 | 2018-05-29 | 余绍志 | A kind of traffic safety automatic evaluation system based on big data |
| CN108765627A (en) * | 2018-04-12 | 2018-11-06 | 深圳市拓保软件有限公司 | A kind of method of driving data risk quantification |
| CN110263836A (en) * | 2019-06-13 | 2019-09-20 | 南京师范大学 | A kind of bad steering state identification method based on multiple features convolutional neural networks |
| CN110688729A (en) * | 2019-08-26 | 2020-01-14 | 南京航空航天大学 | LSTM-IDM (least squares inverse discrete cosine transform) -following characteristic fusion method based on adaptive Kalman filtering, storage medium and equipment |
| CN110706517A (en) * | 2019-11-07 | 2020-01-17 | 深圳鼎然信息科技有限公司 | Intelligent safety early warning method, device and system for vehicle and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111231971A (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111231971B (en) | Automobile safety performance analysis and evaluation method and system based on big data | |
| JP7263233B2 (en) | Method, system and program for detecting vehicle collision | |
| CN111137284B (en) | Early warning method and early warning device based on driving distraction state | |
| CN113487874B (en) | System and method for collecting, identifying and classifying following behavior scene data | |
| JP5691237B2 (en) | Driving assistance device | |
| CN113635897B (en) | Safe driving early warning method based on risk field | |
| JP2022532941A (en) | Devices and methods for processing vehicle signals to calculate behavioral risk measures | |
| CN115027484A (en) | Human-computer fusion perception method for high-degree automatic driving | |
| CN112874515B (en) | System and method for carrying out safety reminding on driving assistance system by using driving posture | |
| CN112606831A (en) | Anti-collision warning information external interaction method and system for passenger car | |
| CN109987090A (en) | Driving assistance system and method | |
| CN113570747A (en) | Driving safety monitoring system and method based on big data analysis | |
| CN115366907A (en) | Driver state abnormity prompting method and device, vehicle and storage medium | |
| KR102518662B1 (en) | Driving safety control system in use with around noise and the method of it | |
| JP2007334554A (en) | Obstacle detection device and obstacle detection method | |
| CN113264042B (en) | Hidden danger situation warning | |
| CN108515967B (en) | Anti-collision early warning method and device based on license plate recognition | |
| CN114274954A (en) | Vehicle active pipe connecting system and method combining vehicle internal and external sensing | |
| JP7427078B2 (en) | Method, driving assistance system and vehicle for avoiding collisions in road traffic based on adaptive setting of stay area | |
| KR20150066308A (en) | Apparatus and method for determining driving condition of deiver | |
| CN113370973B (en) | Forward collision early warning algorithm considering prediction of driving intention of front vehicle | |
| CN114103966A (en) | Control method, device and system for driving assistance | |
| CN113920734A (en) | Lane change early warning method based on logistic model | |
| JP6992285B2 (en) | Driving support device | |
| CN114999134B (en) | Driving behavior early warning method, device and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |