CN117022323A - Intelligent driving vehicle behavior analysis and prediction system and method - Google Patents
Intelligent driving vehicle behavior analysis and prediction system and method Download PDFInfo
- Publication number
- CN117022323A CN117022323A CN202311060538.7A CN202311060538A CN117022323A CN 117022323 A CN117022323 A CN 117022323A CN 202311060538 A CN202311060538 A CN 202311060538A CN 117022323 A CN117022323 A CN 117022323A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- accident
- data
- machine learning
- training
- 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.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 33
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000012549 training Methods 0.000 claims abstract description 87
- 238000010801 machine learning Methods 0.000 claims abstract description 77
- 230000006399 behavior Effects 0.000 claims abstract description 41
- 206010039203 Road traffic accident Diseases 0.000 claims abstract description 32
- 230000006870 function Effects 0.000 claims description 21
- 230000001133 acceleration Effects 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 12
- 238000011156 evaluation Methods 0.000 claims description 8
- 238000007405 data analysis Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000006378 damage Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000035484 reaction time Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- B60W60/00—Drive control systems specially adapted for autonomous road vehicles
- B60W60/001—Planning or execution of driving tasks
- B60W60/0015—Planning or execution of driving tasks specially adapted for safety
- B60W60/0016—Planning or execution of driving tasks specially adapted for safety of the vehicle or its occupants
-
- 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
- B60W40/02—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 ambient conditions
- B60W40/04—Traffic conditions
-
- 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
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W50/08—Interaction between the driver and the control system
- B60W50/14—Means for informing the driver, warning the driver or prompting a driver intervention
-
- 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
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W50/08—Interaction between the driver and the control system
- B60W50/14—Means for informing the driver, warning the driver or prompting a driver intervention
- B60W2050/143—Alarm means
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention belongs to the technical field of vehicle behavior analysis and prediction, and discloses an intelligent driving vehicle behavior analysis and prediction system and method, comprising the following steps: collecting vehicle type characteristic data; training a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle; collecting accident training data; generating an accident assessment value based on the accident training data; training a second machine learning model for predicting the type of traffic accident occurring in surrounding vehicles in real time based on the accident assessment value, the accident training data and the output of the first machine learning model; a driving adjustment instruction is generated based on an output of the second machine learning model.
Description
Technical Field
The invention relates to the technical field of vehicle behavior analysis and prediction, in particular to an intelligent driving vehicle behavior analysis and prediction system and method.
Background
The intelligent driving vehicle is a vehicle capable of realizing autonomous road running and interaction without human intervention through a self-carried sensing, decision and control system.
The Chinese patent of the grant bulletin number CN107862862B discloses a vehicle behavior analysis method and device, which can determine the behavior information of a vehicle to be analyzed in a preset time according to the stay time between checkpoints, track the vehicle to be analyzed according to the behavior information, and improve the accuracy of tracking the vehicle to be analyzed.
However, the residence time between the checkpoints cannot precisely reflect the behavior of the vehicle in running, cannot predict the temporary behavior of other vehicles and perform safety evaluation, cannot specifically analyze different types of vehicles, and cannot effectively reduce the probability of occurrence of road traffic accidents.
In view of the above, the present invention provides an intelligent driving vehicle behavior analysis and prediction system and method to solve the above-mentioned problems.
Disclosure of Invention
In order to overcome the above-mentioned drawbacks of the prior art, embodiments of the present invention provide an intelligent driving vehicle behavior analysis and prediction system and method.
In order to achieve the above purpose, the present invention provides the following technical solutions: an intelligent driving vehicle behavior analysis and prediction method, the method comprising:
collecting vehicle type characteristic data;
training a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle;
collecting accident training data;
generating an accident assessment value based on the accident training data;
training a second machine learning model for predicting traffic accident behaviors of surrounding vehicles in real time based on the accident assessment value, the accident training data and the output of the first machine learning model;
a driving adjustment instruction is generated based on an output of the second machine learning model.
An intelligent driving vehicle behavior analysis and prediction system, comprising:
the first data acquisition module is used for acquiring vehicle type characteristic data;
the second data acquisition module is used for acquiring accident training data;
the data analysis module generates an accident assessment value based on the accident training data;
the model training module trains a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle; training a second machine learning model for predicting traffic accident behaviors of surrounding vehicles in real time based on the accident assessment value, the accident training data and the output of the first machine learning model;
the control module generates a driving adjustment instruction and a danger announcement instruction based on the output of the second machine learning model;
and the cloud notification module is used for sending an early warning notification to surrounding vehicles based on the danger announcement instruction.
An electronic device, comprising: a processor and a memory, wherein the memory stores a computer program for the processor to call; the processor executes the intelligent driving vehicle behavior analysis and prediction system and method by calling the computer program stored in the memory.
A computer readable storage medium storing instructions that, when executed on a computer, cause the computer to perform an intelligent driving vehicle behavior analysis and prediction system and method as described above.
The intelligent driving vehicle behavior analysis and prediction system and the intelligent driving vehicle behavior analysis and prediction method have the technical effects and advantages that:
collecting vehicle type characteristic data; training a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle; collecting accident training data; generating an accident assessment value based on the accident training data; training a second machine learning model for predicting the type of traffic accident that may occur to surrounding vehicles in real time based on the accident assessment value, the accident training data, and the output of the first machine learning model; generating a driving adjustment instruction based on the output of the second machine learning model; the method and the device realize the real-time identification of the types of surrounding vehicles in driving, carry out accident assessment according to different vehicle types, predict the most probable behaviors of the surrounding vehicles when traffic accidents are possible, and carry out early warning in advance, thereby reducing the operation requirements of drivers, improving the dangerous identification capability of the drivers and reducing the probability of road safety accidents.
Drawings
FIG. 1 is a schematic diagram of an intelligent driving vehicle behavior analysis and prediction system according to the present invention;
FIG. 2 is a schematic diagram of an intelligent driving vehicle behavior analysis and prediction method according to the present invention;
FIG. 3 is a schematic diagram of an electronic device according to the present invention;
fig. 4 is a schematic diagram of an intelligent driving vehicle behavior analysis and prediction system according to embodiment 2 of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
Referring to fig. 1, the intelligent driving vehicle behavior analysis and prediction system according to the present embodiment includes a first data acquisition module 1, a second data acquisition module 2, a model training module 3, a data analysis module 4, and a control module 5, where the modules are connected by a wired and/or wireless network.
The first data acquisition module 1 is used for acquiring vehicle type characteristic data.
The vehicle type characteristic data comprise vehicle dynamic video data and vehicle three-dimensional point cloud data, wherein the vehicle dynamic video data are obtained by collecting video data when vehicles appear around a test vehicle in real time through a camera when the test vehicle runs on a road, and the vehicle three-dimensional point cloud data are three-dimensional point cloud data of the vehicles appearing around the test vehicle; the three-dimensional point cloud data of the vehicle are obtained in real time through a laser radar installed on the test vehicle;
the vehicle dynamic video data and the vehicle three-dimensional point cloud data are synchronously collected by the test vehicle, so that the vehicle type appearing in the vehicle dynamic video data and the vehicle type in the vehicle three-dimensional point cloud data are in one-to-one correspondence;
the vehicle dynamic video data may provide visual characteristics of the vehicle appearance, shape, color, etc., as well as movement information of the vehicle on the road. By analyzing the video image data, vehicle detection, localization, and classification can be performed.
A point cloud refers to data consisting of a large number of discrete three-dimensional points. In the fields of three-dimensional modeling, computer vision, robots and the like, point clouds are generally used for representing information such as shape, position, attitude and the like of an object; three-dimensional point cloud data are generally obtained by means of three-dimensional laser scanning, camera capturing or sensor acquisition and the like; three-dimensional point cloud data consists of a series of points with spatial coordinates, each of which contains the position of the point and possibly other attribute information such as color, normal vector, reflectivity, etc.
The vehicle three-dimensional point cloud data can provide geometric information of the spatial location, shape, and surrounding environment of the vehicle. By analyzing the vehicle three-dimensional point cloud data, vehicle detection and classification can be performed.
Meanwhile, vehicle dynamic video data and vehicle three-dimensional point cloud data are used, so that more comprehensive and rich vehicle information can be obtained, more geometric features are provided, and the accuracy and the robustness of vehicle type identification are improved.
Robustness refers to stability and reliability in the face of uncertainty, interference, or accidents. For example, a vehicle may be partially occluded by other vehicles, obstacles, or obstructions in the actual scene resulting in a deformation of the vehicle shape in the vehicle type feature data; the more robust the vehicle is, the more accurately the vehicle is classified as partially occluded or deformed.
The first data acquisition module 1 transmits the vehicle type characteristic data to the model training module 3
The model training module 3 trains a first machine learning model for identifying the vehicle type in real time based on the vehicle type characteristic data;
the training mode of the first machine learning model is as follows: manually setting a tag for a real-time group of vehicle type feature data, wherein the tag is a positive integer and represents a vehicle type; illustratively, set motorcycle to 1, car to 2, and truck to 3; the labels corresponding to each group of vehicle type characteristic data and each group of vehicle type characteristic data are constructed into one sample, and a plurality of samples are collected to construct a machine learning data set. The data set is divided into a training set, a verification set and a test set, wherein the training set accounts for 70% of the data set, and the verification set and the test set each account for 15% of the data set.
Taking the training set as input of a first machine learning model, taking a tag of a vehicle type as output, taking a tag corresponding to real-time vehicle type characteristic data as a prediction target, and taking a minimized machine learning model loss function as a training target; and stopping training when the first machine learning model loss function is smaller than or equal to the target loss value.
The first machine learning model loss function may be a Mean Square Error (MSE) or a Cross Entropy (CE);
exemplary, mean Square Error (MSE) is determined by applying a loss functionThe model is trained for the purpose of minimization, so that the machine learning model is better fitted with data, and the performance and accuracy of the model are improved; mse in the loss function is a loss function value, and i is a vehicle type characteristic data group number; u is the number of the characteristic data sets of the vehicle type; y is i Tag corresponding to the i-th group of vehicle type characteristic data, < > for the i-th group of vehicle type characteristic data>Tags predicted for the i-th set of real-time vehicle type characteristic data.
It should be noted that, the first machine learning model is a neural network model, and other model parameters of the first machine learning model, such as a target loss value, a depth of the network model, the number of neurons in each layer, an activation function used by the network model, and optimization of the loss function, are all realized through actual engineering, and are obtained after experimental tuning is continuously performed;
after the first machine learning model is trained, vehicle types around the vehicle can be identified in real time, and it can be understood that acceleration of different vehicle types is inconsistent with the highest speed per hour and the like, and the degree and reason of occurrence of road traffic accidents are different.
The second data acquisition module 2 is used for acquiring accident training data.
The accident training data comprises the type of the hit-vehicle, the acceleration of the hit-vehicle, the maximum speed of the hit-vehicle and the distance of the hit-vehicle from the driving vehicle n seconds before the traffic accident occurs, when several kinds of traffic accidents occur.
It should be noted that different types of vehicles have different characteristics and potential risks in traffic accidents. For example, large trucks may cause more serious damage in collisions, while motorcycles may lose control more easily due to poor stability. Therefore, the type of hit vehicle needs to be considered when considering the potential impact of traffic accidents.
Acceleration of the hit vehicle reflects the change in speed of the vehicle per unit time. If the acceleration of the hit vehicle is high, meaning that it may approach other vehicles or pedestrians at a faster rate, the probability of an accident increases. In addition, acceleration can also affect the braking ability of the vehicle in an emergency, and high acceleration can lead to increased braking distances and increased accident risk.
The maximum speed of a hit vehicle is also one of the key factors for the occurrence of an accident. If the hit vehicle is traveling at a high speed, the impact force is greater in the event of a collision with another vehicle or pedestrian, causing more serious injury or damage. In addition, the high-speed running can influence the perception and reaction time of a driver to road conditions, so that the driver is more difficult to control the vehicle, and the risk of accidents is increased.
The distance between the onset vehicle and the driving vehicle directly influences the occurrence of the accident. A closer distance reduces the reaction time of driving the vehicle, especially in an emergency situation, resulting in a failure to avoid a collision in time. In addition, the proper holding distance may provide more operating space and time for the driver to avoid accidents.
The accident training data collection environment includes, but is not limited to, one of a historical traffic accident profile based, an Event Data Recorder (EDR) or a simulated collision experiment.
The accident training data further includes a number of sets of types of surrounding vehicles without traffic accidents, accelerations of the surrounding vehicles, maximum speeds of the surrounding vehicles, and distances of the surrounding vehicles from the driving vehicle.
It is to be understood that the hit vehicle is also part of the surrounding vehicles before the traffic accident, and will therefore hereinafter be collectively referred to as surrounding vehicles.
The acceleration of the accident training data is obtained by an accelerometer or an on-board radar in real time, the distance and the maximum speed in the accident training data are all obtained by the on-board radar in real time, and the data are stored in a vehicle storage system, and after the accident happens, a traffic accident handling department manager reads the type of surrounding vehicles, the acceleration of the surrounding vehicles, the maximum speed of the surrounding vehicles and the distance between the surrounding vehicles and driving vehicles from the on-board storage system.
The second data acquisition module 2 sends the incident training data to the data analysis module 4,
the data analysis module 4 generates an accident assessment value based on the accident training data, and the calculation formula of the accident assessment value is as follows:
in the formula of danger r An accident evaluation value for the r-th vehicle in the surrounding vehicles, A r For the vehicle type influence coefficient of the r-th vehicle in the surrounding vehicles, the vehicle type influence coefficient of each vehicle type is a fixed value, and is composed ofThe staff comprehensively sets according to the volumes and the weights of different vehicle types, A r The value range is 1-10, the larger the vehicle is, the larger the weight is, because the speed reduction and steering are difficult, the probability of traffic accidents is greatly increased, and the A is r The larger the danger r The greater the value will be, the greater the volume and weight of an exemplary truck of one type, the fixed set at 8, the lighter the volume and weight of a car of one type, the fixed set at 5, the staff setting the vehicle type influence coefficients for several types of vehicles;
a r for acceleration of the r-th vehicle of the surrounding vehicles, V r For the current maximum speed of the r-th vehicle in the surrounding vehicles, d r The distance between the r-th vehicle and the own vehicle in the surrounding vehicles is set; a, a r The larger V r The larger d r Smaller indicates a greater possibility of traffic accidents, and danger r The larger the value, the danger r The value reflects the degree of risk during the current vehicle travel.
Setting an accident assessment threshold danger_go, when danger r <When the danger_go is not operated, the safety driving state is marked; when danger r And when the accident evaluation value is not less than the danger_go, marking the accident evaluation value as dangerous driving state, and sending the accident evaluation value to the model training module 3 by the data analysis module 4.
The accident assessment threshold danger_go is set by traffic related staff according to the historical accident occurrence condition, and is exemplified by no speed limit in a region, the accident assessment threshold is set lower if the traffic accident is frequent, and the dangerous driving state is judged to be sensitive and timely at the moment, otherwise, the accident assessment threshold is opposite.
The model training module 3 trains a second machine learning model that predicts in real time the type of traffic accident that is likely to occur in the surrounding x-th type of vehicle, based on the accident assessment value, the accident training data, and the output of the first machine learning model.
The training mode of the second machine learning model is as follows:
taking each group of accident training data and corresponding accident assessment values as characteristic data, taking a plurality of accident type numbers and taking the accident type number corresponding to each group of characteristic data as a label, wherein if the rear-end accident number is 1, the label is set to be 1, the left-turn overtaking collision accident number is 2, and the label is set to be 2;
taking the characteristic data as input of a second machine learning model, wherein the second machine learning model takes an accident type number as output, takes a label corresponding to the real-time characteristic data as a prediction target, and takes a minimized machine learning model loss function as a training target; and stopping training when the loss function of the second machine learning model is smaller than or equal to the target loss value.
The loss function of the second machine learning model is consistent with that of the first machine learning model, and other model parameters of the second machine learning model are realized through actual engineering, and are obtained after experimental tuning is continuously carried out.
The second machine learning model is used for rapidly analyzing the most probable behaviors of surrounding vehicles according to the types of the surrounding vehicles identified by the first machine learning model when dangerous driving states possibly occur, and reducing the probability of traffic accidents.
The model training module 3 sends the output of the second machine learning model to the control module 5;
the control module 5 generates a driving adjustment instruction based on the output of the second machine learning model. The driving adjustment instruction comprises voice reminding of a driver, intelligent speed reduction of the vehicle and other operations.
The embodiment 1 realizes the real-time identification of the types of surrounding vehicles in driving, carries out accident assessment according to different vehicle types, predicts the most probable behavior of the surrounding vehicles when traffic accidents are possible, and carries out early warning in advance, thereby reducing the probability of road safety accidents, reducing the operation requirement of drivers and improving the capability of the drivers to identify dangers.
Example 2
Referring to fig. 4, the detailed description of the embodiment 1 is omitted, and an intelligent driving vehicle behavior analysis and prediction system is provided, which includes: a control module 5 and a cloud notification module 6;
after the control module 5 generates the driving adjustment instruction, the danger announcement instruction is also generated and sent to the cloud notification module 6.
The cloud notification module 6 operates based on hazard advertisement instructions including: the method includes the steps that early warning notices are sent to all vehicle-mounted central control devices of surrounding vehicles in a wireless network mode, the early warning notices inform the types of possible accidents to occur, early warning and evacuation are carried out in advance, the probability of traffic accidents is further reduced, and serious traffic accidents caused by continuous collision are avoided.
Example 3
Referring to fig. 2, the detailed description of the embodiment is not shown in the description of embodiment 1, and an intelligent driving vehicle behavior analysis and prediction method is provided. The method comprises the following steps:
collecting vehicle type characteristic data;
training a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle;
collecting accident training data;
generating an accident assessment value based on the accident training data;
training a second machine learning model for predicting the type of traffic accident occurring in surrounding vehicles in real time based on the accident assessment value, the accident training data and the output of the first machine learning model;
a driving adjustment instruction is generated based on an output of the second machine learning model.
Further, the vehicle type feature data comprises vehicle dynamic video data and vehicle three-dimensional point cloud data, and the vehicle dynamic video data and the vehicle three-dimensional point cloud data are synchronously collected for the test vehicle.
Further, the training manner of the first machine learning model is as follows:
setting a tag for a real-time set of vehicle type feature data, wherein the tag is a positive integer and represents a vehicle type; constructing labels corresponding to each group of vehicle type characteristic data and each group of vehicle type characteristic data into one sample, and collecting a plurality of samples to construct a machine learning data set; the data set is divided into a training set, a verification set and a test set, wherein the training set accounts for 70% of the data set, and the verification set and the test set respectively account for 15% of the data set;
taking the training set as input of a first machine learning model, wherein the first machine learning model takes a tag of a vehicle type as output, takes a tag corresponding to real-time vehicle type characteristic data as a prediction target, and minimizes a machine learning model loss function valueAs a training target; mse in the loss function is a loss function value, and i is a vehicle type characteristic data group number; u is the number of the characteristic data sets of the vehicle type; y is i Tag corresponding to the i-th group of vehicle type characteristic data, < > for the i-th group of vehicle type characteristic data>A tag predicted for the i-th set of real-time vehicle type feature data; stopping training when the loss function of the first machine learning model is smaller than or equal to a preset target loss value; the first machine learning model is a neural network model.
Further, the accident training data comprises the types of surrounding vehicles, the acceleration of the surrounding vehicles and the maximum speed of the surrounding vehicles when a plurality of kinds of traffic accidents occur, and the accident training data also comprises the distance between the surrounding vehicles and the driving vehicles n seconds before the traffic accidents occur;
the accident training data further includes a number of sets of types of surrounding vehicles without traffic accidents, accelerations of the surrounding vehicles, maximum speeds of the surrounding vehicles, and distances of the surrounding vehicles from the driving vehicle.
Further, the calculation formula of the accident evaluation value is as follows:in the formula of danger r An accident evaluation value for the r-th vehicle in the surrounding vehicles, A r A is the vehicle type influence coefficient of the r-th vehicle in the surrounding vehicles r For acceleration of the r-th vehicle of the surrounding vehicles, V r For the current maximum speed of the r-th vehicle in the surrounding vehicles, d r Is the (r) vehicle and the host in the surrounding vehiclesA vehicle distance;
setting an accident assessment threshold danger_go, when danger r <When the danger_go is, marking as a safe driving state; when danger r And marking dangerous driving states when the traffic is not less than danger_go.
Further, the training manner of the second machine learning model is as follows:
taking each group of accident training data and corresponding accident assessment values as characteristic data, and taking a plurality of accident type numbers as labels, wherein each group of accident type numbers corresponds to each group of characteristic data;
taking the characteristic data as input of a second machine learning model, wherein the second machine learning model takes an accident type number as output, takes a label corresponding to the real-time characteristic data as a prediction target, and takes a minimized machine learning model loss function value as a training target; and stopping training when the loss function of the second machine learning model is smaller than or equal to the target loss value.
Further, the driving adjustment instruction comprises voice reminding of a driver and intelligent deceleration operation of the vehicle;
after the driving adjustment instruction is generated, a danger announcement instruction is generated; the danger announcement instruction comprises the steps of sending an early warning notice to all surrounding vehicle-mounted central control devices, informing about the type of the impending accident, and early warning and evacuation.
Example 4
Referring to fig. 3, an electronic device according to an exemplary embodiment includes: a processor and a memory, wherein the memory stores a computer program for the processor to call;
the processor executes the intelligent driving vehicle behavior analysis and prediction method by calling the computer program stored in the memory.
Example 5
A computer readable storage medium having stored thereon a computer program that is erasable according to an exemplary embodiment is shown; the computer program, when run on a computer device, causes the computer device to perform an intelligent driving vehicle behavior analysis and prediction method as described above.
The above embodiments may be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above-described embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with embodiments of the present invention are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center over a wired network or a wireless network. The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more sets of available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium may be a solid state disk.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely one, and there may be additional divisions in actual implementation, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Finally: the foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. An intelligent driving vehicle behavior analysis and prediction method is characterized by comprising the following steps:
collecting vehicle type characteristic data;
training a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle;
collecting accident training data;
generating an accident assessment value based on the accident training data;
training a second machine learning model for predicting the type of traffic accident occurring in surrounding vehicles in real time based on the accident assessment value, the accident training data and the output of the first machine learning model;
a driving adjustment instruction is generated based on an output of the second machine learning model.
2. The intelligent driving vehicle behavior analysis and prediction method according to claim 1, wherein the vehicle type feature data comprises vehicle dynamic video data and vehicle three-dimensional point cloud data, and the vehicle dynamic video data and the vehicle three-dimensional point cloud data are synchronously collected for the test vehicle.
3. The intelligent driving vehicle behavior analysis and prediction method according to claim 2, wherein the training manner of the first machine learning model is as follows:
setting a tag for a real-time set of vehicle type feature data, wherein the tag is a positive integer and represents a vehicle type; constructing labels corresponding to each group of vehicle type characteristic data and each group of vehicle type characteristic data into one sample, and collecting a plurality of samples to construct a machine learning data set; the data set is divided into a training set, a verification set and a test set, wherein the training set accounts for 70% of the data set, and the verification set and the test set respectively account for 15% of the data set;
taking the training set as input of a first machine learning model, wherein the first machine learning model takes a tag of a vehicle type as output and corresponds to real-time vehicle type characteristic dataThe labels are predictive targets to minimize machine learning model loss function valuesAs a training target, i is a vehicle type characteristic data group number; u is the number of the characteristic data sets of the vehicle type; y is i Tag corresponding to the i-th group of vehicle type characteristic data, < > for the i-th group of vehicle type characteristic data>A tag predicted for the i-th set of real-time vehicle type feature data; stopping training when the loss function of the first machine learning model is smaller than or equal to a preset target loss value; the first machine learning model is a neural network model.
4. A method for intelligent driving vehicle behavior analysis and prediction according to claim 3, wherein the accident training data comprises types of surrounding vehicles, accelerations of the surrounding vehicles and maximum speeds of the surrounding vehicles when a traffic accident of several kinds occurs, and the accident training data further comprises distances between the surrounding vehicles and the driving vehicles n seconds before the traffic accident occurs;
the accident training data further includes a number of sets of types of surrounding vehicles without traffic accidents, accelerations of the surrounding vehicles, maximum speeds of the surrounding vehicles, and distances of the surrounding vehicles from the driving vehicle.
5. The intelligent driving vehicle behavior analysis and prediction method according to claim 4, wherein the accident assessment value is calculated as follows:
in the formula of danger r An accident evaluation value for the r-th vehicle in the surrounding vehicles, A r A is the vehicle type influence coefficient of the r-th vehicle in the surrounding vehicles r For acceleration of the r-th vehicle of the surrounding vehicles, V r For the r-th vehicle in the surrounding vehiclesFront maximum speed, d r The distance between the r-th vehicle and the own vehicle in the surrounding vehicles is set;
setting an accident assessment threshold danger_go, when danger r <When the danger_go is, marking as a safe driving state; when danger r And marking dangerous driving states when the traffic is not less than danger_go.
6. The intelligent driving vehicle behavior analysis and prediction method according to claim 5, wherein the training manner of the second machine learning model is as follows:
taking each group of accident training data and corresponding accident assessment values as characteristic data, and taking a plurality of accident type numbers as labels, wherein each group of accident type numbers corresponds to each group of characteristic data;
taking the characteristic data as input of a second machine learning model, wherein the second machine learning model takes an accident type number as output, takes a label corresponding to the real-time characteristic data as a prediction target, and takes a minimized machine learning model loss function value as a training target; and stopping training when the loss function of the second machine learning model is smaller than or equal to the target loss value.
7. The intelligent driving vehicle behavior analysis and prediction method according to claim 6, wherein the driving adjustment instruction comprises a voice prompt for a driver and an intelligent deceleration operation for the vehicle; after the driving adjustment instruction is generated, a danger announcement instruction is generated; the danger announcement instruction comprises the step of sending an early warning notice to all surrounding vehicle-mounted central control devices to inform about the type of the impending accident.
8. An intelligent driving vehicle behavior analysis and prediction system, comprising:
the first data acquisition module (1) is used for acquiring vehicle type characteristic data;
the second data acquisition module (2) is used for acquiring accident training data;
a data analysis module (4) that generates an accident assessment value based on the accident training data;
the model training module (3) is used for training a first machine learning model for identifying the type of the vehicle in real time based on the characteristic data of the type of the vehicle; training a second machine learning model for predicting the type of traffic accident occurring in surrounding vehicles in real time based on the accident assessment value, the accident training data and the output of the first machine learning model;
a control module (5) for generating a driving adjustment instruction and a danger announcement instruction based on the output of the second machine learning model;
the cloud notification module (6) transmits an early warning notification to surrounding vehicles based on the danger announcement instruction.
9. An electronic device, comprising: a processor and a memory, wherein the memory stores a computer program for the processor to call;
the processor performs an intelligent driving vehicle behavior analysis and prediction method according to any one of claims 1 to 7 by calling a computer program stored in the memory.
10. A computer-readable storage medium, characterized by: instructions stored thereon which, when executed on a computer, cause the computer to perform an intelligent driving vehicle behavior analysis and prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311060538.7A CN117022323A (en) | 2023-08-22 | 2023-08-22 | Intelligent driving vehicle behavior analysis and prediction system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311060538.7A CN117022323A (en) | 2023-08-22 | 2023-08-22 | Intelligent driving vehicle behavior analysis and prediction system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117022323A true CN117022323A (en) | 2023-11-10 |
Family
ID=88602182
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311060538.7A Pending CN117022323A (en) | 2023-08-22 | 2023-08-22 | Intelligent driving vehicle behavior analysis and prediction system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117022323A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117494589A (en) * | 2024-01-03 | 2024-02-02 | 北京中机车辆司法鉴定中心 | Vehicle accident prediction method, device and storage medium based on vehicle body color |
| CN117690303A (en) * | 2024-02-04 | 2024-03-12 | 四川三元环境治理股份有限公司 | Noise early warning system, device and early warning method based on traffic data acquisition |
-
2023
- 2023-08-22 CN CN202311060538.7A patent/CN117022323A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117494589A (en) * | 2024-01-03 | 2024-02-02 | 北京中机车辆司法鉴定中心 | Vehicle accident prediction method, device and storage medium based on vehicle body color |
| CN117494589B (en) * | 2024-01-03 | 2024-04-09 | 北京中机车辆司法鉴定中心 | Vehicle accident prediction method, device and storage medium based on vehicle body color |
| CN117690303A (en) * | 2024-02-04 | 2024-03-12 | 四川三元环境治理股份有限公司 | Noise early warning system, device and early warning method based on traffic data acquisition |
| CN117690303B (en) * | 2024-02-04 | 2024-04-26 | 四川三元环境治理股份有限公司 | Noise early warning system, device and early warning method based on traffic data acquisition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11024165B2 (en) | Driver behavior monitoring | |
| US10769456B2 (en) | Systems and methods for near-crash determination | |
| CN109789777B (en) | Unintended pulse-change collision detector | |
| CN112389448B (en) | Abnormal driving behavior identification method based on vehicle state and driver state | |
| EP4283575A2 (en) | Detection of driving actions that mitigate risk | |
| US7579942B2 (en) | Extra-vehicular threat predictor | |
| CN109345829B (en) | Unmanned vehicle monitoring method, device, equipment and storage medium | |
| CN117022323A (en) | Intelligent driving vehicle behavior analysis and prediction system and method | |
| CN110400478A (en) | A kind of road condition notification method and device | |
| WO2018026733A1 (en) | Determining causation of traffic events and encouraging good driving behavior | |
| EP3403219A1 (en) | Driver behavior monitoring | |
| CN112885144A (en) | Early warning method and system for vehicle crash event in construction operation area | |
| CN112373467A (en) | Intelligent obstacle avoidance system of unmanned automobile | |
| WO2020051168A1 (en) | Systems and methods for classifying driver behavior | |
| CN113635897B (en) | Safe driving early warning method based on risk field | |
| EP3886076A1 (en) | Warning system for a host automotive vehicle | |
| JP2023535928A (en) | Systems, apparatus, and methods for operating a vehicle using sensor-monitored parameters | |
| CN112180951A (en) | Intelligent obstacle avoidance method for unmanned vehicle and computer readable storage medium | |
| CN113487912A (en) | Traffic accident early warning and protecting method, automobile and readable storage medium | |
| CN108831144B (en) | Collision risk avoidance processing method | |
| CN113393594A (en) | Automobile visual safety situation perception method based on collision risk of peripheral objects | |
| CN117864075A (en) | Active brake control method, device and storage medium | |
| CN117516579A (en) | Navigation line generation method, system and automatic driving method based on road section identification | |
| CN117445937A (en) | Unmanned vehicle machine vision deep learning system | |
| CN117831343A (en) | Confidence judging method, device, equipment and medium for traffic moving target |
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 |