CN110009903B - Traffic accident scene restoration method - Google Patents
Traffic accident scene restoration method Download PDFInfo
- Publication number
- CN110009903B CN110009903B CN201910162408.1A CN201910162408A CN110009903B CN 110009903 B CN110009903 B CN 110009903B CN 201910162408 A CN201910162408 A CN 201910162408A CN 110009903 B CN110009903 B CN 110009903B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- accident
- data
- wheel
- traffic accident
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G06T11/23—
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/025—Services making use of location information using location based information parameters
- H04W4/027—Services making use of location information using location based information parameters using movement velocity, acceleration information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/029—Location-based management or tracking services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
本发明涉及一种交通事故现场还原方法，包括以下步骤：S1、利用设置在车上的GPS装置和传感器采集的数据绘制出事故发生前一分钟车辆的行驶轨迹；S2、以绘制出的事故发生前一分钟车辆的行驶轨迹和发生事故一瞬间的车辆记录仪中的数据作为输入，运用深度学习算法得到还原的车辆行驶轨迹。与现有技术相比，本发明利用安装部署在车辆上车载装置采集的数据，经过反复训练的深度学习算法，能有效去除随机误差和人为主观因素的干扰，大大降低事故误判率，省去了不必要的时间、人力的浪费，对维护良好的社会秩序有极大促进作用。
The present invention relates to a method for restoring a traffic accident scene, comprising the following steps: S1, using a GPS device arranged on the vehicle and data collected by a sensor to draw the driving track of the vehicle one minute before the accident occurs; S2, using the drawn accident occurrence The driving trajectory of the vehicle in the previous minute and the data in the vehicle recorder at the moment of the accident are used as input, and the restored vehicle driving trajectory is obtained by using the deep learning algorithm. Compared with the prior art, the present invention utilizes the data collected by the on-board device installed and deployed on the vehicle, and the deep learning algorithm after repeated training can effectively remove the interference of random errors and human subjective factors, greatly reduce the accident misjudgment rate, and save the It saves unnecessary waste of time and manpower, and greatly promotes the maintenance of good social order.
Description
Technical Field
The invention relates to a traffic accident scene simulation and reproduction technology, in particular to a traffic accident scene restoration method.
Background
The main task of the simulation and reproduction of the traffic accident is to estimate the running speed, the acceleration, the course angle, the running track of the vehicle after the vehicle is collided and the like in the collision process by using the running marks left on the road, the collision marks on the vehicle body and the stopping position of the vehicle at the accident site so as to estimate the occurrence reason of the traffic accident, determine the accident responsibility and carry out corresponding punishment and the like. This process is a vehicle travel track reproduction technique.
Under the background of rapid development of current transportation, the occurrence frequency of traffic accidents is high. In a traffic accident, if the two parties of the traffic accident can not agree on who is the main responsible party of the accident in the current accident, the light accident often falls into a rather troublesome processing flow and must wait for the traffic police to arrive at the scene. In order to obtain specific information of an accident site, a traffic police needs to measure and evaluate the damage condition of a vehicle on the site, the stop position of the vehicle and the like so as to realize simulation and reproduction of the accident site. However, this method usually consumes a lot of time for both parties, so that the following vehicles cannot run normally to cause traffic jam, and also causes a certain waste of police resources. Moreover, the artificial measurement process includes artificial subjectivity and objective errors, and the result of accident reproduction often deviates greatly from the actual situation. In the existing research on vehicle track reproduction in traffic accidents, the accident scene is reproduced by collecting video based on a camera installed and deployed on a vehicle, and the track reproduction is realized by modeling the vehicle only through the stop position of the vehicle. However, these methods ignore the system dynamic response of the vehicle in operation and are therefore often less than satisfactory.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a traffic accident scene restoration method.
The purpose of the invention can be realized by the following technical scheme:
a traffic accident scene restoration method comprises the following steps:
s1, drawing the driving track of the vehicle one minute before the accident by using the data collected by the GPS device and the sensor arranged on the vehicle;
and S2, taking the drawn driving track of the vehicle one minute before the accident and the data in the vehicle recorder at the moment of the accident as input, and obtaining the restored driving track of the vehicle by applying a deep learning algorithm.
Preferably, in step S1, the traveling direction of the vehicle is determined by the correspondence between the difference between the tension sensors mounted on both sides of the wheel and the turning direction and angle of the vehicle, and the emergency braking of the vehicle is determined by the pressure sensors mounted on the wheel.
Preferably, data acquired by the sensor in the method is filtered by a Kalman filtering algorithm and then is subjected to other processing.
Preferably, the deep learning algorithm is specifically a neural network algorithm.
Preferably, the neural network algorithm specifically adopts LSTM to build a recurrent neural network.
Preferably, the data collected by the GPS device and the sensor includes: position of vehicle, vehicle speed, acceleration, course angle, tire pressure, tire surface tension, yaw rate, angular acceleration.
Preferably, the restored vehicle driving track is specifically obtained through simulation of an OpenCv algorithm.
Preferably, the restored vehicle driving track is sent to a traffic police or a mobile terminal of an owner of the vehicle.
Compared with the prior art, the invention has the following advantages:
1. by using data acquired by a vehicle-mounted device installed and deployed on a vehicle and a repeatedly trained deep learning algorithm, random errors and interference of human subjective factors can be effectively removed, and accident misjudgment rate is greatly reduced; the mobile phone terminal is matched to directly generate a simulation reproduction picture of the vehicle running track of an accident scene, so that two accident parties do not need to wait for traffic polices in situ, congestion possibly caused by the accident is greatly reduced, unnecessary waste of time and manpower is saved, and good social order maintenance is greatly promoted.
2. High accuracy: the method completely avoids subjective assumption during accident responsibility confirmation from the realization method, eliminates random errors and is beneficial to treating traffic accidents fairly and equitably.
3. High adaptability: the method adopts a neural network model with large data flow, so that the adaptability of the method is greatly improved, and after full training, the method can be competent for route prediction under any condition; meanwhile, each successful prediction can be brought into a new data set, so that the neural network is continuously updated, and the accuracy of the neural network can be ensured without spending manpower on maintenance.
4. High real-time performance: the wireless communication technology used by the method has small transmission time delay, can ensure that the acquired motion state information and the position information of the vehicle are always real-time, has low algorithm complexity and small time consumption, and can draw a conclusion as soon as possible.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a simulation diagram of a Kalman filter model with Gaussian random noise as the noise established in the embodiment;
fig. 3 is a registration flowchart of a mobile APP in an embodiment;
FIG. 4 is a flowchart illustrating the operation of the owner of the vehicle on the APP after the accident in the embodiment;
FIG. 5 is a flow chart of the operation of the traffic police on the APP after an accident in an embodiment;
FIG. 6 is a Kalman filtered data of the angular acceleration value of a certain vehicle in the embodiment.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
Examples
As shown in fig. 1, the present application provides a traffic accident scene restoration method, which records a driving track of a vehicle in real time through a vehicle-mounted device, can automatically extract the driving track of the vehicle within one minute before a car accident occurs after the car accident occurs, generates an original driving track of the vehicle on the basis of the image, and finally outputs a final image. The method comprises the following steps:
s1, installing a pressure sensor and a tension sensor on a vehicle wheel, starting a vehicle-mounted GPS device, and drawing a driving track of the vehicle one minute before an accident occurs by using data collected by the GPS device and the sensor provided on the vehicle, where the data collected by the GPS device and the sensor is transmitted to a microprocessor unit in the vehicle in this embodiment, and the method specifically includes:
acquiring data such as longitude and latitude, vehicle speed, course angle and the like of a vehicle through a GPS device, and mapping the vehicle coordinate to a two-dimensional rectangular coordinate system by utilizing coordinate system conversion;
judging the vehicle traveling direction: when the vehicle turns in the advancing process, the tension force applied to the left surface and the right surface of the wheel is different due to the action of centripetal force, based on the principle, tension sensors can be arranged on the two sides of the wheel, and then the advancing direction of the vehicle is determined according to the corresponding relation between the difference value of the tension force on the two sides of the wheel and the turning direction and angle of the vehicle;
judging whether the vehicle is suddenly braked: the real-time speed of the vehicle can be reflected, the pressure between a front side tire and the ground can be increased when the vehicle is braked emergently under the normal condition, and the pressure can be obtained through a pressure sensor arranged on the wheel; the tire is slightly deformed under the action of pressure, so that the trace left on the ground by the wheel is wider than that of the wheel in normal running, and based on the principle, the change of the running trace of the wheel of the vehicle can be simulated in a computer through the acceleration during the braking of the vehicle;
drawing an image: and drawing the running track of the vehicle one minute before the accident through the values measured by the GPS device, the tension sensor and the pressure sensor.
And S2, obtaining a restored vehicle running track by applying a deep learning algorithm by taking the drawn vehicle running track one minute before the accident and data such as speed, acceleration and the like in a vehicle recorder at the moment of the accident as input.
Because the data collected by the sensors is accompanied by a lot of noise, the data must first be filtered using a Kalman (Kalman) filtering algorithm.
The Kalman filter estimates the process state by a feedback control method: the filter estimates the state of the process at a certain moment and then obtains feedback in the form of a (noisy) measured variable. The final Kalman estimation algorithm becomes a kind of pre-estimation-correction algorithm with numerical solution.
Kalman filter for estimating state variable x ∈ R of discrete time processn. The state equation of the discrete linear steady-state system is as follows:
xk＝Axk-1+Buk-1+wk-1 (1)
defining an observed variable z ∈ RmAnd obtaining a measurement equation:
zk＝Hxk+vk (2)
wherein wkAnd vkRepresenting process excitation noise and observation noise, respectively. Assuming that they are normally distributed white noises independent of each other, the Kalman filter estimates the next state mainly by equation (1), according to zkCorrecting the estimator:
wherein x ∈ Rn-represents prior and ^ represents estimation) is prior state estimation of the kth step under the condition that the prior state of the kth step is known; x is formed by RnFor a known measured variable zkEstimating the posterior state of the kth step; pkEstimating error covariance for the prior; pk -Estimating covariance for the posteriori; r is the variance of the observed noise.
Matlab is used for establishing a Kalman filter model with Gaussian random noise as noise for simulation, the effect is shown in figure 2, it can be seen that the measured data of the sensor can not be directly used for calculation in practical application, the deviation between the data result filtered by Kalman and the theoretical predicted value is not large, the precision requirement of calculation is met, and the filtering effect is good.
The deep learning algorithm adopted by the method is specifically a neural network algorithm. In the practical application scenario of the method, the controlled object is a running vehicle, if the parameters (coordinates) of the traveling track of the vehicle without an accident are regarded as the output of the neural network, and the parameters (the track before the accident, the speed, the acceleration, the course angle, the data of each sensor and the like) of the running state of the vehicle are regarded as the input of the neural network. The coordinates of each moment of the vehicle are determined by the coordinates of the previous moment and the operation state parameters, and a causal relationship exists between the moments, so that the neural network is not suitable for being built by a common full-connection layer neural network DNN, but is suitable for being built by a recurrent neural network RNN.
A multilayer feedback RNN (Recurrent neural Network) neural Network is an artificial neural Network with nodes directionally connected into a ring. The internal state of such a network may exhibit dynamic timing behavior. Unlike feed-forward neural networks, the RNN can use its internal memory to process an input sequence at an arbitrary timing, highlighting the logical relationship between time points and time points.
As known from the major forward and backward algorithms of RNN, a big problem faced by RNN is that there may be gradient vanishing or bursting for long span time. In this embodiment, a Long Short-Term Memory network (LSTM) is used to build the RNN network, so as to solve the above problem.
The LSTM is one in which the cell has its own state and can change its state by internal update, and the gate structure adds or deletes information to or from the state of the cell, and the gate is divided into an input gate, an output gate, and a forgetting gate. Through the LSTM, the time information of the input sequence can be guaranteed not to be lost, the situation of gradient explosion or disappearance under the long-span time condition of the RNN can be solved, and the robustness of the neural network is guaranteed.
The input and output data of the neural network are as follows:
the input data is divided into two parts: the running track parameters (coordinates) of one minute before the accident and various parameters (including position, speed, acceleration, course angle, yaw rate, angular acceleration, tire pressure and tire surface tension) of the filtered running state of the vehicle, which are obtained by real-time positioning through a GPS device, are in the form of a sequence consisting of digital streams and containing time logic. The output data is the predicted vehicle travel trajectory (coordinates) without an accident, and is also a series of sequences including time logic.
The training result of the neural network is as follows:
training uses a TensorFlow deep learning framework and GPU rendering + CUDA accelerated training under a Ubuntu 16.04LTS system. The training data sets are 2500 groups in total, each group uses Numpy to generate data to simulate various parameters in the vehicle running process, and matplotlib is used to simulate the vehicle running track within one minute before the accident happens and draw the track.
In this embodiment, after 1000 rounds of training are performed on the training data set, the accuracy of the test on 50 verification data sets with the same structure is as high as 98.3%. The random error can be greatly eliminated by 1000 groups of training data sets, and the accuracy of the deep learning method is proved by repeated verification of 50 different verification data sets.
In order to verify the feasibility of the method as much as possible under the conditions of ensuring safety and avoiding property loss, the obtained model is simulated under the framework of the OpenCv algorithm. The simulation result shows that the preset particle motion track and the prediction motion track at every moment calculated according to the deep learning network model almost have no deviation. The simulation well proves that the track predicted by the network can well approach the track of the vehicle without accidents under the real condition.
The restored track obtained by the method can be sent to a mobile terminal application program (APP) of a user:
1) APP main function design: there is the communication function between this APP and the car-mounted device. After an accident occurs, a user logs in the APP and clicks a key to operate, and a driving track of the vehicle before the accident occurs and an image simulated by the original driving track are obtained.
The car owner and the traffic police can log in the APP platform to perform related operations. The basic logical framework is: the vehicle owner enters the operation interface of the APP to upload the related data and the pictures, and then the background program uploads the data to the database. When the traffic police enters the operation interface to process the traffic accident, the background program calls the corresponding data from the database.
2) The use process comprises the following steps:
registering: when using this APP, user registration is first required. The user only needs to download the APP in the mobile phone APP store, click to register and operate according to the following operation flow, and fig. 3 is a registration flow chart.
The specific operation sequence is as follows: for car owners, the operation of the entire application after an accident is shown in fig. 4, and for traffic police, the operation of the entire application after an accident is shown in fig. 5.
In the present embodiment, it is assumed that there are two vehicles at the first vehicle (121.21348E,31.28772N) and the second vehicle (121.21352E,31.28771N), respectively. And the microprocessor unit performs Kalman filtering on the read sensor data and records the sensor data, and the GPS synchronizes the running position information of the vehicle in real time. FIG. 6 shows real-time readings of the angular acceleration of the vehicle over the horizon and Kalman filtered data, which is used as input data to the neural network along with the GPS recorded trajectory. And (3) assuming that the two vehicles collide with each other, quickly calculating according to input data by the neural network, drawing a preset running track of the two vehicles when no accident occurs after the calculation is completed, and finally obtaining that the first vehicle is in a straight-going state before the collision and the second vehicle turns right before the collision.
Claims (6)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910162408.1A CN110009903B (en) | 2019-03-05 | 2019-03-05 | Traffic accident scene restoration method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910162408.1A CN110009903B (en) | 2019-03-05 | 2019-03-05 | Traffic accident scene restoration method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110009903A CN110009903A (en) | 2019-07-12 |
| CN110009903B true CN110009903B (en) | 2022-02-18 |
Family
ID=67166479
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910162408.1A Expired - Fee Related CN110009903B (en) | 2019-03-05 | 2019-03-05 | Traffic accident scene restoration method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110009903B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110782670A (en) * | 2019-11-05 | 2020-02-11 | 北京汽车集团有限公司 | Scene restoration method based on data fusion, vehicle cloud platform and storage medium |
| CN112863244B (en) * | 2019-11-28 | 2023-03-14 | 大众汽车股份公司 | Method and apparatus for promoting safe driving of a vehicle |
| CN111640308B (en) * | 2020-04-24 | 2022-03-08 | 合肥湛达智能科技有限公司 | Deep learning red light running detection method based on embedded terminal |
| CN111651829B (en) * | 2020-06-12 | 2022-03-29 | 招商局重庆交通科研设计院有限公司 | Road simulation platform based on intelligent traffic flow simulation |
| CN114510533B (en) * | 2022-01-06 | 2023-03-17 | 北京中交兴路车联网科技有限公司 | Accident recovery method and device, electronic equipment and storage medium |
| CN114373311B (en) * | 2022-01-27 | 2023-06-06 | 支付宝(杭州)信息技术有限公司 | Vehicle driving guiding method and device |
| TWI824788B (en) * | 2022-10-21 | 2023-12-01 | 財團法人車輛研究測試中心 | System and method of integrating traffic accident assistance investigation and safety of the intended functionality scene establishment |
| CN117912259B (en) * | 2024-03-19 | 2024-06-21 | 中汽数据有限公司 | Traffic accident reproduction method and device based on automobile electronic data, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105975721A (en) * | 2016-05-27 | 2016-09-28 | 大连楼兰科技股份有限公司 | Accident recurrence collision simulation establishing method and accident recurrence collision simulation method based on vehicle real-time motion state |
| CN107291972A (en) * | 2017-03-10 | 2017-10-24 | 清华大学 | The Intelligent Vehicle Driving System efficiency evaluation method excavated based on multi-source data |
| CN108665757A (en) * | 2018-03-29 | 2018-10-16 | 斑马网络技术有限公司 | Simulating vehicle emergency system and its application |
| CN109272745A (en) * | 2018-08-20 | 2019-01-25 | 浙江工业大学 | Vehicle trajectory prediction method based on deep neural network |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2768092Y (en) * | 2004-12-29 | 2006-03-29 | 何崇中 | Work information recorder for automobile tyre |
| US7536247B2 (en) * | 2005-11-30 | 2009-05-19 | Ford Global Technologies | Method and apparatus for reconstructing an accident using side satellite lateral acceleration sensors |
| US9067565B2 (en) * | 2006-05-22 | 2015-06-30 | Inthinc Technology Solutions, Inc. | System and method for evaluating driver behavior |
| CN102073789B (en) * | 2010-12-30 | 2013-03-20 | 长安大学 | Analysis, calculation and simulative representation system for combined accident of vehicle-vehicle collision and vehicle-fixed object collision |
| CN102063573B (en) * | 2010-12-31 | 2012-09-12 | 长安大学 | Computer system for analysis, computation, simulation and reconstruction of car falling accident |
| US9296263B2 (en) * | 2011-12-23 | 2016-03-29 | Prasad Muthukumar | Smart active tyre pressure optimising system |
| US20210256614A1 (en) * | 2014-09-22 | 2021-08-19 | State Farm Mutual Automobile Insurance Company | Theft identification and insurance claim adjustment using drone data |
| EP3159853B1 (en) * | 2015-10-23 | 2019-03-27 | Harman International Industries, Incorporated | Systems and methods for advanced driver assistance analytics |
| CN105956265A (en) * | 2016-04-29 | 2016-09-21 | 大连楼兰科技股份有限公司 | Method and system for simulation and reappearance of crash accident simulation with process-based simulation parameters |
| CN106897948B (en) * | 2017-01-04 | 2021-01-01 | 北京工业大学 | Riding and pushing traffic accident identification method |
| CN108791276B (en) * | 2017-04-26 | 2020-05-19 | 湖南大学 | A quick method for judging the linear/non-linear working state of tire lateral force |
-
2019
- 2019-03-05 CN CN201910162408.1A patent/CN110009903B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105975721A (en) * | 2016-05-27 | 2016-09-28 | 大连楼兰科技股份有限公司 | Accident recurrence collision simulation establishing method and accident recurrence collision simulation method based on vehicle real-time motion state |
| CN107291972A (en) * | 2017-03-10 | 2017-10-24 | 清华大学 | The Intelligent Vehicle Driving System efficiency evaluation method excavated based on multi-source data |
| CN108665757A (en) * | 2018-03-29 | 2018-10-16 | 斑马网络技术有限公司 | Simulating vehicle emergency system and its application |
| CN109272745A (en) * | 2018-08-20 | 2019-01-25 | 浙江工业大学 | Vehicle trajectory prediction method based on deep neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110009903A (en) | 2019-07-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110009903B (en) | Traffic accident scene restoration method | |
| CN110177374B (en) | V2X functional application testing method, device and system based on vehicle-road cooperation | |
| CN107169402B (en) | vehicle lane positioning | |
| CN109784254B (en) | A method, device and electronic device for vehicle violation event detection | |
| CN110796007B (en) | Method and computing device for scene recognition | |
| CN107491736A (en) | A kind of pavement adhesion factor identifying method based on convolutional neural networks | |
| CN114274972A (en) | Scene recognition in an autonomous driving environment | |
| CN116901975B (en) | Vehicle-mounted AI security monitoring system and method thereof | |
| CN111183464B (en) | System and method for estimating saturated flow at signalized intersections based on vehicle trajectory data | |
| US20210042642A1 (en) | Driving action evaluating device, driving action evaluating method, and recording medium storing driving action evaluating program | |
| CN113126624B (en) | Automatic driving simulation test method, device, electronic equipment and medium | |
| CN115879294B (en) | A method and system for generating full-sample vehicle flow trajectories based on multi-vehicle environment perception | |
| JP2020096286A (en) | Determination device, determination program, determination method, and neural network model generation method | |
| CN113183988B (en) | Method, device and equipment for supervising automatic driving of vehicle and storage medium | |
| CN112419345B (en) | High-precision tracking method for patrol cars based on echo state network | |
| CN114970805A (en) | Method for operating a motor vehicle | |
| CN113092135A (en) | Test method, device and equipment for automatically driving vehicle | |
| CN112434782A (en) | Architecture and method for state estimation fault detection using crowdsourcing and deep learning | |
| EP4220581A1 (en) | Estimation of risk exposure for autonomous vehicles | |
| EP4219261A1 (en) | Estimation of risk exposure for autonomous vehicles | |
| CN113848073A (en) | A driving assistance method and system for predicting the probability of collision between people and vehicles | |
| CN114968187A (en) | Platform for perception system development of an autopilot system | |
| CN114968189A (en) | Platform for perception system development of an autopilot system | |
| CN114037969A (en) | Automatic driving lane information detection method based on radar point cloud and image fusion | |
| CN113795010A (en) | Method and device for testing automobile queue, electronic equipment and storage medium thereof |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220218 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |