CN115482659B - Intelligent autonomous decision-making method based on deep reinforcement learning - Google Patents
Intelligent autonomous decision-making method based on deep reinforcement learning Download PDFInfo
- Publication number
- CN115482659B CN115482659B CN202210992167.5A CN202210992167A CN115482659B CN 115482659 B CN115482659 B CN 115482659B CN 202210992167 A CN202210992167 A CN 202210992167A CN 115482659 B CN115482659 B CN 115482659B
- Authority
- CN
- China
- Prior art keywords
- time
- turning
- objects
- road
- comparison
- 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
Classifications
-
- 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
-
- 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
- G08G1/0145—Measuring and analyzing of parameters relative to traffic conditions for specific applications for active traffic flow control
-
- 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
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The application discloses an intelligent agent autonomous decision-making method based on deep reinforcement learning, which relates to the technical field of intelligent agent autonomous decision-making, and obtains license plate numbers, right turn time and vehicle identifications of all comparison objects by acquiring basic data aiming at intersections in a forbidden standard state; the comparison object is automatically deleted after the storage T1 time; then, primarily screening all potential objects, and determining all primarily suspicious objects and corresponding interval time according to the license plate numbers, the entering time and the vehicle identifications of the real objects and the comparison among the license plate numbers, the right turn time and the vehicle identifications of the comparison objects; and secondly auditing the primarily suspicious object, determining the illegal object through the secondarily auditing, facilitating the processing of law enforcement departments, and solving the problem that the problem cannot be processed in the prior art, so that potential safety hazards are caused.
Description
Technical Field
The application belongs to the technical field of autonomous decision making, and particularly relates to an intelligent autonomous decision making method based on deep reinforcement learning.
Background
Patent number CN111833597a discloses autonomous decisions in traffic situations with planned control. A control device for generating steering decisions of an autonomous vehicle in a traffic scenario is proposed. The control device comprises a first module comprising a trained self-learning model configured to receive data comprising information about the surroundings of the autonomous vehicle, by means of which an action to be performed by the autonomous vehicle is determined based on the received data. The control device includes a second module configured to receive the determined action, receive data including information about an ambient environment of the autonomous vehicle during the limited time range, predict an environmental state of the limited time range for a first period of time, determine a trajectory of the autonomous vehicle based on the received action of the limited time range and the environmental state of the first period of time, send a signal to control the autonomous vehicle during the first period of time according to the determined trajectory.
However, for the patent, an important circle is lacking for autonomous decision of traffic scenes, in the existing road condition driving process, some red lights are often avoided by turning right in advance and turning right again after turning around, and in the normal condition, the driving mode is allowed, but in the prior art, the situation of illegal turning around after turning right often exists, so that some potential safety hazards are caused, and a solution is provided for solving the problem.
Disclosure of Invention
The present application aims to solve at least one of the technical problems existing in the prior art; therefore, the application provides an intelligent agent autonomous decision-making method based on deep reinforcement learning.
To achieve the above object, an embodiment according to a first aspect of the present application provides an agent autonomous decision method based on deep reinforcement learning, which specifically includes the following steps:
step one: basic data acquisition is carried out on the crossing in the forbidden standard state, and the specific mode is as follows:
acquiring all right turning vehicle pictures at the intersection commanded by the target object in a forbidden target state, and marking the right turning vehicle pictures as comparison objects;
synchronously and automatically acquiring license plate numbers, right turning time and vehicle identifications of all comparison objects; the comparison object is automatically deleted after the storage T1 time;
step two: then, primarily screening all potential objects, and determining all primarily suspicious objects and corresponding interval time according to the license plate numbers, the entering time and the vehicle identifications of the real objects and the comparison among the license plate numbers, the right turn time and the vehicle identifications of the comparison objects;
step three: the secondary auditing method for the primarily suspicious object comprises the following specific steps:
s1: acquiring all the suspicious objects and the corresponding interval time;
s2: then, acquiring a pre-turning road of an object command road of the right turning entering target of the primarily suspicious object, wherein the pre-turning road is the road where the primarily suspicious object is located when the right turning entering target object, acquiring an object lane of the pre-turning road, and marking the object lane as a through lane;
s3: obtaining the distance that the right corner of the target object directs the road to enter the intersection position inserted through the lane and reach the position inserted through the turning point of the lane, marking the distance as the turning distance, and turning the turning point to indicate that the vehicle can turn around at the position;
s4: then, the distance from the turning point to the object command road reaching the standard from the road before turning is obtained, and the distance is marked as an intra-standard distance;
s5: obtaining a first speed limit value inserted through a lane, and dividing the turning distance by the first speed limit value to obtain a turning list;
then obtaining a second speed limit value of the road before turning, and dividing the intra-gauge distance by the second speed limit value to obtain a second turning time; adding a second turning time to the turning time to obtain short limit time;
s6: marking the suspicious object with the interval time exceeding the short limit as a speculative object;
s7: all the speculative objects are obtained.
Compared with the prior art, the application has the beneficial effects that:
according to the application, basic data acquisition is carried out on intersections in a forbidden standard state, so that license plate numbers, right turn time and vehicle identifications of all comparison objects are obtained; the comparison object is automatically deleted after the storage T1 time; then, primarily screening all potential objects, and determining all primarily suspicious objects and corresponding interval time according to the license plate numbers, the entering time and the vehicle identifications of the real objects and the comparison among the license plate numbers, the right turn time and the vehicle identifications of the comparison objects;
and secondly auditing the primarily suspicious object, determining the illegal object through the secondarily auditing, facilitating the processing of law enforcement departments, and solving the problem that the problem cannot be processed in the prior art, so that potential safety hazards are caused.
Detailed Description
The technical solutions of the present application will be clearly and completely described in connection with the embodiments, and it is obvious that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The application provides an agent autonomous decision method based on deep reinforcement learning, which as an embodiment of the application, specifically comprises the following steps:
step one: synchronizing the state of the target object, when any driving direction is in a forbidden target state, starting autonomous decision judgment, and keeping silence in the rest states; any driving direction is the direction in which the road corresponding to the traffic light is responsible for commanding can pass;
the target is referred to as a traffic light, and the state of the target is the specific of the real-time state of the traffic light in red, green and yellow; the mark forbidden state is a state that the vehicles in the straight direction corresponding to the straight lanes are in red light forbidden traffic;
step two: basic data acquisition is carried out on the crossing in the forbidden standard state, and the specific mode is as follows:
acquiring all right turning vehicle pictures at the intersection commanded by the target object in a forbidden target state, marking the right turning vehicle pictures as comparison objects, and synchronously and automatically acquiring license plate numbers, right turning time and vehicle identifications of all the comparison objects; the right turn time is the specific time point of the corresponding comparison object at the right turn; the comparison object is automatically deleted after storing the T1 time, T1 is a preset value, the value is usually fifteen minutes, and the comparison object can be set to other values;
the vehicle identification refers to the color of a vehicle engine cover and the brand of the vehicle, and the brand of the vehicle can be known by automatically identifying the brand through a camera;
step three: and then, initially screening all potential objects, wherein the initial screening method comprises the following specific steps:
s01: the method comprises the steps of obtaining pictures of all vehicles on a road commanded by a right turning entering target object, marking the corresponding vehicles as real entering objects through image recognition, and obtaining license plate numbers, entering time and vehicle identifications of all the real entering objects; the entering time is the time node when the real entering object enters the road commanded by the target object;
s02: firstly, carrying out any pairwise comparison on an actual object and a comparison object, comparing whether license plate numbers and vehicle identifications of the actual object and the comparison object are consistent, and if so, marking the consistent actual object as a preliminary object;
s03: synchronously obtaining interval time according to the entering time of the suspicious object and the right turn time of the comparison object, wherein the interval time is determined according to the time length of the interval from the entering time to the right turn time;
s04: obtaining all the suspicious objects and the corresponding interval time;
step four: the secondary auditing method for the primarily suspicious object comprises the following specific steps:
s1: acquiring all the suspicious objects and the corresponding interval time;
s2: then, acquiring a pre-turning road of an object command road of the right turning entering target of the primarily suspicious object, wherein the pre-turning road is the road where the primarily suspicious object is located when the right turning entering target object, acquiring an object lane of the pre-turning road, and marking the object lane as a through lane; the inserted lane meets the condition that the vehicle can enter through the right turn of the road commanded by the target object;
s3: obtaining the distance that the right corner of the target object directs the road to enter the intersection position inserted through the lane and reach the position inserted through the turning point of the lane, marking the distance as the turning distance, and turning the turning point to indicate that the vehicle can turn around at the position;
s4: then, the distance from the turning point to the object command road reaching the standard from the road before turning is obtained, and the distance is marked as an intra-standard distance;
s5: obtaining a first speed limit value inserted through a lane, and dividing the turning distance by the first speed limit value to obtain a turning list;
then obtaining a second speed limit value of the road before turning, and dividing the intra-gauge distance by the second speed limit value to obtain a second turning time; adding a second turning time to the turning time to obtain short limit time;
s6: marking the suspicious object with the interval time exceeding the short limit as a speculative object;
s7: obtaining all the speculative objects;
step five: all the speculative objects and pictures thereof are transmitted to the intelligent terminal ports of the corresponding supervision departments, so that the processing is convenient; the speculative object refers to the illegal situation that a user does not wait to turn around from a right turning road and turns around again at the illegal position when going straight through the red light, so as to avoid the traffic light;
as an embodiment two of the present application, on the basis of the embodiment one, the fourth step of the present embodiment is slightly different from the embodiment one, specifically:
step four: the secondary auditing method for the primarily suspicious object comprises the following specific steps:
s1: acquiring all the suspicious objects and the corresponding interval time;
s2: then, acquiring a pre-turning road of an object command road of the right turning entering target of the primarily suspicious object, wherein the pre-turning road is the road where the primarily suspicious object is located when the right turning entering target object, acquiring an object lane of the pre-turning road, and marking the object lane as a through lane; the inserted lane meets the condition that the vehicle can enter through the right turn of the road commanded by the target object;
s3: obtaining the distance that the right corner of the target object directs the road to enter the intersection position inserted through the lane and reach the position inserted through the turning point of the lane, marking the distance as the turning distance, and turning the turning point to indicate that the vehicle can turn around at the position;
s4: then, the distance from the turning point to the object command road reaching the standard from the road before turning is obtained, and the distance is marked as an intra-standard distance;
s5: obtaining a first speed limit value inserted through a lane, and dividing the turning distance by the first speed limit value to obtain a turning list;
then obtaining a second speed limit value of the road before turning, and dividing the intra-gauge distance by the second speed limit value to obtain a second turning time; adding a second turning time to the turning time to obtain short limit time; if the lane is inserted and the turning position does not exist, marking the short-limit time consumption as 0;
the absence of a u-turn position is specifically referred to herein as:
along the straight line of the inserted lane, before encountering the second traffic light and including the position of the second traffic light, no turning position exists;
s6: marking the suspicious object with the interval time exceeding the short limit as a speculative object;
s7: obtaining all the speculative objects;
of course, as the third embodiment of the present application, the present application further needs to perform the following steps after performing the treatment of the fifth step on the basis of the first embodiment, and the specific manner is as follows:
SS1: acquiring all the speculative objects and the time points marked as the speculative objects, and obtaining all the speculative objects and the speculative time points;
SS2: obtaining all the speculative objects from the current approaching stage and the times of marking the speculative objects as speculative objects, removing the times lower than X1, wherein X1 is a preset value and is generally 2; the approach stage is obtained by reckoning three months forward from the current time;
SS3: marking the rest of the gambling objects as inertial objects, and marking the times corresponding to the marked gambling objects as inertial times to obtain all the inertial objects and the inertial times;
SS4: optionally, an inertial object, wherein the interval time from the first time to the last time of the period marked as the speculative object is obtained, and each time the period marked as the last time of the speculative object is marked as a inter-projection period value, and the interval time value is correspondingly marked as Ji, i=1..n, and the interval time value is expressed as n inter-projection period values, namely the period is marked as n+1 times of the speculative object;
SS5: then automatically calculating the mean value of Ji, marking the mean value as P, and calculating the polymerization degree D of Ji according to a formula, wherein the specific calculation formula is as follows:
SS6: defining a doubling value according to the polymerization degree D, wherein when D is smaller than X2, the doubling value is equal to 2;
when X2 is more than or equal to D is more than or equal to X3, defining a doubling value to be 1.5;
when D > X3, a doubling value of 1 is defined;
SS7: then, carrying out the same processing of steps SS4-SS7 on all the inertial objects to obtain doubling values of all the inertial objects;
SS8: calculating the measurement values of all inertial objects by using the formula, wherein the specific formula is as follows:
measurement value = number of inertias x doubling value;
SS9: sorting according to the measurement values of the inertial objects from large to small, and marking the corresponding inertial objects with thirty-five percent of the top ranking as habit targets;
SS10: when the processing of any habit object in the step three of the embodiment is marked as a first suspected object, the processing is synchronously marked as a half-speculation object, and the half-speculation object is transmitted to an intelligent terminal port of a corresponding supervision department for processing, and certainly, the half-speculation object is not illegally determined to be a speculation object, and can be slightly started, and if the processing is marked as a speculation object later, the processing is normally penalized according to the mode of the speculation object.
As the fourth embodiment of the present application, the first to third embodiments are fused and implemented.
The partial data in the formula are all obtained by removing dimension and taking the numerical value for calculation, and the formula is a formula closest to the real situation obtained by simulating a large amount of collected data through software; the preset parameters and the preset threshold values in the formula are set by those skilled in the art according to actual conditions or are obtained through mass data simulation.
The above embodiments are only for illustrating the technical method of the present application and not for limiting the same, and it should be understood by those skilled in the art that the technical method of the present application may be modified or substituted without departing from the spirit and scope of the technical method of the present application.
Claims (7)
1. The intelligent autonomous decision-making method based on deep reinforcement learning is characterized by comprising the following steps of:
step one: basic data acquisition is carried out on the crossing in the forbidden standard state, and the specific mode is as follows:
acquiring all right turning vehicle pictures at the intersection commanded by the target object in a forbidden target state, and marking the right turning vehicle pictures as comparison objects;
synchronously and automatically acquiring license plate numbers, right turning time and vehicle identifications of all comparison objects; the comparison object is automatically deleted after the storage T1 time;
step two: then, primarily screening all potential objects, and determining all primarily suspicious objects and corresponding interval time according to the license plate numbers, the entering time and the vehicle identifications of the real objects and the comparison among the license plate numbers, the right turn time and the vehicle identifications of the comparison objects;
step three: the secondary auditing method for the primarily suspicious object comprises the following specific steps:
s1: acquiring all the suspicious objects and the corresponding interval time;
s2: then, acquiring a pre-turning road of an object command road of the right turning entering target of the primarily suspicious object, wherein the pre-turning road is the road where the primarily suspicious object is located when the right turning entering target object, acquiring an object lane of the pre-turning road, and marking the object lane as a through lane;
s3: obtaining the distance that the right corner of the target object directs the road to enter the intersection position inserted through the lane and reach the position inserted through the turning point of the lane, marking the distance as the turning distance, and turning the turning point to indicate that the vehicle can turn around at the position;
s4: then, the distance from the turning point to the object command road reaching the standard from the road before turning is obtained, and the distance is marked as an intra-standard distance;
s5: obtaining a first speed limit value inserted through a lane, and dividing the turning distance by the first speed limit value to obtain a turning list;
then obtaining a second speed limit value of the road before turning, and dividing the intra-gauge distance by the second speed limit value to obtain a second turning time; adding a second turning time to the turning time to obtain short limit time;
s6: marking the suspicious object with the interval time exceeding the short limit as a speculative object;
s7: all the speculative objects are obtained.
2. The method for autonomous decision making of an agent based on deep reinforcement learning according to claim 1, wherein the following steps are further performed before the step one is performed:
synchronizing the state of the target object, starting autonomous decision making when the target object is in a forbidden target state, and keeping silence in the rest states.
3. The method for autonomous decision making by an agent based on deep reinforcement learning of claim 2, wherein the target object is denoted by a traffic light; and the mark forbidden state is a state that the vehicles in the straight direction corresponding to the straight lanes are in red light forbidden traffic.
4. The intelligent agent autonomous decision-making method based on deep reinforcement learning according to claim 1, wherein in the first step, the right turn time is a specific time point of the corresponding comparison object at the right turn;
the vehicle identification refers to the vehicle hood color and the vehicle brand; and T1 is a preset value.
5. The method for autonomous decision making of an agent based on deep reinforcement learning according to claim 1, wherein the primary screening in the second step is specifically as follows:
s01: the method comprises the steps of obtaining pictures of all vehicles on a road commanded by a right turning entering target object, marking the pictures as an actual entering target object, and obtaining license plate numbers, entering time and vehicle identifications of all the actual entering target objects; the entering time is the time node when the real entering object enters the road commanded by the target object;
s02: firstly, carrying out any pairwise comparison on an actual object and a comparison object, comparing whether license plate numbers and vehicle identifications of the actual object and the comparison object are consistent, and if so, marking the consistent actual object as a preliminary object;
s03: synchronously obtaining interval time according to the entering time of the suspicious object and the right turn time of the comparison object, wherein the interval time is determined according to the time length of the interval from the entering time to the right turn time;
s04: all the suspicious objects and the corresponding interval time are obtained.
6. The intelligent autonomous decision-making method based on deep reinforcement learning according to claim 1, wherein the inserted lane in step S2 satisfies that the vehicle can enter through a right turn of the road directed by the subject;
in step S5, if the lane is inserted and there is no turning position, the short-limit time is marked as 0.
7. The method for autonomous decision making by an agent based on deep reinforcement learning according to claim 1, wherein after the step three is performed, the following steps are further performed:
and transmitting all the speculative objects to the intelligent terminal ports of the corresponding supervision departments.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210992167.5A CN115482659B (en) | 2022-08-18 | 2022-08-18 | Intelligent autonomous decision-making method based on deep reinforcement learning |
| JP2023051884A JP7518993B2 (en) | 2022-08-18 | 2023-03-28 | Autonomous decision-making method for learning subjects based on advanced reinforcement learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210992167.5A CN115482659B (en) | 2022-08-18 | 2022-08-18 | Intelligent autonomous decision-making method based on deep reinforcement learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115482659A CN115482659A (en) | 2022-12-16 |
| CN115482659B true CN115482659B (en) | 2023-09-19 |
Family
ID=84422523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210992167.5A Active CN115482659B (en) | 2022-08-18 | 2022-08-18 | Intelligent autonomous decision-making method based on deep reinforcement learning |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP7518993B2 (en) |
| CN (1) | CN115482659B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108831168A (en) * | 2018-06-01 | 2018-11-16 | 江苏数翰信息科技有限公司 | A kind of method for controlling traffic signal lights and system based on association crossing visual identity |
| CN109949579A (en) * | 2018-12-31 | 2019-06-28 | 上海眼控科技股份有限公司 | A kind of illegal automatic auditing method that makes a dash across the red light based on deep learning |
| WO2020192122A1 (en) * | 2019-03-27 | 2020-10-01 | 江苏智通交通科技有限公司 | Off-site law enforcement picture intelligent auditing method and system for vehicles running red light |
| WO2021093335A1 (en) * | 2019-11-14 | 2021-05-20 | Suzhou Zhijia Science & Technologies Co., Ltd. | Method for automatically labeling lane changing intention based on high-noise trajectory data of vehicle |
| CN113682312A (en) * | 2021-09-23 | 2021-11-23 | 中汽创智科技有限公司 | Autonomous lane changing method and system integrating deep reinforcement learning |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106114507B (en) * | 2016-06-21 | 2018-04-03 | 百度在线网络技术（北京）有限公司 | Local path planning method and device for intelligent vehicle |
| JP7028066B2 (en) | 2018-05-30 | 2022-03-02 | 株式会社Ihi | Detection device and detection system |
| CN112749681A (en) | 2021-01-25 | 2021-05-04 | 长威信息科技发展股份有限公司 | Violation detection method based on edge equipment and deep learning |
-
2022
- 2022-08-18 CN CN202210992167.5A patent/CN115482659B/en active Active
-
2023
- 2023-03-28 JP JP2023051884A patent/JP7518993B2/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108831168A (en) * | 2018-06-01 | 2018-11-16 | 江苏数翰信息科技有限公司 | A kind of method for controlling traffic signal lights and system based on association crossing visual identity |
| CN109949579A (en) * | 2018-12-31 | 2019-06-28 | 上海眼控科技股份有限公司 | A kind of illegal automatic auditing method that makes a dash across the red light based on deep learning |
| WO2020192122A1 (en) * | 2019-03-27 | 2020-10-01 | 江苏智通交通科技有限公司 | Off-site law enforcement picture intelligent auditing method and system for vehicles running red light |
| WO2021093335A1 (en) * | 2019-11-14 | 2021-05-20 | Suzhou Zhijia Science & Technologies Co., Ltd. | Method for automatically labeling lane changing intention based on high-noise trajectory data of vehicle |
| CN113682312A (en) * | 2021-09-23 | 2021-11-23 | 中汽创智科技有限公司 | Autonomous lane changing method and system integrating deep reinforcement learning |
Non-Patent Citations (1)
| Title |
|---|
| 面向智能避障场景的深度强化学习研究;刘庆杰;林友勇;李少利;;智能物联技术(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115482659A (en) | 2022-12-16 |
| JP7518993B2 (en) | 2024-07-19 |
| JP2024028097A (en) | 2024-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110647839B (en) | Method and device for generating automatic driving strategy and computer readable storage medium | |
| US10885777B2 (en) | Multiple exposure event determination | |
| CN112700470B (en) | Target detection and track extraction method based on traffic video stream | |
| CN113212454B (en) | Method and device for adjusting running state of vehicle, computer equipment and storage medium | |
| US20090326796A1 (en) | Method and system to estimate driving risk based on a heirarchical index of driving | |
| CN112292719A (en) | Adapting the trajectory of an ego-vehicle to a moving foreign object | |
| CN113071487B (en) | Automatic driving vehicle control method and device and cloud equipment | |
| CN107077606A (en) | For the method and apparatus for the travel situations for determining there is currently | |
| CN112673379A (en) | Dynamic spatial scene analysis | |
| CN111539268A (en) | Road condition early warning method and device during vehicle running and electronic equipment | |
| CN114694060A (en) | Road shed object detection method, electronic equipment and storage medium | |
| CN111315623A (en) | Vehicle control device | |
| CN116547187A (en) | Selecting driving maneuvers for at least partially automated driven vehicles | |
| CN116244902A (en) | Road environment scene simulation method, electronic equipment and medium for vehicle-road cloud fusion | |
| CN115482659B (en) | Intelligent autonomous decision-making method based on deep reinforcement learning | |
| CN113989772A (en) | Traffic light detection method and device, vehicle and readable storage medium | |
| CN117671617A (en) | Real-time lane recognition method in container port environment | |
| CN106601024A (en) | Lane change vehicle recognition apparatus using surrounding lane identification and method thereof | |
| CN112896166A (en) | Vehicle lane changing method and device and electronic equipment | |
| CN110745143A (en) | Vehicle control method, device, equipment and storage medium | |
| Tewari et al. | AI-based autonomous driving assistance system | |
| CN112758107B (en) | Automatic lane changing method for vehicle, control device, electronic equipment and automobile | |
| CN113593253A (en) | Method and device for monitoring red light running of vehicle | |
| CN118323197B (en) | Training and using method and device of predictive model in automatic driving vehicle and vehicle | |
| CN115257807B (en) | Urban road scene automatic driving decision-making method and device based on knowledge graph |
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 |