CN110349407B - Regional traffic signal lamp control system and method based on deep learning - Google Patents
Regional traffic signal lamp control system and method based on deep learning Download PDFInfo
- Publication number
- CN110349407B CN110349407B CN201910610269.4A CN201910610269A CN110349407B CN 110349407 B CN110349407 B CN 110349407B CN 201910610269 A CN201910610269 A CN 201910610269A CN 110349407 B CN110349407 B CN 110349407B
- Authority
- CN
- China
- Prior art keywords
- module
- unit
- traffic
- information
- data
- 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
- 238000013135 deep learning Methods 0.000 title claims abstract description 52
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 58
- 238000004088 simulation Methods 0.000 claims abstract description 52
- 238000004364 calculation method Methods 0.000 claims abstract description 46
- 238000004891 communication Methods 0.000 claims abstract description 37
- 238000005457 optimization Methods 0.000 claims description 28
- 238000012216 screening Methods 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 15
- 230000000694 effects Effects 0.000 claims description 7
- 238000005516 engineering process Methods 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 9
- 238000005259 measurement Methods 0.000 description 6
- 230000006872 improvement Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012546 transfer Methods 0.000 description 3
- 206010039203 Road traffic accident Diseases 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
-
- 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
- G08G1/048—Detecting movement of traffic to be counted or controlled with provision for compensation of environmental or other condition, e.g. snow, vehicle stopped at detector
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/095—Traffic lights
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
- H04L67/025—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP] for remote control or remote monitoring of applications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computing Systems (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Multimedia (AREA)
- Entrepreneurship & Innovation (AREA)
- Analytical Chemistry (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Chemical & Material Sciences (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
A regional traffic signal lamp control system and method based on deep learning comprises an information acquisition unit, a storage unit, a 5G communication unit, a cloud data processing and database unit, a regional road network model building unit, a Synchro-based simulation calculation unit, a deep learning-based traffic prediction unit and a traffic signal control unit, wherein data such as traffic flow are acquired through the acquisition unit, 5G communication transmission is used, data information such as traffic flow of each road is integrated in a cloud server, prediction is carried out based on the deep learning, then the predicted traffic flow data are subjected to simulation based on the Synchro to obtain an optimal control scheme, and finally the optimal control scheme is transmitted to each road red and green lamp for execution, so that the management and control of traffic signals in a region can be effectively optimized, and an effective method is provided for relieving traffic congestion.
Description
Technical Field
The invention relates to a regional traffic signal lamp control system and method based on deep learning, and belongs to the technical field of intelligent traffic.
Background
Along with economic development, the urban scale is gradually expanded, the traffic volume is increased day by day, and in order to meet the increasing travel demand of people, solve the problem of the way struggle between people and vehicles, improve the travel quality and relieve the traffic jam problem, more traffic signal lamps are needed to be added to improve the phenomenon. However, the conversion cycles of the currently adopted traffic signal lamps are fixed time duration, and the traffic signal lamps cannot adapt to real-time road conditions of different intersections and cannot solve the problems of uneven traffic flow peak driving direction distribution and the like.
The initial application of the existing intelligent traffic signal lamp control system makes up part of the problems, but still has the following defects:
1. information collection is less comprehensive. The current situations of different corresponding intersections such as weather and special events are not collected in real time.
2. There is hysteresis in the data. The control scheme obtained from data acquisition to calculation is longer in duration, so that a larger time difference exists between information transmission and command execution, and the beneficial effect brought by the implementation of the control scheme is seriously reduced.
3. The optimization control target is single. The regulation and control schemes obtained by different calculation methods are based on the shortest delay consideration, and lack of comprehensive consideration for other optimization control targets.
4. The traffic signal control system is centralized on single intersection signal lamp regulation and control, and the overall control and coordination of regional traffic conditions are lacked, so that conflicts and contradictions between traffic signal control time length and traffic flow speed control exist in a region.
Disclosure of Invention
The invention aims to provide a regional traffic signal lamp control system based on deep learning, a regional road network model is established according to collected effective information, traffic simulation calculation is carried out according to different control targets on the basis to obtain a traffic signal lamp control scheme, and urban traffic travel quality is improved.
In order to achieve the purpose, the invention adopts the following technical scheme:
a regional traffic signal control system based on deep learning comprises a single-intersection traffic signal lamp and a cloud server connected with the traffic signal lamp, wherein the single-intersection traffic signal lamp comprises an information acquisition unit, a storage unit, a 5G communication unit and a traffic signal control unit; the cloud server comprises a cloud data processing and database unit, a deep learning-based traffic prediction unit, a Synchro-based simulation calculation unit and a regional road network model building unit; the information acquisition unit is connected with the storage unit, the storage unit is connected with the 5G communication unit, the 5G communication unit is connected with the data processing and database unit, the data processing and database unit is connected with the traffic prediction unit based on deep learning, the traffic prediction unit based on deep learning is connected with the simulation calculation unit based on Synchro, the regional road network model building unit is connected with the simulation calculation unit based on Synchro, and the simulation calculation unit based on Synchro is also connected with the traffic signal control unit through the 5G communication unit and the storage unit.
The invention has the further improvement that the information acquisition unit comprises a weather acquisition module, a traffic flow image acquisition module, a pedestrian flow image acquisition module, a time acquisition module and a special event acquisition module;
the weather acquisition module comprises a light intensity sensor, a humidity sensor and a temperature sensor, wherein the 3 sensors are uniformly arranged at the lower part of the traffic signal lamp post at the single intersection; the light intensity sensor, the humidity sensor and the temperature sensor are all connected with the storage unit.
The traffic flow image acquisition module comprises a plurality of cameras for acquiring vehicle image information of each direction of a single intersection;
the pedestrian traffic image acquisition module is used for acquiring pedestrian traffic image information of a sidewalk;
the time acquisition module is used for recording the time period corresponding to the information;
the special event acquisition module is arranged in the cloud server database module and is used for acquiring important festivals, activities and special vehicle passing information of a current city and area through the Internet.
The invention has the further improvement that the storage unit comprises a single intersection signal lamp self-storage module and a simulation calculation control result storage module; the single intersection signal lamp storage module is used for storing and numbering all information of the weather acquisition module, the traffic flow image acquisition module, the pedestrian flow image acquisition module, the time acquisition module and the special event acquisition module; and the simulation calculation control result storage module is used for storing the control information downloaded by the 5G communication unit.
The invention has the further improvement that the 5G communication unit is arranged at the top of the signal lamp post of the single intersection; the 5G communication unit comprises a data uploading module and a data downloading module, the data uploading module is connected with the storage unit 202, the data downloading module is connected with the simulation computing unit based on Synchro, and the data uploading module is used for uploading data; the data downloading module is used for downloading data.
The cloud data processing and database unit comprises an information and image processing module and a database module, wherein the information and image processing module is used for realizing vehicle identification and face identification processing on the collected image received by the cloud based on a video detection technology to obtain traffic flow data of different vehicle types in corresponding time periods and the number of pedestrians in different age stages in a waiting area of a sidewalk, and collecting data of each road junction through corresponding numbers to form road network data of the area;
and the database module is used for receiving and storing various traffic volume data, pedestrian flow data, environment information at corresponding moments and special event data which are obtained by the information and image processing module.
The invention has the further improvement that the traffic prediction unit based on deep learning comprises a database information reading module, a traffic prediction model training module and a prediction result storage module, wherein the database information reading module is used for automatically calling and storing various traffic data and pedestrian traffic data stored in the cloud data processing and database unit, and environment information and special event data at corresponding moments; the traffic prediction model training module is used for establishing a plurality of prediction models based on deep learning, and calling data in the database information reading module to continuously train the established prediction models based on deep learning to obtain a traffic prediction result of the regional road network; the prediction result storage module is used for storing the prediction result of the regional road network traffic volume obtained by the traffic volume prediction model training module.
The invention has the further improvement that the Synchro-based simulation calculation unit comprises a storage module, a main intersection screening module and an intersection signal timing optimization module, wherein the main intersection screening module is used for determining a current traffic signal control target according to the current temperature, humidity, visibility and relevant information of special events; the intersection signal timing optimization module is used for calling the regional road network model of the regional road network model building unit and carrying out synchronous simulation calculation on the control target of the main intersection screening module to obtain the control duration of the traffic signal; the storage module is used for receiving road network model files of the regional road network model building unit, regional road network traffic prediction results predicted by the traffic prediction unit based on deep learning, control target data of the main intersection screening module and calculation results of the intersection signal timing optimization module.
The regional road network model building unit comprises a road information measuring module and a model building module, wherein the road information measuring module is used for carrying out oblique photography on a control region road network, collecting image data and carrying out measurement on road length, number of lanes, lane width, number of inlet roads, change positions and positions of inlet roads and stop lines; the model building module is used for building a plane and three-dimensional road network model by combining the image data, the road length, the number of lanes, the lane width and the number of the inlet roads and the lane number conversion position data acquired by the road information measuring module with the image information of the satellite map.
The invention is further improved in that the control unit comprises an instruction receiving and processing module, a signal lamp automatic control module and a display screen control module, wherein the instruction receiving and processing module is used for receiving information generated by the signal lamp automatic control module and correspondingly matching all intersection signal lamp control schemes in a road network with the numbers of signal lamps at single intersections, the signal lamp automatic control module is used for controlling the color and the duration of traffic signals, and the display screen control module is used for implementing the intersection signal lamp control information and displaying the traffic volume information according to the information of the instruction receiving and processing module.
A control method of a regional traffic signal control system based on deep learning comprises the steps of collecting information of a single intersection through an information collection unit, storing the collected information through a storage unit, uploading the collected information to a cloud data processing and database unit in a cloud server through a 5G communication unit, processing the information through an information and image processing module and storing the information in a database module, reading and importing the processed information through a database information reading module of a traffic prediction unit based on deep learning, predicting through a traffic prediction module, storing a prediction result in a prediction result storage module, reading a regional traffic network model established by a predicted traffic and regional traffic network model establishing unit through the storage module, optimizing and simulating a control target through a main screening module and an intersection signal optimization module, and finally, obtaining control information, storing the control scheme in a storage module, downloading the control scheme by using a 5G communication unit, and finally receiving the control scheme and controlling the lipstick and green light of each road through the control unit to realize the management and control of traffic signals in the road network.
Compared with the prior art, the invention has the following beneficial effects:
according to the invention, the single-intersection traffic signal lamp comprises the information acquisition unit, so that the road information of different intersections in an updated area can be acquired in real time, the information is butted with the cloud server, and the cloud database is established based on the information and image processing technology, so that the measurement index is refined. Through setting up 5G communication unit, improved data transmission download speed, save time cost improves efficiency. According to the invention, through arranging the traffic prediction unit based on deep learning, the future can be directly predicted according to the traffic and pedestrian volume data in the database, so that the simulation calculation is carried out on the prediction result, and the influence of information transfer hysteresis is avoided. If a traffic prediction unit based on deep learning is not adopted, simulation calculation can be carried out only after summary processing and image recognition are carried out on the data acquired in real time after the data are uploaded, the control of traffic signals can be realized only by feeding back the result calculated by the simulation method to a traffic signal control system, and the simulation calculation needs a long time so that the information acquired by the traffic signals is lagged, so that the influence of information transmission lag is avoided.
The invention collects information of a single intersection through an information collection unit, stores the collected information through a storage unit, uploads the collected information to a cloud data processing and database unit in a cloud server through a 5G communication unit, processes the information through an information and image processing module and stores the information in a database module, reads and imports the processed information data through a database information read-in module of a traffic prediction unit based on deep learning, predicts the traffic prediction module, stores the prediction result in a prediction result storage module, reads the predicted traffic and a road network model established by a regional road network model establishment unit through the storage module, and performs control target optimization and simulation calculation through a main intersection screening module and an intersection signal timing optimization module, and finally, obtaining a control scheme, storing the control scheme in a storage module, downloading the control scheme by using a 5G communication unit, and finally receiving the control scheme and controlling the lipstick and green light of each road through the control unit to realize the management and control of traffic signals in the road network. The method is based on the Synchro simulation calculation control result, determines the target of optimization control under different time conditions according to the effective information in the database, considers the influence of special events on road intersections, and obtains signal lamp time control schemes aiming at different conditions. By establishing a regional road network model, road condition information and traffic flow change in the whole region are recorded in detail, and on the basis, simulation calculation is carried out by combining with cloud database information, so that coordination work of traffic lights in the whole region is realized, the traffic jam phenomenon is effectively relieved, the purpose of realizing regional control of the traffic lights is achieved, and an effective method is provided for relieving traffic jam.
Drawings
FIG. 1 is a general block diagram of a regionalized traffic signal control system of the present invention;
FIG. 2 is a flow chart of a regionalized traffic signal control system of the present invention;
FIG. 3 is a block diagram of an information collection unit according to the present invention;
FIG. 4 is a block diagram of a memory cell structure according to the present invention;
FIG. 5 is a block diagram of a 5G communication unit according to the present invention;
FIG. 6 is a block diagram of a cloud data processing and database unit architecture according to the present invention;
FIG. 7 is a block diagram of a traffic prediction unit based on deep learning according to the present invention;
FIG. 8 is a block diagram of a Synchro-based simulation computing unit according to the present invention;
FIG. 9 is a block diagram of a regional road network model building unit according to the present invention;
FIG. 10 is a block diagram of a traffic signal control unit according to the present invention;
fig. 11 is a schematic view of a traffic signal light matched with the invention.
Wherein: 010 is an information acquisition unit; 020 is a storage unit; 030 is a 5G communication unit; 040 is a cloud data processing and database unit; 050 is a traffic prediction unit based on deep learning; 060 is a regional road network model building unit; 070 is a simulation calculation unit based on Synchro; 080 a traffic signal control unit; 011 is that the information acquisition unit comprises a weather acquisition module; 011-1 is a temperature sensor, 011-2 is a humidity sensor, 011-3 is a light intensity sensor, 012 is a traffic flow image acquisition module, 013 is a pedestrian flow image acquisition module, 014 is a time acquisition module, and 015 is a special event acquisition module; 020 is a storage unit, 021 is a single intersection signal lamp image self-storage module, 022 is a simulation calculation control result storage module; 030 is a 5G communication unit, 031 is a data uploading module, 032 is a data downloading module; 040 is a cloud data processing and database unit, 041 is an information and image processing module, and 042 is a database module; 050 is a traffic prediction unit based on deep learning, 051 is a database information reading module, 052 is a traffic prediction model training module, and 053 is a prediction result storage module; 060 is a simulation calculation unit based on Synchro, 061 is a storage module, and 062 is a main intersection screening module; 063 is an intersection signal timing optimization module; 070 is a regional road network model building unit, 071 is a road information measuring module, and 072 is a model building module; 080 to control unit, 081 to command receiving and processing module, 082 to signal lamp automatic control module, 083 to display screen control module; 091 is a camera, 092 is a signal lamp time display module, and 093 is a signal lamp color display module.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
Unless otherwise defined, the relative arrangement of the components and steps, the numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present invention. It should also be understood that the dimensions of the various features shown in the drawings are not drawn to scale for ease of illustration. Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate. All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" is not to be interpreted as implying any or all combinations of one or more of the associated listed items.
Referring to fig. 1 to 11, the invention provides a regional traffic signal control system based on deep learning, which includes a single intersection traffic signal lamp and a cloud server connected to the traffic signal lamp, wherein the single intersection traffic signal lamp includes an information acquisition unit 010, a storage unit 020, a 5G communication unit 030, and a traffic signal control unit 080; the cloud server comprises a cloud data processing and database unit 040, a traffic prediction unit 050 based on deep learning, a simulation calculation unit 060 based on Synchro and a regional road network model building unit 070, and forms a single-way traffic signal lamp for being matched with a traffic signal control system and improved on the basis of an existing traffic signal lamp.
Referring to fig. 1, the information acquisition unit 010, the storage unit 020, the 5G communication unit 030, and the traffic signal control unit 080 form a single intersection traffic signal lamp, referring to fig. 11, the information acquisition unit 010 realizes single intersection information acquisition through the weather acquisition module 011 and the camera 091, and the traffic signal lamp control unit 080 realizes reasonable control of a traffic signal lamp through the display screen control module 083, the signal lamp time display module 092, and the signal lamp color display module 093.
Specifically, referring to fig. 1, the information acquisition unit 010 is connected to the storage unit 020, the storage unit 020 is connected to the 5G communication unit 030, the 5G communication unit 030 is connected to the data processing and database unit 040, the data processing and database unit 040 is connected to the deep learning-based traffic prediction unit 050, the deep learning-based traffic prediction unit 050 is connected to the Synchro-based simulation calculation unit 060, the regional road network model creation unit 070 is connected to the Synchro-based simulation calculation unit 060, and the syncchro-based simulation calculation unit 060 is further connected to the traffic signal control unit 080 via the 5G communication unit 030 and the storage unit 020.
Referring to fig. 3, the information collecting unit 010 includes a weather collecting module 011, a traffic flow image collecting module 012, a pedestrian flow image collecting module 013, a time collecting module 014, and a special event collecting module 015;
referring to fig. 11, the weather collection module 011 is disposed at the middle lower part of the traffic signal light post, and the weather collection module 011 includes a light intensity sensor 011-3, a humidity sensor 011-2 and a temperature sensor 011-1, wherein the 3 sensors are disposed at the lower part of the traffic signal light post at the single intersection. Specifically, a light intensity sensor 011-3, a humidity sensor 011-2 and a temperature sensor 011-1 are respectively arranged on the single-intersection traffic signal lamp post from top to bottom and are respectively used for acquiring visibility, humidity and temperature of the intersection at all times all day, and therefore the three types of information are taken as environmental factors and are also taken as a consideration factor for optimizing a simulation model. The light intensity sensor 011-3, the humidity sensor 011-2 and the temperature sensor 011-1 are all connected with the storage unit 020.
The single intersection traffic signal lamp post is further provided with a display screen control module 083, a signal lamp time display module 092 and a signal lamp color display module 093.
The traffic flow image collecting module 012 includes several cameras 091, and the several cameras 091 are arranged on the upper part of the signal lamp beam, have different quantity, and are determined according to the road width. Specifically, cameras 091 with different numbers are arranged at the upper part of a single intersection traffic signal lamp beam, and the cameras 091 acquire vehicle image information of each direction of a single intersection at the speed of 10 frames per second; the pedestrian flow image acquisition module 013 (the pedestrian flow image acquisition module 013 is a camera) acquires pedestrian traffic image information of the sidewalk through the camera 091; the time acquisition module 014 is used for recording the time period corresponding to the information; the special event acquisition module 015 is arranged in the cloud server database module 042 and acquires major festivals, activities and special vehicle passing information of a current city and area through the Internet, wherein the major festivals are five, six, mid-autumn, eleven, Christmas, and denier; the activities are major traffic accidents, road maintenance, concerts, star propaganda, food festivals and the like are carried out on the closed part of the road section, and special vehicles are ambulances, police cars for executing tasks of fire fighting vehicles and the like.
Referring to fig. 11, the weather collection module 011 is arranged at the middle lower part of the traffic signal light post, and the weather collection module 011 comprises light intensity sensors 011-3, humidity sensors 011-2 and temperature sensors 011-1, 3 sensors all arranged at the lower part of the traffic signal light post at the single intersection. Specifically, the light intensity sensor 011-3, the humidity sensor 011-2 and the temperature sensor 011-1 are respectively arranged from top to bottom and are respectively used for collecting visibility, humidity and temperature of the intersection at each moment all day, and the three types of information are taken as environmental factors and taken as a consideration factor for optimizing the simulation model; the traffic flow image collecting module 012 includes several cameras 091, and the several cameras 091 are arranged on the upper part of the signal lamp beam, have different quantity, and are determined according to the road width. Specifically, cameras 091 with different numbers are arranged at the upper part of a single intersection traffic signal lamp beam, and the cameras 091 acquire vehicle image information of each direction of a single intersection at the speed of 10 frames per second; the pedestrian flow image acquisition module 013 (the pedestrian flow image acquisition module 013 is a camera) acquires pedestrian traffic image information of the sidewalk through the camera 091; the time acquisition module 014 is used for recording the time period corresponding to the information; the special event acquisition module 015 is arranged in the cloud server database module 042 and acquires major festivals, activities and special vehicle passing information of a current city and area through the Internet, wherein the major festivals are five, six, mid-autumn, eleven, Christmas, and denier; the activities are major traffic accidents, road maintenance, concerts, star propaganda, food festivals and the like are carried out on the closed part of the road section, and special vehicles are ambulances, police cars for executing tasks of fire fighting vehicles and the like.
Referring to fig. 4, the storage unit 020 includes a single intersection signal lamp self-storage module 021 and a simulation calculation control result storage module 022. The single intersection signal lamp storage module 021 stores and codes all information of the weather acquisition module 011, the traffic flow image acquisition module 012, the pedestrian flow image acquisition module 013, the time acquisition module 014 and the special event acquisition module 015 in the driving direction information acquisition unit. The simulation calculation control result storage module 022 is configured to store the control scheme downloaded by the 5G communication unit 030.
Referring to fig. 5, the 5G communication unit 030 includes a data uploading module 031 and a data downloading module 032, where the data uploading module 031 is configured to upload an image of a single intersection signal lamp from the storage module 021 to the cloud server. The data downloading module 032 is configured to download data, and download a control scheme obtained by simulation calculation of the simulation calculation unit 060 based on Synchro. The 5G communication unit 030 is arranged on the top of the signal lamp post.
Referring to fig. 6, the cloud data processing and database unit 040 includes an information and image processing module 041 and a database module 042, where the information and image processing module 041 is configured to summarize and process information and image files transmitted at each intersection, implement vehicle identification and face identification processing based on a video detection technology, process a summarized image received by the cloud (the processing is the video detection technology), obtain traffic flow numbers of different vehicle types in corresponding time periods and pedestrian numbers in different age periods in a sidewalk waiting area, and summarize data at each intersection to form road network data in the area through corresponding numbers. The traffic time required by pedestrians in different vehicle types and different ages is different, the distinguishing and counting of different vehicle types and ages of the pedestrians are effectively realized through a video detection technology, the data has great significance in final prediction and simulation, and in addition, the data measured by all the road ports form traffic volume data of the road network area, so that the regional control is realized. The database module 042 is configured to receive the road network related information and the traffic data obtained by the information and image processing module 041, and store various types of traffic data and pedestrian traffic data obtained by the information and image processing module 041, and environmental information at a corresponding time, data of a special event, and the like in a database.
The information and image processing unit 041 identifies vehicles by adopting a BP neural network and a digital image processing technology to realize vehicle type identification and automatic division; the face recognition adopts an age characteristic extraction algorithm based on PCA (principal component analysis) to estimate the age stage of the waiting area of the pedestrian and realize the division, and the age stage is specifically divided into an old stage and a non-old stage.
Referring to fig. 7, the deep learning based traffic prediction unit 050 includes a database information reading module 051, a traffic prediction model training module 052 and a prediction result storage module 053, wherein the database information reading module 051 is used for automatically calling various types of traffic data and pedestrian traffic data stored in a database module 042, and environmental information, special events and other data at corresponding moments; the traffic prediction model training module 052 is used for establishing a prediction model based on deep learning, and calling data in the database information reading module 051 to continuously train the established prediction model based on deep learning to obtain a traffic prediction result of the regional road network; the prediction result storage module 053 is used for storing the prediction result of the road network traffic of the region of the traffic prediction model training module 052. And continuously training the established prediction model based on deep learning according to the collected data, so that the prediction result is more and more accurate.
The built prediction model based on deep learning is an SAE short-time traffic prediction model, the SAE short-time traffic prediction model is continuously trained, and then the SAE short-time traffic prediction model is adopted to predict the traffic of the current intersection; the traffic volume of the current intersection is closely related to the time sequence and the traffic flow of the connected intersections.
The concrete process of training the SAE short-time traffic prediction model is as follows: the traffic volume l in the data extraction module 042 is divided by taking every 5min as a time period and is taken as an input feature vector which is marked as lt＝{liWhere i is 1, 2, … …, i indicates the number of adjacent junctions, l_i＝{l(i,t),l(i,t-5),l(i,t-10),l(i,t-15),……},l_iRepresenting the amount of traffic on the ith road segment at time t.
Firstly, inputting a characteristic vector, training a first layer, carrying out measurement by using a mean square error loss function to minimize the error between an initial input value and a reconstructed output value, and taking the first layer as a next automatic encoder; taking the output of the first layer as the input of the second layer as an automatic encoder; repeating the second step until the required number of layers is reached; the output of the last layer is used as the input of a prediction layer and is input into a logistic regression layer, and the parameters of the logistic regression layer are initialized through supervision and training; finally, parameters of all layers are finely adjusted by using a BP method.
Specifically, the traffic volume in the first five periods of the predicted traffic volume period is taken as an input vector, the number of hidden layers is 2, and the number of hidden layer nodes is 400.
Referring to fig. 8, the Synchro-based simulation calculation unit 060 includes a storage module 061, a main intersection screening module 062, and an intersection timing optimization module 063, where the storage module 061 is configured to receive the road network model file of the regional road network model building unit 070, a result predicted by the deep learning-based traffic prediction unit 050 (including information such as traffic volume, traffic type, and pedestrian volume of each intersection of the road network), control target data of the main intersection screening module 062, and a calculation result of the intersection timing optimization module 063.
The main intersection screening module 062 is used for determining the importance degree of an intersection based on a PageRank algorithm according to data of the special event acquisition module and the road information measurement module; the intersection signal timing optimization module 063 is configured to invoke various data of the storage module 061, perform synchoro simulation calculation on the road network model and the intersection importance ranking result in the main intersection screening module 062 to obtain a traffic signal control scheme, and further, during simulation optimization, when the screened main intersection is subjected to matching and locking timing, realize regional signal timing coordination control by optimizing other adjacent intersections.
The method for screening the main intersections by the main intersection screening module 062 based on the PageRank improved algorithm is as follows: taking traffic flow as a measurement parameter of the relevance between intersections, and firstly calling the traffic volume composition of different advancing directions (straight advancing, left turning and right turning) of the inlet and outlet channels of each intersection in a database; the influence of outlet flow on inlet flow is calculated by adopting a PageRank algorithm, and the importance of the I-th inlet channel of any intersection R is calculated by adopting the following formula;
wherein, Z (R)I) -intersection R, entrance lane importance;
q is the flow correction factor;
U(rj) -intersection M, entrance lane j, turn ratio, where R is intersection R adjacent intersection;
C(rj) -the capacity of the intersection M entrance lane j;
m is the number of inlet channels related to the I inlet of the R intersection.
And updating and iterating for multiple times to obtain a stable intersection R entrance road importance value, and finally taking the maximum value of each entrance of the intersection R as the ranking basis of the intersection importance, wherein the intersections ranked at the top 25% are main intersections.
Referring to fig. 9, the regional road network model building unit 070 includes a road information measuring module 071 and a model building module 072, the road information measuring module 070 performs oblique photography on the road network of the control region by using modes such as an unmanned aerial vehicle, acquires image data, and performs measurement of road length, number of lanes, lane width, number of lanes of an inlet road, position of a parking line of the inlet road and the like by using modes such as laser scanning (ranging); the model building module 072 is used for building a plane and three-dimensional road network model by combining the image data, the road length, the lane number, the lane width, the number of the inlet roads and the lane change positions and the like collected by the road information measuring module 071 with the image information of the satellite map.
Referring to fig. 10, the control unit 080 includes an instruction receiving and processing module 081 for receiving a traffic signal control scheme generated by the signal lamp automatic control module 082, a signal lamp automatic control module 082 for controlling a color, a time duration, etc. of a traffic signal, and a display screen control module 083 for displaying traffic volume information of each road, which are received by the instruction receiving and processing module 081.
Referring to fig. 2, the control method based on the above-mentioned regionalized traffic signal control system is: the method comprises the steps of collecting information of a single intersection through a weather collection module 011, a traffic flow image collection module 012, a pedestrian flow image collection module 013, a time collection module 014 and a special event collection module 015 of an information collection unit 010, storing the collected information from a storage module 021 through a single intersection signal lamp of a storage unit 020, uploading the collected information to a cloud data processing and database module 042 of a database unit 040 through a data uploading module 031 of a 5G communication unit 030, processing the information through an information and image processing module 041 to obtain heavy vehicle proportions, pedestrian flow, intersection inlet and outlet traffic flow and turning proportions, storing the heavy vehicle proportions, the pedestrian flow, the intersection inlet and outlet traffic flow and the turning proportions in the database module 042, reading and importing the processed information data through a database information reading-in module 051 of a traffic flow prediction unit 050 based on deep learning, the traffic prediction module 052 predicts according to the information of the database information read-in module 051, the prediction result is stored in the prediction result storage module 053, then the prediction traffic and the regional road network model established by the regional road network model establishing unit 070 are read through the storage module 061, the intersection type is determined according to the regional road network model established by the storage module 070, and the intersection type is divided into three-way intersection, four-way intersection and intersection with more than four ways; road network intersection signal timing optimization is carried out through the main intersection screening module 062 and the intersection signal timing optimization module 063, a control scheme is finally obtained, the control scheme is stored in the storage module 061, at the moment, the data downloading module 032 of the 5G communication unit 030 is used for downloading the control scheme, and finally control scheme receiving and red and green light control of each intersection are carried out through the instruction receiving and processing module 081 of the control unit 080, the signal lamp automatic control module 082 and the display screen control module 083, so that management and control of traffic signals in the road network are achieved.
According to the special event acquisition module 015, when no special event exists, the intersection determined by the main intersection screening module 062 is directly called as the intersection for locking timing to optimize, when the special event exists, the property of the special event is judged, if the special event exists, traffic control is adopted, and the effective green light passing time is prolonged until the special vehicle passes; if the situation is other, the main intersection should be screened again by taking the traffic flow under the special event as a parameter.
According to the invention, data such as traffic flow and the like are collected through the collection unit, 5G communication transmission is used, data information such as traffic flow of each intersection and the like is integrated in the cloud server, prediction is carried out based on deep learning, then the predicted traffic flow data is subjected to simulation based on Synchro, parameters such as heavy vehicle ratio and pedestrian flow are input to obtain an optimal control scheme, and finally the optimal control scheme is transmitted to red and green lamps of each intersection for execution, so that the control of traffic signals in an area can be effectively optimized, and a feasible method is provided for relieving traffic congestion.
In the prior art, the regulation and control schemes generally obtained by different calculation methods are based on the shortest delay consideration and lack of consideration for other optimization control factors. The Synchro system adopted by the invention takes delay, queuing length and parking times as comprehensive optimization indexes to measure, and realizes signal timing optimization control based on multiple targets.
The invention has the following advantages:
(1) the information collection is relatively comprehensive. The signal lamp matched with the control method can collect the weather condition of the intersection in real time, remind traffic participants on a display screen, simultaneously collect whether special events occur, and determine whether the main intersection is screened and needs to be recalculated according to whether the special events occur or not; and determining whether to carry out traffic control according to whether the property of the special event is the special vehicle passing.
(2) Avoiding information transfer hysteresis effects. According to the invention, through arranging the traffic prediction unit based on deep learning, the future can be directly predicted according to the traffic and pedestrian volume data in the database, so that the simulation calculation is carried out on the prediction result, and the influence of information transfer hysteresis is avoided. If a traffic prediction unit based on deep learning is not adopted, because simulation calculation can be carried out only after the data acquired by uploading implementation needs to be subjected to summary processing and image recognition, the result calculated by the simulation method is fed back to a traffic signal control system to realize the control of traffic signals, and the simulation calculation needs a long time, so that the information acquired by the traffic signals is delayed.
(3) And diversifying the optimization control target. The signal lamp control scheme is based on a Synchro simulation calculation unit based on Synchro, and a Synchro system measures delay, queue length and parking times as comprehensive optimization indexes.
(4) And the regional control of the traffic signal lamp is realized. According to the method, a regional road network model is established, traffic flow prediction is carried out, road condition information and traffic flow change in a region are recorded in detail, intersection importance degree sequencing in the road network is combined, when intersection signal optimization is carried out, main intersection optimization timing is locked, optimization timing of adjacent intersections is adjusted, and coordination work of traffic lights in the whole region is achieved.
The above embodiments are merely illustrative of the technical ideas and features of the present invention, and are intended to enable those skilled in the art to understand the contents of the present invention and implement the present invention, and not to limit the scope of the present invention. All equivalent changes or modifications made according to the spirit of the present invention should be covered within the protection scope of the present invention.
Claims (7)
1. A regional traffic signal control system based on deep learning is characterized by comprising a single-intersection traffic signal lamp and a cloud server connected with the traffic signal lamp, wherein the single-intersection traffic signal lamp comprises an information acquisition unit (010), a storage unit (020), a 5G communication unit (030) and a traffic signal control unit (080); the cloud server comprises a cloud data processing and database unit (040), a deep learning-based traffic prediction unit (050), a Synchro-based simulation calculation unit (060) and a regional road network model building unit (070); the information acquisition unit (010) is connected with the storage unit (020), the storage unit (020) is connected with the 5G communication unit (030), the 5G communication unit (030) is connected with the data processing and database unit (040), the data processing and database unit (040) is connected with the traffic prediction unit (050) based on deep learning, the traffic prediction unit (050) based on deep learning is connected with the simulation calculation unit (060) based on Synchro, the regional road network model building unit (070) is connected with the simulation calculation unit (060) based on Synchro, and the simulation calculation unit (060) based on Synchro is also connected with the traffic signal control unit (080) through the 5G communication unit (030) and the storage unit (020);
the information acquisition unit (010) comprises a weather acquisition module (011), a traffic flow image acquisition module (012), a pedestrian flow image acquisition module (013), a time acquisition module (014) and a special event acquisition module (015);
the weather collection module (011) comprises a light intensity sensor (011-3), a humidity sensor (011-2) and a temperature sensor (011-1), wherein the 3 sensors are all arranged at the lower part of the traffic signal lamp post at the single intersection; the light intensity sensor (011-3), the humidity sensor (011-2) and the temperature sensor (011-1) are all connected with the storage unit (020);
the traffic flow image acquisition module (012) comprises a plurality of cameras (091) for acquiring vehicle image information of each direction of a single intersection;
the pedestrian flow image acquisition module (013) is used for acquiring pedestrian traffic image information of sidewalks;
the time acquisition module (014) is used for recording the time period corresponding to the information;
the special event acquisition module (015) is arranged in the cloud server database module (042) and is used for acquiring important festivals, activities and special vehicle passing information of a current city and area through the Internet;
the traffic prediction unit (050) based on deep learning comprises a database information reading module (051), a traffic prediction model training module (052) and a prediction result storage module (053), wherein the database information reading module (051) is used for automatically calling and storing various types of traffic data and pedestrian flow data stored in the cloud data processing and database unit (040) as well as environment information and special event data at corresponding moments; the traffic prediction model training module (052) is used for establishing a plurality of prediction models based on deep learning, and calling data in the database information reading module (051) to continuously train the plurality of established prediction models based on deep learning to obtain a regional road network traffic prediction result; the prediction result storage module (053) is used for storing the prediction result of the regional road network traffic, which is obtained by the traffic prediction model training module (052);
the Synchro-based simulation computing unit (060) comprises a storage module (061), a main intersection screening module (062) and an intersection signal timing optimization module (063), wherein the main intersection screening module (062) is used for determining a current traffic signal control target according to current temperature, humidity, visibility and information related to special events; the intersection signal timing optimization module (063) is used for calling the regional road network model of the regional road network model building unit (070) and carrying out synchronous simulation calculation on a control target of the main intersection screening module (062) to obtain the control duration of the traffic signal; the storage module (061) is used for receiving road network model files of the regional road network model establishing unit (070), regional road network traffic prediction results predicted by the deep learning traffic prediction unit (050), control target data of the main intersection screening module (062) and calculation results of the intersection signal timing optimization module (063).
2. The regional traffic signal control system based on deep learning of claim 1, wherein the storage unit (020) comprises a single intersection signal lamp storage module (021) and a simulation calculation control result storage module (022); the single intersection signal lamp storage module (021) is used for storing and numbering all information of the weather acquisition module (011), the traffic flow image acquisition module (012), the pedestrian flow image acquisition module (013), the time acquisition module (014) and the special event acquisition module (015); the simulation calculation control result storage module (022) is used for storing the control information downloaded by the 5G communication unit (030).
3. The regionalized traffic signal control system based on deep learning of claim 1, characterized in that the 5G communication unit (030) is arranged on top of a single intersection signal post; the 5G communication unit (030) comprises a data uploading module (031) and a data downloading module (032), the data uploading module (031) is connected with the storage unit (020), the data downloading module (032) is connected with the Synchro-based simulation computing unit (060), and the data uploading module (031) is used for uploading data; the data downloading module (032) is used for data downloading.
4. The regional traffic signal control system based on deep learning of claim 1, wherein the cloud data processing and database unit (040) comprises an information and image processing module (041) and a database module (042), the information and image processing module (041) is configured to implement vehicle identification and face identification processing on a summarized image received by a cloud based on a video detection technology, obtain traffic flow data of different vehicle types in corresponding time periods and pedestrian numbers in different age periods of a sidewalk waiting area, and summarize road junction data by corresponding numbers to form the regional road network data;
and the database module (042) is used for receiving and storing various traffic volume data, pedestrian flow data, environment information at corresponding moments and special event data which are obtained by the information and image processing module (041).
5. The regional traffic signal control system based on deep learning of claim 1, wherein the regional road network model building unit (070) comprises a road information measuring module (071) and a model building module (072), the road information measuring module (071) is used for oblique photography of the road network of the control region, collecting image data, and measuring road length, number of lanes, lane width, number of entrance lanes, and position of a stop line of an entrance lane; the model building module (072) is used for building a plane and three-dimensional road network model by combining the image data, the road length, the lane number, the lane width and the lane number conversion position data of the inlet road collected by the road information measuring module (071) with the image information of the satellite map.
6. The regionalized traffic signal control system based on deep learning of claim 1, characterized in that the control unit (080) comprises an instruction receiving and processing module (081), a signal light automatic control module (082) and a display screen control module (083), the instruction receiving and processing module (081) is used for receiving information generated by the signal light automatic control module (082) and correspondingly matching all intersection signal light control information in the road network with the signal light number, the signal light automatic control module (082) is used for controlling the color and the time length of the traffic signal, and the display screen control module (083) is used for implementing the intersection signal light control information and simultaneously displaying the traffic volume information according to the information of the instruction receiving and processing module (081).
7. The control method of the regional traffic signal control system based on the deep learning as claimed in claim 1, wherein the information acquisition unit (010) is used for acquiring information of a single intersection, the acquired information is stored through the storage unit (020), the 5G communication unit (030) is used for uploading the acquired information to the cloud data processing and database unit (040) in the cloud server, the information is processed through the information and image processing module (041) and stored in the database module (042), the processed information data is read and imported through the database information reading module (051) of the traffic prediction unit (050) based on the deep learning, the traffic prediction model training module (052) is used for predicting, and the prediction result is stored in the prediction result storage module (053), and then reading the regional road network model established by the predicted traffic volume and regional road network model establishing unit (070) through a storage module (061), optimizing and simulating the control target through a main intersection screening module (062) and an intersection signal timing optimization module (063), finally obtaining control information, storing the control information in the storage module (061), downloading the control information through a 5G communication unit (030), and finally receiving the control information and controlling red and green lamps of each road through a control unit (080), thereby realizing the management and control of traffic signals in the road network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910610269.4A CN110349407B (en) | 2019-07-08 | 2019-07-08 | Regional traffic signal lamp control system and method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910610269.4A CN110349407B (en) | 2019-07-08 | 2019-07-08 | Regional traffic signal lamp control system and method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110349407A CN110349407A (en) | 2019-10-18 |
| CN110349407B true CN110349407B (en) | 2021-08-13 |
Family
ID=68178341
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910610269.4A Active CN110349407B (en) | 2019-07-08 | 2019-07-08 | Regional traffic signal lamp control system and method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110349407B (en) |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111798675B (en) * | 2020-07-02 | 2022-09-13 | 南京华捷艾米软件科技有限公司 | Signal lamp control method and system |
| CN111882887B (en) * | 2020-07-16 | 2021-11-23 | 浙江工业大学 | Method for synchronously displaying SCATS phase signals and integrating monitoring data of flow equipment |
| CN112115007B (en) * | 2020-08-26 | 2023-04-21 | 深圳市黑马启航科技有限公司 | Debugging method and device of countdown display |
| CN114882696B (en) * | 2020-10-28 | 2023-11-03 | 华为技术有限公司 | Road capacity determination method, device and storage medium |
| CN112785859B (en) * | 2020-12-17 | 2022-01-25 | 漳州职业技术学院 | System and method for controlling uniform timing of signal lamps in urban area |
| CN112785840B (en) * | 2020-12-17 | 2022-03-08 | 漳州职业技术学院 | Urban intelligent traffic computer control system and method |
| CN113192334B (en) * | 2021-04-30 | 2023-05-16 | 山东高速建设管理集团有限公司 | System for warning constructor and using method thereof |
| CN113283341A (en) * | 2021-05-25 | 2021-08-20 | 华北理工大学 | Data analysis terminal device |
| CN113643528A (en) * | 2021-07-01 | 2021-11-12 | 腾讯科技（深圳）有限公司 | Signal lamp control method, model training method, system, device and storage medium |
| CN113658437A (en) * | 2021-10-20 | 2021-11-16 | 枣庄智博智能科技有限公司 | Traffic signal lamp control system |
| CN114241762B (en) * | 2021-12-14 | 2022-11-15 | 深圳市鹏城交通网络股份有限公司 | Smart city traffic comprehensive data acquisition and analysis system based on Internet of things |
| CN114973661B (en) * | 2022-05-16 | 2024-05-10 | 深圳市中建恒峰电子有限公司 | Traffic signal lamp with intelligent control and traffic flow data storage |
| CN114973672A (en) * | 2022-05-24 | 2022-08-30 | 上海富骏交通科技有限公司 | Holographic road network traffic signal control system and method based on deep reinforcement learning |
| CN114882707A (en) * | 2022-07-05 | 2022-08-09 | 华录易云科技有限公司 | Crossing traffic signal control system of high in clouds |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1936999A (en) * | 2006-10-17 | 2007-03-28 | 大连理工大学 | City area-traffic cooperative control method based wireless sensor network |
| WO2010042973A1 (en) * | 2008-10-15 | 2010-04-22 | National Ict Australia Limited | Tracking the number of vehicles in a queue |
| CN102906800A (en) * | 2010-02-01 | 2013-01-30 | 影视技术集成公司 | System and method for modeling and optimizing the performance of transportation networks |
| CN103456180A (en) * | 2013-07-18 | 2013-12-18 | 吴建平 | Signal control system and signal control method based on traffic simulation |
| CN107705557A (en) * | 2017-09-04 | 2018-02-16 | 清华大学 | Road network signal control method and device based on depth enhancing network |
| CN108389403A (en) * | 2018-04-23 | 2018-08-10 | 白冰 | The traffic lights and speed(-)limit sign intelligent adjusting method captured based on vehicle flowrate |
| CN108417057A (en) * | 2018-05-15 | 2018-08-17 | 哈尔滨工业大学 | A kind of intelligent signal lamp timing system |
| CN109410606A (en) * | 2018-03-22 | 2019-03-01 | 合肥革绿信息科技有限公司 | A kind of trunk roads synergistic signal machine control method based on video |
| CN109493617A (en) * | 2018-10-29 | 2019-03-19 | 沈阳天久信息技术工程有限公司 | A kind of traffic signal optimization control method and device |
-
2019
- 2019-07-08 CN CN201910610269.4A patent/CN110349407B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1936999A (en) * | 2006-10-17 | 2007-03-28 | 大连理工大学 | City area-traffic cooperative control method based wireless sensor network |
| WO2010042973A1 (en) * | 2008-10-15 | 2010-04-22 | National Ict Australia Limited | Tracking the number of vehicles in a queue |
| CN102906800A (en) * | 2010-02-01 | 2013-01-30 | 影视技术集成公司 | System and method for modeling and optimizing the performance of transportation networks |
| CN103456180A (en) * | 2013-07-18 | 2013-12-18 | 吴建平 | Signal control system and signal control method based on traffic simulation |
| CN107705557A (en) * | 2017-09-04 | 2018-02-16 | 清华大学 | Road network signal control method and device based on depth enhancing network |
| CN109410606A (en) * | 2018-03-22 | 2019-03-01 | 合肥革绿信息科技有限公司 | A kind of trunk roads synergistic signal machine control method based on video |
| CN108389403A (en) * | 2018-04-23 | 2018-08-10 | 白冰 | The traffic lights and speed(-)limit sign intelligent adjusting method captured based on vehicle flowrate |
| CN108417057A (en) * | 2018-05-15 | 2018-08-17 | 哈尔滨工业大学 | A kind of intelligent signal lamp timing system |
| CN109493617A (en) * | 2018-10-29 | 2019-03-19 | 沈阳天久信息技术工程有限公司 | A kind of traffic signal optimization control method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110349407A (en) | 2019-10-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110349407B (en) | Regional traffic signal lamp control system and method based on deep learning | |
| KR102021992B1 (en) | Apparatus for controling a trafic signal, method for controling a trafic signal, and recoding medium for controling a tarfic signal | |
| CN114418468B (en) | Smart city traffic scheduling strategy control method and Internet of things system | |
| CN110111592B (en) | Method for dynamically matching optimal signal timing scheme based on traffic signal controller | |
| CN110517487B (en) | Urban area traffic resource regulation and control method and system based on thermodynamic diagram change identification | |
| CN106652483B (en) | The method for laying traffic information test point in regional highway network using detection device | |
| CN110969871B (en) | Intelligent traffic light control system and control method | |
| JP2021531462A (en) | Intelligent navigation methods and systems based on topology maps | |
| CN112437501B (en) | Multi-sensor beyond-the-horizon ad hoc network method based on traffic semantics and game theory | |
| CN108665715A (en) | A kind of road junction intelligent traffic is studied and judged and signal optimizing method | |
| CN107067759A (en) | A kind of parallel traffic lights real-time control method based on ACP | |
| CN113096418B (en) | Traffic network traffic light control method, system and computer readable storage medium | |
| CN102024330A (en) | Intelligent traffic signal control system, method and equipment based on high-definition video technology | |
| CN111027430A (en) | Traffic scene complexity calculation method for intelligent evaluation of unmanned vehicles | |
| CN114819305B (en) | Path planning method based on carbon emission measurement scale | |
| CN111161545B (en) | Intersection region traffic parameter statistical method based on video | |
| CN110287995B (en) | Multi-feature learning network model method for grading all-day overhead traffic jam conditions | |
| CN115687709A (en) | Traffic dynamic control method based on traffic data dimension reduction reconstruction and multidimensional analysis | |
| CN109410608B (en) | Picture self-learning traffic signal control method based on convolutional neural network | |
| CN113140108B (en) | Cloud traffic situation prediction method in internet-connected intelligent traffic system | |
| CN116415756A (en) | Urban virtual scene experience management system based on VR technology | |
| CN113505346B (en) | Urban street lamp data processing and combined regulation and control system based on artificial intelligence | |
| CN115188192A (en) | Automatic control parking system and method based on travel prediction | |
| CN113808404A (en) | Fixed-time-length traffic signal lamp time length prediction method | |
| Guo et al. | Research on optimization model of multisource traffic information collection combination based on genetic algorithm |
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 |