CN110494902A - For managing the system, apparatus and method of the traffic in geographical location - Google Patents

For managing the system, apparatus and method of the traffic in geographical location Download PDF

Info

Publication number
CN110494902A
CN110494902A CN201780088995.3A CN201780088995A CN110494902A CN 110494902 A CN110494902 A CN 110494902A CN 201780088995 A CN201780088995 A CN 201780088995A CN 110494902 A CN110494902 A CN 110494902A
Authority
CN
China
Prior art keywords
traffic
intersection
environment
density
signal profile
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201780088995.3A
Other languages
Chinese (zh)
Inventor
S.施皮克曼
V.苏德哈卡兰
L.坦贾
V.拉文德兰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yunnix Co ltd
Original Assignee
Siemens Mobility GmbH
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Siemens Mobility GmbH filed Critical Siemens Mobility GmbH
Publication of CN110494902A publication Critical patent/CN110494902A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/07—Controlling traffic signals
    • G08G1/08—Controlling traffic signals according to detected number or speed of vehicles
    • 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
    • G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
    • G08G1/0112—Measuring and analyzing of parameters relative to traffic conditions based on the source of data from the vehicle, e.g. floating car data [FCD]
    • 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
    • G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
    • G08G1/0116—Measuring and analyzing of parameters relative to traffic conditions based on the source of data from roadside infrastructure, e.g. beacons
    • 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
    • G08G1/04—Detecting movement of traffic to be counted or controlled using optical or ultrasonic detectors
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/07—Controlling traffic signals
    • G08G1/081—Plural intersections under common control

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Traffic Control Systems (AREA)

Abstract

Provide a kind of method for managing the traffic flow in geographical location.This method include based at least one intersection (130) traffic environment in real time determine traffic density, wherein traffic environment includes vehicular traffic and pedestrian traffic.This method includes based on traffic environment associated at least one intersection (130) and historical traffic environment come prediction traffic density.In addition, this method includes based on traffic density and forecasting that traffic density selects the traffic signals profile of at least one intersection (130).Traffic flow in traffic signals profile management geographical location based at least one intersection (130).

Description

System, device and method for managing traffic in a geographic location
As the number of vehicles has increased, managing the flow of traffic on roads has become a daunting task. Generally, traffic flow is managed at intersections with traffic sign posts having static signal times to control the traffic flow. An increase in vehicle or traffic density results in significant traffic congestion because static signal times do not account for variations in traffic density. In addition, several factors, such as reckless or unstable driving and poor road conditions contribute to increased traffic congestion and lead to a cluttered traffic environment.
To manage traffic flow in situations where traffic density varies, intersections are remotely managed using human agents (i.e., police) either personally or by using cameras. This method is limited to controlling only one intersection and is difficult to coordinate with other intersections.
To address scalability issues, a large volume of traffic data may be processed using centralized traffic control. Traffic data is collected from several sensors, such as inductive loops, wireless ground sensors, passive infrared detectors, high resolution camera systems, and radar. These sensors are not suitable for chaotic traffic environments where lane discipline is maintained without a vehicle. Additionally, centralized traffic control is expensive to implement in view of the large amount of traffic data that needs to be processed.
Decentralized traffic control may be used to manage traffic to reduce the complexity of centralized traffic control. Such a decentralized traffic control system and method is disclosed in US 20130176146. The method describes a system of agents in a traffic environment, where each agent represents a traffic light controller at an intersection to control the flow of traffic. Each agent collects data local to the junction using one or more sensors and applies a distributed W learning model to adapt to changing traffic conditions in a decentralized, distributed manner. However, this approach makes assumptions based on sensor data used to collect traffic information. This assumption of traffic flow may not be appropriate in situations where a chaotic traffic environment exists.
In view of the above, there is a need for accurately managing traffic flow in geographic locations having cluttered traffic environments. Accordingly, it is an object of the present invention to provide a method, apparatus and system for efficiently managing traffic flow in a chaotic traffic environment.
The method, apparatus and system according to the present invention achieve the above-mentioned objects by detecting traffic density at an intersection based on traffic environment at the intersection and predicting a forecast traffic density. The present invention also teaches selecting a traffic signal contour for the intersection based on the traffic density and the forecasted traffic density, and managing traffic flow in the geographic location based on the traffic signal contour for the intersection.
As used herein, "traffic environment" indicates vehicular traffic and pedestrian traffic for geographic locations in which traffic flow is to be managed. Thus, the traffic environment includes coordinates of the geographic location and data associated with the weather, air quality, time, day, scheduled events, and unscheduled events of the geographic location. The term "geographical location" includes a plurality of intersections and, thus, the traffic environment also includes information associated with each intersection, such as lane closures or road maintenance. Further, "intersection" means a traffic intersection having manned or unmanned traffic signal poles for managing the flow of traffic at the intersection. Each intersection also includes one or more lanes whose flow of traffic is managed by traffic signs.
In accordance with the present invention, a computer-implemented method of managing traffic flow for a traffic environment including vehicular traffic and pedestrian traffic is provided. The method includes determining traffic density in real-time based on traffic environment at one or more intersections. As used herein, "traffic density" includes the number of vehicles of each vehicle type, the number of private vehicles, the number of public vehicles, and passenger occupancy. Since the traffic environment includes information related to vehicle traffic and pedestrian traffic, the traffic environment is used to determine traffic density. For example, traffic density is determined in real time by identifying traffic objects such as cars, scooters, bicycles, trams, etc. in a traffic environment.
The method also includes predicting a forecasted traffic density based on the traffic environment associated with the intersection and the historical traffic environment. As used herein, "historical traffic environment" includes information related to traffic at an intersection for a past instance of time. For example, to predict the forecasted traffic density at time "t", the historical traffic environment includes information about vehicle traffic, pedestrian traffic, and data associated with weather, air quality, time, day, scheduled events, and unscheduled events for the geographic location captured at time "t-x". The historical environment also includes traffic density at time t-x. The forecasted traffic density includes the number of vehicles of each vehicle type, the number of private vehicles, the number of public vehicles, and the passenger occupancy.
Further, the method includes selecting a traffic signal contour for the intersection based on the real-time traffic density and the forecasted traffic density. As used herein, a "traffic signal profile" includes one of a red, yellow, and green color profile to indicate that traffic in a lane should stop, cautiously advance, and advance, respectively. The traffic signal profile also indicates the signal period time of the next green signal. A traffic signal profile is selected by comparing the forecasted traffic density to the traffic density. If the traffic density differs significantly from the forecast traffic density, the traffic signal profile is recalculated. In other words, the signal cycle time of the next green time increases or decreases for the lanes in the intersection. In one embodiment, the forecasted traffic density is used to generate a traffic signal profile when the traffic density cannot be determined in real-time. For example, when weather conditions result in poor visibility of traffic at an intersection, the forecasted traffic density is used to generate a traffic signal profile. In another example, a capture device used to capture traffic density at an intersection may be faulty, in which case the forecasted traffic density is used to generate a traffic signal profile.
Further, the method includes managing a traffic flow of the traffic environment based on the traffic signal profile of the intersection. The traffic signal profile selected for the intersection is transmitted in real time to adjacent intersections in the geographic location. In one embodiment, the traffic signal profile of an intersection is communicated by means of a wireless communication network and a transmitter associated with the intersection. The traffic signal contour of the neighboring intersection is determined by performing the above steps for each of the intersections based on the traffic signal contour of the intersection. This is further communicated to intersections that are adjacent to neighboring intersections and is repeated until the traffic signal profiles for all intersections in the geographic location are calculated and a coordinated green color is achieved. The term "coordinated green" indicates a synchronized green signal profile across an intersection in a geographic location such that traffic flow is continuous with minimal latency. The coordinated green is an optimization problem as follows: coordinated green signals across the intersection as much as possible are achieved based on various parameters, such as time of day and forecasted traffic density, to ensure smooth traffic flow. Additionally, the coordinated green indication maximizes the distance traveled before having to wait at the intersection. The method effectively achieves coordinated green in a decentralized manner by determining a traffic signal profile at each of the intersections based on the traffic signal profiles of adjacent intersections.
According to an embodiment of the invention, the method includes capturing traffic environment in real-time for an intersection and determining lane occupancy of two or more lanes associated with the intersection based on the traffic environment. Multiple capture devices may be used to capture the traffic environment in real time. In one embodiment, the traffic environment captured by means of a remote server is forwarded in real time. In another embodiment, the captured traffic environment is normalized to ensure that lane occupancy is accurately determined regardless of weather or light conditions. The captured traffic environment is analyzed to determine lane occupancy. The analysis may be performed by using a neural network on the captured image. In an exemplary embodiment, a convolutional neural network is used to analyze the traffic environment. A convolutional neural network is advantageous because it can accurately determine lane occupancy in chaotic traffic environments.
The method also includes determining a traffic object in the traffic environment based on the predetermined identification model. As used herein, "traffic object" includes vehicle type, vehicle occupancy, and pedestrian type. For example, the transportation object includes, but is not limited to, an automobile, a bicycle, a scooter, a tricycle, a bus, a public transportation vehicle, an ambulance, and a vehicle for a disabled person. Further, the traffic object includes pedestrian types such as elderly pedestrians, physically disabled pedestrians, visually impaired pedestrians, children's pedestrians, and animals. The predetermined identification model comprises a training data set in which valid vectors are to be positively identified and opposite examples are rejected. At the end of the training, an appropriate number of neurons with appropriate fields of influence are trained to recognize objects, patterns, and decisions required to identify traffic objects. Thus, using a predetermined identification model results in back propagation, in which the weighted inputs to the neurons are determined by an iterative process. The predetermined identification model leads to a decision, which is defined by a prototype example of correct recognition of the traffic object.
The method also includes determining a traffic anomaly associable with the intersection. As used herein, "traffic anomaly" means an accident, vehicle malfunction, traffic violation, and the like. In one embodiment, the traffic anomaly may be at a location that may cause an irregular flow of traffic at an intersection. For example, an accident at an adjacent intersection may cause traffic on one lane to change course. Therefore, the traffic signal profile of the lane must be altered based on the change in traffic on the lane. Thus, the method is advantageous in that it is capable of quickly adapting to changes in the traffic environment and selecting a traffic signal profile of the intersection accordingly.
According to another embodiment, a traffic environment is captured in real-time using one or more capture devices placed at a plurality of capture points. The capture device may be placed on street lights, drones, traffic signs at intersections, street fences. By distributing the capture devices along the lanes of the intersection, the traffic environment is captured evenly across the lanes of the intersection. This enables an accurate determination of the traffic density at the intersection. As used herein, a capture device may be a manned or unmanned device equipped with an image sensor, a motion sensor, a Global Positioning System (GPS) device. For example, the capture device may be a camera located on a traffic sign pole, street light, drone, or the like. In another example, the capture device may be a mobile computing device with a GPS sensor that is capable of relaying details of the latency time at a location due to heavy traffic congestion.
According to yet another embodiment, the method includes capturing media data that includes social media corresponding to an intersection. For example, media and social media sites provide real-time information of traffic at various geographic locations. The server parses through the media data to detect whether there is a traffic anomaly that can be associated with the intersection. The information is transmitted from the server to a computing device located at the intersection, which selects a traffic signal profile for the intersection.
According to one embodiment, the method includes generating a traffic image of a traffic environment. The traffic image is a representation of a traffic environment. The traffic images are processed to determine traffic density, lane occupancy, and traffic objects. The method also includes identifying traffic objects independent of vehicle location, relative vehicle proximity, and pedestrian location based on the predetermined identification model. In chaotic traffic environments, a predetermined identification model is trained taking into account vehicle location and relative vehicle proximity. For example, in a cluttered traffic environment, inductive loops provided on the road will not be able to accurately determine traffic density. The predetermined identification model maps traffic objects by using a neural network. This enables accurate identification of traffic objects and thereby ensures that traffic density is correctly estimated.
In an exemplary embodiment, the predetermined identification model is trained by inputting a set of parameters to one or more nodes of an input layer of a neural network (such as a convolutional neural network). The set of parameters includes identification parameters to identify the vehicle and the pedestrian, such as vehicle shape, pedestrian shape, vehicle position, pedestrian position, vehicle height, vehicle width, vehicle depth, and pedestrian height. Traffic objects are identified by calculating at each node of the input layer the product of the set of parameters and the associated weighting factor, and calculating the sum of the products of the set of parameters and the weighting factor at the respective node of the first intermediate layer of the neural network.
Further, the traffic object is identified by calculating a product of a sum of products and an associated weighting coefficient at each node of the intermediate layer. As the identification parameters increase, additional intermediate layers are added to the neural network to accurately identify traffic objects. Thereafter, the traffic object is identified by correlating the sum of the products using a look-up table.
In another embodiment, the predetermined identification model is an untrained model generated using a generative confrontation network (GAN) or a deep convolutional generative confrontation network (DCGAN). GAN/DCGAN attempts to learn the joint probabilities of traffic images and traffic objects simultaneously. The GAN/DCGAN model is advantageous because the underlying structure of the traffic image is mapped even when there are no identifying parameters. This is desirable when a new vehicle type is captured at the intersection.
According to yet another embodiment, the method includes communicating the traffic image through a plurality of layers of a neural network, wherein the plurality of layers includes a convolutional layer and a pooling layer. In one embodiment, the traffic image is represented as a three-dimensional array of pixels with intensity for each parameter (such as color, height, width). The three-dimensional array is transformed by the convolutional and pooling layers using a maximum function. The maximum function is used to aggregate the maximum of the spatial region across the convolutional layer.
According to one embodiment, the method includes comparing the traffic density to a forecast traffic density to generate a traffic difference. The traffic difference is compared to a threshold to determine whether the traffic difference has exceeded the threshold. The threshold is determined based on the traffic environment of the intersection. If the traffic difference is within a threshold, a traffic signal profile for the intersection is selected based on the forecasted traffic density. For example, if traffic at an intersection is 18 on every monday of the first three weeks: and severe congestion at 00, determining a forecast traffic density based on the historical traffic environment, wherein the forecast traffic density implies that traffic is at 18: and will be heavily congested at 00 f. Selecting 18 based on the forecasted traffic density: traffic signal profile at 00 deg.f. At 18: at 00 hours, the traffic density is determined in real time and compared to the forecast traffic density. If the traffic density is comparable to the forecasted traffic density, then a traffic signal profile selected based on the forecasted traffic density is used. If the traffic density is different from the forecast traffic density, the traffic signal profile is recalculated.
According to another embodiment, the method comprises: if the traffic difference exceeds a threshold, an optimal traffic signal profile for the intersection is determined. The optimal traffic signal profile is based on traffic differences. In one embodiment, a look-up table is used to arrange the optimal traffic signal profile and traffic differences. The lookup table is generated based on the historical traffic environment of the intersection.
The optimal traffic signal profile is tested in real time by identifying whether a traffic anomaly has occurred at the intersection. For example, if the optimal traffic signal profile has a signal cycle time of 60 seconds before the next green color, and a traffic anomaly is identified at the intersection. The traffic density is checked to see if there is traffic congestion. Thus, it may be determined that a traffic anomaly at an intersection is likely to be associated with an optimal traffic signal profile. If a traffic anomaly is identified, the optimal traffic signal profile is recalculated. If no traffic anomaly is identified, the best traffic signal profile is selected as the traffic signal profile.
According to one embodiment, the method includes transferring a traffic signal contour of an intersection to an adjacent intersection. Thereafter, based on the transmitted traffic signal profile of the intersection, a traffic signal profile of an adjacent intersection is generated. This step is repeated across all intersections in the geographic location. Accordingly, the method includes iteratively generating traffic signal profiles for a plurality of intersections based on traffic signal profiles for adjacent intersections. Further, the method includes managing traffic flow for the plurality of intersections based on traffic signal contours of adjacent intersections by generating coordinated green colors for the geographic locations. The method effectively achieves coordinated green in a decentralized manner by determining a traffic signal profile at each of the intersections based on the traffic signal profiles of adjacent intersections.
Also disclosed in accordance with the present invention are computing devices and systems for managing traffic flow in a geographic location. The computing device includes one or more processors, a memory coupled to the processors, a communication unit, and a database. The memory is configured to store computer program instructions defined by modules, such as a traffic density estimator, a traffic prediction module, a contour determination module, and the like. Each processor is configured to execute instructions in memory. The memory may be volatile memory such as Random Access Memory (RAM), static RAM (sram), dynamic RAM (dram), or any other memory accessible by the processor. The memory may operate as a routing table, a register, virtual memory, and/or a swap space for the processor. Further, the storage module may represent any known or later developed non-volatile storage device, such as, but not limited to, a hard drive, a Solid State Drive (SSD), a flash ROM (read only memory), or an optical drive.
In one embodiment, the computing device includes an edge device and a capture component communicatively coupled to the edge device. Edge devices are compact computing devices with resource constraints in terms of computing power.
The computing device includes a traffic density estimator configured to determine traffic density in real-time based on a traffic environment of the intersection, and a traffic prediction module configured to predict a forecasted traffic density based on the traffic environment associated with the intersection and a historical traffic environment. Further, the computing device includes a contour determination module configured to select a traffic signal contour for the intersection based on the traffic density and the forecasted traffic density.
The communication unit of the computing device is capable of communicating a traffic signal contour of the intersection to a traffic sign post associated with the intersection for managing a flow of traffic of the traffic environment. Thus, the computing device is advantageous because it can efficiently manage traffic flow in a traffic environment.
According to one embodiment, the computing device includes a traffic analyzer configured to determine lane occupancy of two or more lanes associated with an intersection based on a traffic environment. The computing device also includes a traffic image generator configured to generate a traffic image of the traffic environment and a traffic object determination module configured to determine a traffic object in the traffic environment based on the predetermined identification model.
According to another embodiment, the traffic object determination module includes a neural network module configured to communicate traffic images through a plurality of layers of a neural network, wherein the plurality of layers includes convolutional layers and max-pooling layers.
According to one embodiment, the traffic analyzer includes a comparator configured to compare the traffic density to a forecast traffic density to generate a traffic difference. Further, the traffic analyzer includes a threshold module to determine whether the traffic difference has exceeded a threshold, wherein the threshold is determined based on the traffic environment of the intersection.
According to another embodiment, a traffic analyzer comprises an optimal profile module configured to: if the traffic difference exceeds a threshold, an optimal traffic signal profile for at least one intersection is determined. The traffic analyzer further includes an anomaly identifier configured to identify a traffic anomaly at the at least one intersection associable to the optimal traffic signal profile determined by the anomaly identifier. If a traffic anomaly is identified, the optimal traffic signal profile is re-determined, and wherein if no traffic anomaly is identified, the optimal traffic signal profile is selected as the traffic signal profile.
According to one embodiment, the communication unit includes a transmitter communicatively coupled to the processor to transmit a traffic signal profile of the intersection to an adjacent intersection. The communication unit also includes a receiver communicatively coupled to the processing to receive traffic signal profiles of adjacent intersections, wherein the traffic signal profile of the at least one intersection is based on the traffic signal profiles of the adjacent intersections.
The present invention also provides a system for managing traffic flow in a geographic location. The system includes a server, a network communicatively coupled to the server, and a computing device communicatively coupled to the server via the network. The computing device is associated with an intersection in a traffic environment. The computing device can manage traffic flow at the intersection based on the traffic environment at the intersection.
According to one embodiment, the system includes a plurality of capture devices communicatively coupled to a server via a network, the plurality of capture devices being one of selected manned devices and unmanned devices including: image sensors, motion sensors, GPS devices, and communication devices. The plurality of capture devices are capable of capturing media data that includes social media corresponding to the at least one intersection.
According to another embodiment, the server includes a database storing data associated with a traffic environment, wherein the traffic environment includes data associated with: weather, air quality, time, day, scheduled events, and unscheduled events for at least one geographic location including the plurality of intersections.
According to a further embodiment, the server comprises a traffic model generator to generate a predetermined identification model based on the traffic environment. The predetermined identification model comprises a training data set in which valid vectors are to be positively identified and opposite examples are rejected, at the end of which a suitable number of neurons with suitable influence fields are trained to identify objects, patterns and decisions required for identifying traffic objects. Thus, using a predetermined identification model results in back propagation, where the weighted inputs to the neurons are determined by an iterative process. The predetermined identification model leads to a decision, which is defined by a prototype example of correct recognition of the traffic object.
The server also includes a forecasting model generator to generate a forecasting model based on the historical traffic environment. The historical traffic environment includes information related to traffic at the intersection for a past instance of time. In one embodiment, the predictive model generator uses a neural network, such as a Recurrent Neural Network (RNN), to generate the predictive model. An advantage of a recurrent neural network is that the RNN maintains an internal state that is used to predict sequence values of the predicted traffic density. In RNN, the current value of the forecasted traffic density depends on the previous state. For example, if there is traffic congestion at an intersection due to a nearby accident or an unplanned aggregate/event, the forecasted traffic density in the next instance in time will be highly dependent on the current instance in time. Thus, it can be observed that the predicted traffic density at the intersection at time t +1 has a strong correlation with the density at time t, and therefore, RNN is used to accurately predict the predicted traffic density.
The above-mentioned and other features of the present invention will now be presented with reference to the drawings of the present invention. The illustrated embodiments are intended to illustrate, but not to limit the invention.
The invention is further described hereinafter with reference to the illustrated embodiments shown in the drawings, in which:
FIG. 1 is a block diagram of a system for managing a geographically located traffic flow;
FIG. 2 is a block diagram of an edge device for managing traffic at an intersection;
FIG. 3 illustrates the determination of lane occupancy at an intersection;
FIG. 4 illustrates identification of a traffic object by the edge device of FIG. 2;
FIG. 5 illustrates generation of a predictive model by the system of FIG. 1;
FIG. 6 is a block diagram of a system for managing traffic flow at multiple intersections;
FIG. 7 illustrates a method of managing traffic flow at a plurality of intersections using the system according to FIG. 6; and
FIG. 8 illustrates a method of managing traffic flow in a traffic environment.
Various embodiments are described with reference to the drawings, wherein like reference numerals are used to refer to like elements throughout. Furthermore, numerous specific details are set forth in order to provide a thorough understanding of one or more embodiments of the present invention. These examples should not be considered as limiting the application of the invention to the configurations disclosed in the figures. It may be evident that such embodiment(s) may be practiced without these specific details.
FIG. 1 is a block diagram of a system 100 for managing a traffic flow including a geographic location of an intersection 130. As shown in fig. 1, traffic at intersection 130 includes vehicular traffic and pedestrian traffic. Intersection 130 connects to lanes 132, 134, 136, and 138. The intersection 130 includes a traffic sign post 140 to indicate whether the traffic at the lanes 132, 134, 136, and 138 should stop (red), proceed cautiously (yellow), or proceed green. In addition, the intersection 130 is equipped with a camera 120 to capture the traffic environment associated with the intersection 130.
The system 100 includes a server 102, a network 108, a database 110, a camera 120, and a computing device 150. The system 100 also includes computing devices 152 and 154 associated with adjacent intersections.
As shown in fig. 1, server 102 is communicatively coupled to database 110. The database 110 is, for example, a Structured Query Language (SQL) data store or not only an SQL (nosql) data store. In one embodiment, the database 110 may be a location on a file system that is directly accessible by the server 102. In another embodiment, database 110 may be configured as a cloud-based database implemented in a cloud computing environment, where computing resources are delivered as services over network 108. As used herein, a "cloud computing environment" refers to a processing environment that includes configurable computing physical and logical resources (e.g., networks, servers, storage, applications, services, etc.) and data distributed over a network 108 (e.g., the internet). A cloud computing environment provides on-demand network access to a shared pool of configurable computing physical and logical resources. The network 108 is, for example, a wired network, a wireless network, a communication network, or a network formed by any combination of these networks.
The server 102 includes a controller 104 and a memory 106. Server 102 is communicatively coupled to network 108. The memory 106 is configured to store computer program instructions defined by modules, such as a traffic model generator 112 and a forecast model generator 114. In the present embodiment, modules 112 and 114 are implemented on a cloud computing environment, as indicated in FIG. 1.
Further, fig. 1 shows that the computing device 150 includes a traffic object determination module 162, a traffic density estimator 164, and a contour determination module 166. The traffic object determination module 162 receives the predetermined identification model from the traffic object generator 112, as indicated by arrow 144. The traffic object generator module 162 uses the captured traffic environment from the camera 120 to determine traffic objects, as indicated by arrow 121. The traffic density estimator 164 receives information from the capture devices 122 and 124 located at the intersection, as indicated by arrow 142. In this example, the capture device 122 is an automotive Global Positioning System (GPS) device 122. The capture device 124 is a mobile computing device 124 that may be used to transmit real-time traffic data to the computing device 150, either directly or via the server 110. The contour determination module 166 receives real-time traffic density from the traffic density estimator 164. The contour determination module 166 also receives real-time data associated with the traffic environment at the intersection 130, as indicated by the arrow 146. For example, the contour determination module 166 receives real-time data from the server 110 relating to the weather and pollution levels at the intersection 130. The contour determination module 166 compares the real-time traffic density against a forecasted traffic density (as indicated by arrow 148) that is based on the forecast model received from the forecast model generator 114. The contour determination module 166 selects a traffic signal contour for the intersection 130 based on the traffic density and the forecasted traffic density. The selected traffic signal profile is reflected on the traffic sign post 140 as indicated by arrow 160. The computing device 150 includes other components, such as a processor, memory, and communication units, as described in detail in fig. 2.
Fig. 2 is a block diagram of an edge device 200 for managing traffic at the intersection 130. The edge device 200 may be used in place of the computing device 150. The edge device 200 has a small form factor that can be connected to the network 108. The edge device 200 includes a processor 202, a memory 204, and a communication unit 208. Memory 204 may include 2 gigabyte stacked package (PoP) stacked Random Access Memory (RAM) and flash memory devices. The communication unit 208 includes a transmitter 248, a receiver 258b, and a gigabit ethernet port. The edge device 200 also includes a high-definition multimedia interface (HDMI) display 206 and a cooling fan (not shown).
The memory 204 of the edge device 200 is equipped with modules, e.g., 210, 212, 214, 216, 218, and 220, stored in the form of computer-readable instructions. The processor 202 is configured to execute defined computer program instructions in the module. Further, the processor 202 is configured to simultaneously execute instructions in the memory 204.
The operation of the edge device 200 in the system 100 is explained as follows. The cameras 120 forward real-time data of the traffic environment at the intersection 130. The traffic environment is processed by the traffic image generator 210 to generate traffic images of the traffic environment for each time instance. The traffic analyzer 212 analyzes the traffic image to determine lane occupancy of the lanes 132, 134, 136, and 138. The determination of the lane occupancy is further explained in fig. 3. In one embodiment, the traffic analyzer 212 analyzes the traffic environment captured by the camera 120 to determine lane occupancy. The traffic analyzer 212 includes a comparator 222, a threshold module 224, an optimal profile module 226, and an anomaly identifier 228.
Thereafter, the traffic object determination module 214 determines traffic objects in the traffic environment based on the predetermined identification model. The traffic objects include vehicle type, vehicle occupancy, and pedestrian type. The predetermined identification model is a training model generated by the traffic model generator 112 using the neural network module 215. The process of determining traffic objects by the traffic object determination module 214 is further explained in fig. 4.
After the traffic object determination module 214 identifies the traffic object, the traffic density estimator 216 determines the traffic density in real-time. The number of vehicles and pedestrians at the intersection is calculated. Further, the number of vehicles in each lane 132, 134, 136, and 138 is calculated by the traffic density estimator 216. Further, vehicles are classified based on their type, such as automobiles, bicycles, buses, public transportation, private transportation, emergency service vehicles, and the like. The number of vehicles in each category is also calculated.
When the edge device 200 is determining traffic density, the traffic prediction module 218 predicts a forecasted traffic density based on the traffic environment associated with the intersection 130 and the historical traffic environment. In particular, the traffic prediction module 218 generates a forecasted traffic density based on the forecast model from the forecast model generator 114. The generation of the forecast traffic density is further explained in fig. 5.
The traffic density and forecast density are then compared by the contour determination module 220. Based on the comparison, a traffic signal profile for the intersection 130 is selected. In one embodiment, a traffic discrepancy between the traffic density and the forecast traffic density is determined. If the traffic difference is within the threshold generated by the threshold module 224 of the traffic analyzer 212, the contour determination module 220 selects a traffic signal contour based on the forecasted traffic density. If the traffic difference is above the generated threshold, the contour determination module 220 uses the optimal contour module 226 to calculate an optimal traffic signal contour. The optimal traffic signal profile is then tested by determining if there are any traffic anomalies at the intersection 130 due to the optimal traffic signal profile. Traffic anomalies are identified by the anomaly identifier 228. If a traffic anomaly is identified, the optimal traffic signal profile is re-determined. If no traffic anomaly is identified, the best traffic signal profile is selected as the traffic signal profile.
Fig. 3 illustrates the determination of the lane occupancy at an intersection. As shown in the figure, the intersection includes four lanes 302, 304, 306, and 308. Each lane 302, 304, 306, and 308 is associated with a traffic signal profile 312, 314, 316, and 318. To determine lane occupancy, a traffic environment, including vehicle traffic, is captured by a capture member, such as a camera 120 and other devices 122 and 124.
As seen in fig. 3, each lane 302, 304, 306, and 308 is divided into four sections, and the traffic density in each section is determined. For example, in the lane 302, the traffic density in the first section 302a is 67.9%, the second section 302b is 23.9%, the third section 302c is 5.3%, and the fourth section 302d is 0%. Similarly, traffic densities of 35.6%, 95.4%, 82.1%, and 2% were determined for sections 304a, 304b, 304c, and 304 d. This determination is done for lanes 306 and 308, with sections 306a, 306b, 306c, and 306d having traffic densities of 82.3%, 84.8%, 23.0%, and 0%, respectively. Sections 308a, 308b, 308c, and 308d have traffic densities of 11.1%, 20.8%, 3.2%, and 0%.
Generally, only traffic environments near an intersection are captured, and traffic flow is managed based on traffic density closest to the intersection. If such an approach is employed, the traffic density at segment 306a is highest and, therefore, lane 306 will be given a green traffic signal profile. However, as seen, the traffic density in lane 304 is higher than the traffic density of 306. According to the prior art, lane 304 is given a green traffic signal profile instead of lane 306. The present invention is therefore advantageous in that it accurately determines lane occupancy as a whole, and also determines lane occupancy in each segment of the lane, and compares lane occupancy across individual segments in other lanes. In addition, the use of other capture devices, such as GPS devices, mobile phones, or cameras located on street lights, more accurately captures the traffic environment at each lane and results in better management of traffic flow.
Fig. 4 illustrates the identification of traffic objects by the edge device 200. The edge device 200 uses a neural network to identify traffic objects. In this embodiment, a convolutional neural network is used to identify traffic objects. As shown in the figure, a traffic image 450 of the traffic environment at an intersection is taken. The convolutional neural network extracts relevant information from the pixels of the traffic image 450 and inputs this information into a fully connected neural network with a normalized index (softmax) output layer 414, the normalized index output layer 414 yielding the traffic density for the vehicle type under consideration (e.g., quad, two-wheel, and special type "other" for other vehicle types or no-vehicle zones). The convolutional neural network is trained on a predetermined identification model that includes a data set mapped to an image of a traffic object.
In particular, the traffic image 450 is represented as a three-dimensional array of pixel intensities for 3 array dimensions of a feature map including color, height, and width. The traffic image 450 is transformed by convolving the feature extraction layers according to the following equation:
wherein,lthe layer index is marked up and,kthe index of the feature map is marked,in correspondence with the array of pixels of the image,andis a filter and an offset, which correspond tolLayers and learning from training exampleskA feature map, and f is a function of an element aspect, such as tanh (x) or max (0, x).
As shown in fig. 4, pooling layers 404 and 408 are used subsequent to convolutional layers 402, 406, and 410. Pooling layers 404 and 408 aggregate the spatial local regions using a maximum function, i.e., select a maximum of the spatial local regions. For example, a maximum function may be used to aggregate spatially local regions of size 2x2, i.e., select the maximum of the 2x2 region. Common aggregation functions are maximum or mean functions, but other functions are possible.
After convolutional layers 402, 406, and 410 and pooling layers 404 and 408, the resulting three-dimensional array is either flattened into a vector of length-wise feature maps (including color, height, and width) or is a global pooling layer 412. In this embodiment, the global pooling layer 412 yields a vector of length-wise number of feature maps.
In another embodiment, the final feature map is flattened into a vector, which can be further transformed by one or more subsequent fully-connected layers according to the following equation:
whereinlThe layer index is marked up and,in correspondence with the said vector(s),W l andb l is learned from predetermined identification datalWeight matrices and deviation vectors in the layers, andis a function of the element, such as tanh (x) or max (0, x). In the normalized-index-output layer 414,is chosen as the normalized exponential function:
the normalized exponential function yields a distribution of traffic densities for the vehicle types based on the predetermined identification model.
In one embodiment, the predetermined identification model includes a mapping of traffic objects for all weather and light conditions. In another embodiment, the traffic image 450 is normalized for brightness before being passed into the convolutional neural network. This normalization step is also performed while generating the predetermined identity model.
FIG. 5 illustrates the generation of a predictive model by the system 100. In the present embodiment, the recurrent neural network 500 is used to generate a predictive model for the system 100. A recurrent neural network with time lag is used to model the non-linear relationship between the forecasted traffic density and the historical traffic density and other external inputs.
According to the recurrent neural network, the following inputs are considered:
weather input 502 (x 1 (t)) including weather data, such as temperature, precipitation, and humidity, is taken as weather input 502. The weather input 502 includes weather data for any period in the past.
The pollution inputs 504 (x 2 (t)) include air quality data such as the amount of SO2, CO, NO2 in the atmosphere.
The time input 506 (x 3 (t)) includes data for the time of day. Traffic density has a strong correlation with time of day, for example at peak hours (8-11 AM and 6-8 PM). Thus, the forecasted traffic density includes the likelihood of congestion based on the time of day.
Day of week 508 (x 4 (t)) is considered an input. During weekends, relatively low traffic density was observed. Thus, days of the week 508 are considered in generating the predictive model.
The holiday input 512 (x 5 (t)) includes information related to a predetermined holiday. The holiday input 510 is used to train the predictive model with intelligence, i.e., during the holiday season, high traffic on expressways going outside city is observed.
Consider event input 514 (x 6 (t)) which includes planned and unplanned events that may affect traffic flow. For example, if information regarding planned road repairs near an intersection is available, traffic congestion may be taken into account when generating the forecast model, and thus the forecast traffic density.
The adjacent traffic density input 516 (x 7 (t)). By considering adjacent traffic densities at adjacent intersections, a coordinated green traffic signal profile may be achieved. The system 100 includes several computing devices 150, 152, and 154 at different intersections to manage traffic flow. Since this is a decentralized approach, it is advantageous to consider adjacent traffic density because if there is traffic congestion at adjacent intersections, there is a high probability that the traffic congestion may penetrate other intersections in the geographic location.
Consider pedestrian density 518 (x 8 (t)), which includes the type of pedestrian at the intersection and the pedestrian occupancy. Pedestrian density 518 is generated by employing a neural network on traffic images of the traffic environment at the intersection.
The historical traffic density 522 (x 9 (t)) is also considered as an input to the recurrent neural network to generate the predictive model. The historical traffic density 522 at the intersection is used to predict the forecasted traffic density.
As shown in fig. 5, hidden neurons 1 to n are indicated as 510-510 n. Further, the weights and recursive weights of hidden neurons 1 to n are denoted ar, respectivelyn530 and a 540. For each time step, the predicted output of the hidden layer is fed back as input at the context neuron y (t) 525. The recurrent neural network 500 determines the weighted sum of the hidden neurons and the output of the hidden neurons indicated by 550-550 n. The weighted sum of the hidden neurons n is. The output of the hidden neuron n is WhereinfA non-linear transfer function.
It can be observed from the above equation that the output at t +1 depends on the past predicted values along with the current input values. For the output neuron, the predicted traffic density 580 is. The forecasting model is deployed on a computing device, such as the edge device 200 in a runtime container commonly referred to as an Analytical Model Container Framework (AMCF). In one embodiment, during operation, preThe traffic model receives historical data for each hour used to generate forecasted traffic densities for the next hour.
Fig. 6 is a block diagram of a system 600 for managing traffic flow for a plurality of intersections 602, 604, and 606. Each intersection 602, 604, and 606 includes a plurality of lanes 610a-610n, 620a-620n, and 630a-630 n. In addition, the intersections 602, 604, and 606 include traffic signs 618, 628, and 638 to manage the flow of traffic in the respective lanes 610a-610n, 620a-620n, and 630a-630 n.
In addition, each intersection 602, 604, and 606 includes a respective traffic agent 615, 625, and 635. The traffic agents 615, 625, and 635 include computing devices such as edge devices 200 and capture devices such as cameras 120. The traffic agents 615, 625, and 635 are connected to the signal controllers 616, 626, and 636, respectively. The functionality of the traffic agents 615, 625, and 635 is indicated by function blocks that determine the traffic signal profile for each of the intersections 602, 604, 606. The traffic agent 615 includes a traffic density estimator 612, a traffic prediction module 614, and a contour determination module 613. Similarly, the traffic agent 625 includes a traffic density estimator 622, a traffic prediction module 624, and a contour determination module 623. The traffic agent 635 includes a traffic density estimator 632, a traffic prediction module 634, and a contour determination module 633.
As shown in FIG. 6, the traffic proxy 615 selects a traffic signal profile for each lane 610a-610n by determining traffic density in real-time based on the traffic environment of the intersection 602 using the traffic density estimator 612. The traffic agent 615 uses the traffic prediction module 614 to predict the forecasted traffic density based on the traffic environment associated with the intersection and the historical traffic environment. In addition, the traffic proxy 615 uses the contour determination module 613 to select a traffic signal contour for the intersection based on the traffic density and the forecasted traffic density. The above steps are completed simultaneously for the lanes 610a-610n in the intersection 602. The traffic agents 625 and 635 select the traffic signal profiles for the intersections 604 and 606 as described for the traffic agent 625. The traffic agents 615, 625, and 635 transmit the selected traffic signal profile to the signal controllers 616, 626, and 636. Additionally, the traffic agents 615, 625, and 635 communicate the selected traffic signal profiles to each other such that a coordinated green color can be achieved across the intersections 602, 604, and 606. Coordinated green is an optimization problem that achieves as many coordinated green signals as possible across the intersections 602, 604, and 606. The method of transmitting the selected traffic signal profile to achieve the coordinated green color is explained in fig. 7.
Fig. 7 illustrates a method 700 of using the system 600 to manage traffic flow for a plurality of intersections 702, 704, and 706. As shown in the figure, the multiple intersections 702, 704, and 706 are adjacent to one another in a geographic location 710. Each intersection 702, 704, and 706 includes a traffic proxy 715, 725, and 735.
The traffic agent 715 observes the traffic environment at the intersection 702 and determines the traffic density at time "t" using a convolutional neural network, as indicated by arrow 711. In addition, the traffic proxy 715 also receives the forecast model and determines a forecast traffic density for time "t", as indicated by arrow 718. A traffic strategy is selected 713 based on the traffic density and the forecasted traffic density. Based on the traffic policy 713, a traffic signal profile is selected at 717. The traffic signal profile is reflected at the traffic sign post of the intersection 702, as indicated by the arrow 720. In addition, the traffic signal profile is transmitted to traffic proxies 725 at adjacent intersections 704, as indicated by arrows 719.
Similarly, the traffic proxy 725 observes the traffic environment at the intersection 704 and determines the traffic density at time "t + 1" using a convolutional neural network, as indicated by arrow 721. In addition, the traffic proxy 725 also receives the predictive model and determines a predicted traffic density for time "t", as indicated by arrow 728. A traffic strategy 723 is selected based on the traffic density and the forecasted traffic density. Based on the traffic policy 723, a traffic signal profile is selected at 727. The traffic signal profile is reflected at the traffic sign post of the intersection 704, as indicated by the arrow 730. In addition, the traffic signal profile is communicated to traffic agents 735 adjacent to the intersection 706, as indicated by arrow 729.
The system 600 employs a multi-agent deep reinforcement learning algorithm for learning the best behavior of the multiple intersections 702, 704, and 706 such that a coordinated green color is achieved. The algorithm trains a global policy that includes traffic policies (i.e., sub-policies) 713 and 723 per traffic agent 715 and 725. The traffic policy for the traffic proxy 735 is not shown in fig. 7, however, it is generated similar to the traffic policies of 713 and 723. Traffic policies 713 and 723 are the traffic signal profiles of intersections 702 and 704 determined by traffic agents 715 and 725.
To generate a global policy, system 600 represents observed or captured asThe traffic environment of (1). Further, the forecast model and the forecast traffic density are expressed as. The traffic policies 713 and 723 are denoted as p, and the traffic signal profiles selected by each of the traffic agents 715, 725, and 735 are denoted as. In the above symbols, t indicates time and a indicates lanes in the intersections 702, 704, and 706.
In addition, each traffic proxy 715, 725, 735 observes the current state (i.e., traffic environment) at time "tAnd selecting a traffic signal profile according to a traffic strategy p. The traffic agents 715, 725 and 735 also observe the reward signalsAnd to new traffic environmentsIs performed. Reward signalBy identifying whether a traffic anomaly has been due to a traffic signal profileBut rather occurs in the geographic location 710. The system 600 attempts to achieve a coordinated green color by maximizing the number of consecutive green traffic signal profiles in discount return. System 600 uses equationsTo perform optimization, whereinIs a reward received at time t, andis a discount factor. The discount factor beta depends on the traffic environment at each intersection. For example, the discount factor β may be greater if the intersection is on a highway and if it is a weekend (where there are more vehicles moving between cities).
The system 600 also uses the Q function to determine the coordinated green color. The Q function of the traffic strategy p is Q (S, U), i.e. a set comprising traffic environment and traffic signal profiles. The Q function includes the Q value of the traffic environment, which is denoted as Qu. In addition, the Q function also includes the Q values of the traffic signal profiles (indicated by arrows 719 and 729) of adjacent intersections, which are represented by QmAnd (4) showing. The system 600 then uses an Elepin greedy strategy to derive the Q value (Q) from the traffic environmentu) And Q value (Q) of traffic signal profile of adjacent intersectionm) Is selectively selectedAnd. The Epsilon greedy strategy is a strategy that works from a set of available actions (i.e., traffic signal contours)In a manner that selects random motion (i.e., having a uniformly distributed traffic signal profile).
Fig. 8 illustrates a method 800 of managing traffic flow in a geographic location. At step 802, the method begins with capturing traffic environment in real-time for an intersection in a geographic location. Traffic environments are captured using cameras placed on street lights, traffic signs at intersections, street fences. The traffic environment may also be determined from cameras provided on unmanned aerial devices associated with the intersection.
At step 804, lane occupancy of two or more lanes associated with the intersection is determined based on the traffic environment. Further, at step 805, a traffic object in the traffic environment is determined based on the predetermined identification model. The traffic objects include vehicle type, vehicle occupancy, and pedestrian type. Identifying a traffic object by: generating a traffic image of the traffic environment, and identifying traffic objects independent of vehicle position, relative vehicle proximity, and pedestrian position based on the predetermined identification model. In an example embodiment, traffic objects are identified by passing traffic images through multiple layers of a neural network. The plurality of layers includes a convolutional layer and a pooling layer.
At step 806, information about the traffic environment is captured from using a device such as a GPS-enabled device, a mobile computing device, or the like. Information is also captured from media including social media. At step 808, this information is used to determine whether there are any traffic anomalies that may be associated with the intersection. Traffic anomalies include accidents, vehicle failures, traffic violations. At step 810, traffic density is determined in real-time based on the traffic environment at the intersection.
The method also includes, at step 820, predicting a forecasted traffic density based on the traffic environment associated with the intersection and the historical traffic environment. As indicated in step 815, the forecasted traffic density is calculated based on several parameters associated with the traffic environment and the historical traffic environment. As indicated at step 825, a recurrent neural network is used to calculate the forecasted traffic density.
Further, the method includes, at step 860, selecting a traffic signal contour for the intersection based on the traffic density and the forecasted traffic density. To select a traffic signal profile, the traffic density is compared to a forecast traffic density to generate a traffic difference. Further, at step 830, it is determined whether the traffic difference has exceeded a threshold. The threshold is determined based on the historical traffic environment of the intersection. If the traffic difference is less than the threshold, a traffic signal profile is selected based on the forecasted traffic density. At step 850, a traffic signal profile is output via the traffic sign post of the intersection.
If the traffic difference is greater than the threshold, then at step 840 the method determines an optimal traffic signal profile for the intersection, the optimal traffic signal profile being based on the traffic difference. If the traffic difference is significantly greater than the threshold, the optimal traffic signal profile will change. The optimal traffic signal profile is tested in real time by identifying whether any traffic anomalies exist at the intersection. If a traffic anomaly exists, the optimal traffic signal profile is recalculated until no traffic anomaly is identified. If there is no traffic anomaly, then at step 850, the optimal traffic signal is output to the traffic sign post.
In addition to the specific components and/or circuits set forth above, the methods, apparatus and systems disclosed above may be implemented via implementations having different or entirely different components. With regard to such other components (e.g., circuits, computing/processing components, etc.) and/or computer-readable media associated with or embodying the present invention, for example, aspects of the present invention herein may be implemented in accordance with numerous general purpose or special purpose computing systems or configurations. Various exemplary computing systems, environments, and/or configurations that may be suitable for use with the disclosed subject matter may include, but are not limited to, various clock-related circuits (such as those within a personal computer), server or server computing devices (such as routing/connection components), hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, smart phones, consumer electronics, network PCs, other existing computer platforms, distributed computing environments that include one or more of the above systems or devices, and the like.
In some examples, aspects of the invention herein may be implemented via logic and/or logic instructions comprising, for example, program modules executed in association with circuitry. Generally, program modules may include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular control, delay or instructions. The invention may also be practiced in the context of a distributed circuit arrangement in which circuits are connected via communications buses, circuits, or links. In a distributed setting, control/instructions may occur from both local and remote computer storage media including memory storage devices.
The systems and computing devices herein, along with their components, may also include and/or utilize one or more types of computer-readable media. Computer-readable media can be any available media that can reside on, be associated with, or be accessed by such circuitry and/or computing components. By way of example, and not limitation, computer readable media may comprise computer storage media and communication media. Computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by the computing components. Communication media may include computer readable instructions, data structures, program modules, or other data that embody the functionality herein. In addition, communication media may include wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared and other wireless media. Combinations of any of the above are also included within the scope of computer readable media.
In this specification, the terms component, module, device, and the like may refer to any type of logical or functional circuit, block, and/or process that may be implemented in various ways. For example, the functions of the various circuits and/or blocks may be combined with one another into any other number of modules. Each module may even be implemented as a software program stored on a tangible memory (e.g., random access memory, read only memory, CD-ROM memory, hard drive) to be read by a central processing unit to implement the functionality of the invention herein. Alternatively, the modules may include programming instructions that are transmitted to a general purpose computer or to processing/graphics hardware via a transmission carrier wave. Further, the modules may be implemented as hardware logic circuitry that implements the functionality encompassed by the invention herein. Finally, the modules may be implemented using dedicated instructions (SIMD instructions), field programmable logic arrays, or any mixture thereof that provides the desired level of performance and cost.
As disclosed herein, implementations and features consistent with the present invention may be implemented by computer hardware, software, and/or firmware. For example, the systems and methods disclosed herein may be embodied in various forms including, for example, a data processor, such as a computer that also includes a database, digital electronic circuitry, firmware, software, or a combination thereof. Further, although some of the disclosed implementations describe components such as software, systems and methods consistent with the invention herein may be implemented with any combination of hardware, software, and/or firmware. Furthermore, the above-identified features and other aspects and principles of the present invention herein may be implemented in various environments. Such environments and related applications may be specially constructed for performing the various processes and operations in accordance with the invention, or they may include a general-purpose computer or computing platform selectively activated or reconfigured by code to provide the necessary functionality. The processes disclosed herein are not inherently related to any particular computer, network, architecture, environment, or other apparatus and may be implemented by a suitable combination of hardware, software, and/or firmware. For example, various general-purpose machines may be used with programs written in accordance with the teachings of the invention herein, or it may be more convenient to construct a specialized apparatus or system to perform the required methods and techniques.
Aspects of the methods and systems described herein, such as logic, may be implemented as functionality programmed into any of a variety of circuits, including programmable logic devices ("PLDs") (such as field programmable gate arrays ("FPGAs")), programmable array logic ("PAL") devices, electrically programmable logic and memory devices, and standard cell-based devices and application specific integrated circuits. Some other possibilities for implementing aspects include: memory devices, microcontrollers with memory (such as EEPROMs), embedded microprocessors, firmware, software, and the like. Further, aspects may be embodied in a microprocessor having: software-based circuit simulation, discrete logic (sequential and combinatorial), custom devices, fuzzy (neural) logic, quantum devices, and hybrids of any of the above device types. The underlying device technology may be provided in a variety of component types, such as metal oxide semiconductor field effect transistor ("MOSFET") technology (e.g., complementary metal oxide semiconductor ("CMOS")), bipolar technology (e.g., emitter coupled logic ("ECL")), polymer technology (e.g., silicon-conjugated polymers and metal-conjugated polymer-metal structures), hybrid analog and digital, and so forth.
It should also be noted that the various logic and/or functions disclosed herein may be implemented using any number of combinations of hardware, firmware, and/or as data and/or instructions embodied in various machine-readable or computer-readable media, depending on their behavior, register transfer, logic component, and/or other characteristics. Computer-readable media in which such formatted data and/or instructions may be embodied include, but are not limited to, non-volatile storage media in various forms (e.g., optical, magnetic or semiconductor storage media) and carrier waves that may be used to transmit such formatted data and/or instructions through wireless, optical, or wired signaling media or any combination thereof. Examples of transmission of such formatted data and/or instructions by a carrier wave include, but are not limited to: transmission (upload, download, email, etc.) over the internet and/or other computer networks via one or more data transfer protocols (e.g., hypertext transfer protocol (HTTP), File Transfer Protocol (FTP), Simple Mail Transfer Protocol (SMTP), etc.).
Unless the context clearly requires otherwise, throughout the description and the claims, the words "comprise", "comprising", and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is, in the meaning of "including but not limited to". Words using the singular or plural number also include the plural or singular number, respectively. Additionally, the words "herein," "below," "upper," "lower," and words of similar import refer to this application as a whole and not to any particular portions of this application.
While certain presently preferred implementations of the present invention have been described in detail herein, it will be apparent to those skilled in the art to which the present invention relates that variations and modifications of the various implementations shown and described herein may be made without departing from the scope of the invention herein. Accordingly, it is intended that the invention be limited only to the extent required by the appended claims and the applicable rules of law.

Claims (16)

1. A method of managing traffic flow in a geographic location, the method comprising:
determining traffic density in real-time based on traffic environment of at least one intersection (130) in the geographic location, wherein the traffic environment includes vehicular traffic and pedestrian traffic;
predicting a forecasted traffic density based on the traffic environment associated with the at least one intersection (130) and a historical traffic environment;
selecting a traffic signal profile for the at least one intersection (130) based on the traffic density and the forecasted traffic density; and
managing traffic flow in the geographic location based on traffic signal profiles of the at least one intersection (130).
2. The method of claim 1, further comprising:
capturing the traffic environment in real-time for the at least one intersection (130);
determining lane occupancy of two or more lanes associated with the at least one intersection (130) based on the traffic environment;
determining traffic objects in the traffic environment based on a predetermined identification model, wherein the traffic objects include vehicle types, vehicle occupancy, and pedestrian types; and
a traffic anomaly associable with the at least one intersection (130) is determined, wherein the traffic anomaly includes an accident, a vehicle fault, and a traffic violation.
3. The method of claim 2, wherein determining traffic objects in the traffic environment based on a predetermined identification model, further comprises:
generating a traffic image of the traffic environment; and
the traffic object is identified based on the predetermined identification model independently of vehicle location, relative vehicle proximity, and pedestrian location.
4. The method of claim 3, wherein identifying traffic objects independent of vehicle location, relative vehicle proximity, and pedestrian location comprises:
communicating the traffic image through a plurality of layers of a neural network, wherein the plurality of layers includes a convolutional layer and a pooling layer.
5. The method of claim 1, wherein selecting a traffic signal profile for the at least one intersection (130) based on the traffic density and the forecasted traffic density comprises:
comparing the traffic density to the forecasted traffic density to generate a traffic differential;
determining whether the traffic difference has exceeded a threshold, wherein the threshold is determined based on the historical traffic environment of the at least one intersection (130); and
selecting a traffic signal profile for the at least one intersection (130) based on the forecasted traffic density if the traffic difference is within the threshold.
6. The method of claim 5, further comprising:
determining an optimal traffic signal profile for the at least one intersection (130) if the traffic difference exceeds the threshold, wherein the optimal traffic signal profile is based on the traffic difference;
identifying a traffic anomaly at the at least one intersection (130) that can be associated to the optimal traffic signal profile;
repeating the step of determining the optimal traffic signal profile if the traffic anomaly is identified; and
selecting the optimal traffic signal profile as a traffic signal profile if the traffic anomaly is not identified.
7. The method of claim 1, further comprising:
communicating the traffic signal profile of the at least one intersection (130) to adjacent intersections;
generating a traffic signal profile for the neighboring intersection based on the transmitted traffic signal profile for the at least one intersection (130);
iteratively generating traffic signal contours of a plurality of intersections based on the traffic signal contours of the adjacent intersections; and
and managing the traffic flows of the multiple intersections based on the traffic signal profiles of the adjacent intersections.
8. A computing device (200) for managing traffic flow in a geographic location, the computing device (200) comprising:
at least one processor (202); and
a memory (204) coupled to the at least one processor (202), the memory (204) comprising:
a traffic density estimator (216) capable of determining traffic density in real-time based on a traffic environment of at least one intersection (130) in the geographic location, the traffic environment including vehicular traffic and pedestrian traffic;
a traffic prediction module (218) capable of predicting a forecasted traffic density based on the traffic environment associated with the at least one intersection (130) and a historical traffic environment;
a contour determination module (220) capable of selecting a traffic signal contour for the at least one intersection (130) based on the traffic density and the forecasted traffic density; and
a communication unit (208) capable of communicating a traffic signal profile of the at least one intersection (130) to a traffic sign post (140) associated with the at least one intersection (130) for managing traffic flow in the geographic location.
9. The computing device (200) of claim 8, further comprising:
a traffic analyzer (212) capable of determining lane occupancy of two or more lanes associated with the at least one intersection (130) based on the traffic environment;
a traffic image generator (210) capable of generating a traffic image of the traffic environment; and
a traffic object determination module (214) capable of determining traffic objects in the traffic environment based on a predetermined identification model, wherein the traffic objects include vehicle types, vehicle occupancy, and pedestrian types.
10. The computing device (200) of claim 9, wherein the traffic analyzer (212) comprises:
a comparator (222) capable of comparing the traffic density with the forecasted traffic density to generate a traffic difference; and
a threshold module (224) capable of determining whether the traffic variance has exceeded a threshold, wherein the threshold is determined based on the traffic environment of the at least one intersection (130).
11. The computing device (200) of claim 9, wherein the traffic analyzer (212) comprises:
an optimal profile module (226) configured to: determining an optimal traffic signal profile for the at least one intersection (130) if the traffic difference exceeds the threshold, wherein the optimal traffic signal profile is based on the traffic difference; and
an anomaly identifier (228) configured to identify a traffic anomaly at the at least one intersection (130) associable to the optimal traffic signal profile determined by the anomaly identifier, wherein the optimal traffic signal profile is re-determined if the traffic anomaly is identified, and wherein the optimal traffic signal profile is selected as a traffic signal profile if the traffic anomaly is not identified.
12. The computing device (200) of claim 8, wherein the contour determination module (220) is capable of selecting a traffic signal contour for the at least one intersection (130) based on the forecasted traffic density if the traffic difference is within the threshold.
13. The computing device (200) of claim 8, wherein the communication unit (208) comprises:
a transmitter (248) communicatively coupled to the processor (202) to transmit a traffic signal profile of the at least one intersection (130) to an adjacent intersection; and
a receiver (258) communicatively coupled to the processor (202) to receive a traffic signal profile of the neighboring intersection, wherein the traffic signal profile of the at least one intersection (130) is based on the traffic signal profile of the neighboring intersection.
14. A system (100, 600) for managing traffic flow in a geographic location, the system (100, 600) comprising:
a server (102);
a network (108) communicatively coupled to the server (102); and
one or more computing devices (200) as recited in claims 8-15, communicatively coupled to the server (102) via the network (108), and wherein the one or more computing devices (200) are associated with one or more intersections (130) in the geographic location.
15. The system (100, 600) of claim 14, further comprising:
a plurality of capture devices (120, 122, 124) communicatively coupled to the server (102) via the network (108), the plurality of capture devices (120, 122, 124) being selected ones of manned and unmanned devices including: image sensors, motion sensors, GPS devices, and communication devices.
16. The system (100) of claim 14, wherein the server (102) comprises:
a database (110) comprising data associated with the traffic environment, wherein the traffic environment comprises data associated with: weather, air quality, time, day, scheduled events, and unscheduled events including geographic locations of the plurality of intersections;
a traffic model generator (112) capable of generating a predetermined identification model based on the traffic environment; and
a forecast model generator (114) capable of generating a forecast model based on historical traffic environment.
CN201780088995.3A 2017-02-03 2017-02-03 For managing the system, apparatus and method of the traffic in geographical location Pending CN110494902A (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/EP2017/052398 WO2018141403A1 (en) 2017-02-03 2017-02-03 System, device and method for managing traffic in a geographical location

Publications (1)

Publication Number Publication Date
CN110494902A true CN110494902A (en) 2019-11-22

Family

ID=57963228

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201780088995.3A Pending CN110494902A (en) 2017-02-03 2017-02-03 For managing the system, apparatus and method of the traffic in geographical location

Country Status (2)

Country Link
CN (1) CN110494902A (en)
WO (1) WO2018141403A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110969871A (en) * 2019-12-18 2020-04-07 浙江大学 Intelligent traffic light control system and control method
CN115280391A (en) * 2020-03-27 2022-11-01 大众汽车股份公司 Computer program, device, method for a vehicle for determining a traffic density from at least one motion curve of the vehicle, and vehicle

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3627472A1 (en) * 2018-09-19 2020-03-25 Deutsche Telekom AG Method and assembly for automated local control of multi-modal urban traffic flows
WO2020077529A1 (en) 2018-10-16 2020-04-23 Beijing DIDI Infinity Technology and Development Co., Ltd Adaptive traffic control using vehicle trajectory data
WO2020077527A1 (en) * 2018-10-16 2020-04-23 Beijing Didi Infinity Technology And Development Co., Ltd. System to optimize scats adaptive signal system using trajectory data
CN109887282B (en) * 2019-03-05 2022-01-21 中南大学 Road network traffic flow prediction method based on hierarchical timing diagram convolutional network
CN110675621B (en) * 2019-05-23 2021-01-15 北京嘀嘀无限科技发展有限公司 System and method for predicting traffic information
CN110188936B (en) * 2019-05-23 2021-11-05 浙江大学 Short-term traffic flow prediction method based on multi-factor space selection deep learning algorithm
CN110782663B (en) * 2019-09-30 2021-09-24 电子科技大学 Road network traffic flow short-time prediction method combining time-space characteristics
US20210118288A1 (en) * 2019-10-22 2021-04-22 Mitsubishi Electric Research Laboratories, Inc. Attention-Based Control of Vehicular Traffic
DE102020116669A1 (en) 2019-11-06 2021-05-06 Fujitsu Technology Solutions Intellectual Property Gmbh Method for controlling a traffic system, device, computer program and computer-readable storage medium
US11521487B2 (en) * 2019-12-09 2022-12-06 Here Global B.V. System and method to generate traffic congestion estimation data for calculation of traffic condition in a region
WO2022033677A1 (en) * 2020-08-12 2022-02-17 Siemens Aktiengesellschaft System and method for adaptive traffic signal planning and control
CN116508081A (en) * 2020-11-09 2023-07-28 华为云计算技术有限公司 Apparatus and method for vehicle traffic signal optimization
CN114189876B (en) * 2021-11-17 2023-11-21 北京航空航天大学 Flow prediction method and device and electronic equipment
CN114973653B (en) * 2022-04-27 2023-12-19 中国计量大学 Traffic flow prediction method based on space-time diagram convolutional network
CN117292551B (en) * 2023-11-27 2024-02-23 辽宁邮电规划设计院有限公司 Urban traffic situation adjustment system and method based on Internet of things

Citations (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0454166A2 (en) * 1990-04-27 1991-10-30 Hitachi, Ltd. Traffic flow measuring method and apparatus
US6339383B1 (en) * 1999-11-05 2002-01-15 Sumitomo Electric Industries, Ltd. Traffic signal control apparatus optimizing signal control parameter by rolling horizon scheme
CN101017080A (en) * 2002-09-30 2007-08-15 石川岛播磨重工业株式会社 Device of measuring object
CN101038700A (en) * 2007-04-20 2007-09-19 东南大学 Mixed controlling method of single dot signal controlling crossing
CN101281685A (en) * 2008-01-30 2008-10-08 吉林大学 Coordination control method for area mixed traffic self-adaption signal
CN101751782A (en) * 2009-12-30 2010-06-23 北京大学深圳研究生院 Crossroad traffic event automatic detection system based on multi-source information fusion
CN102156822A (en) * 2011-04-28 2011-08-17 北京市劳动保护科学研究所 Pedestrian traffic data assembly multi-step forecasting method
DE102011018821A1 (en) * 2010-06-28 2011-12-29 Deutsches Zentrum für Luft- und Raumfahrt e.V. Traffic-dependent controlling method for traffic light for controlling traffic streams at intersection, involves distributing phase-specific release times or phase-specific extension times in dependence of total release time
CN103473926A (en) * 2013-09-11 2013-12-25 无锡加视诚智能科技有限公司 Gun-ball linkage road traffic parameter collection and rule breaking snapshooting system
CN104200680A (en) * 2014-08-18 2014-12-10 浙江大学 Traffic signal coordination control method of arterial street under super saturation traffic state
CN104463194A (en) * 2014-11-04 2015-03-25 深圳市华尊科技有限公司 Driver-vehicle classification method and device
US20160027299A1 (en) * 2014-07-28 2016-01-28 Econolite Group, Inc. Self-configuring traffic signal controller
WO2016022108A1 (en) * 2014-08-06 2016-02-11 Robinson Kurt B Systems and methods involving features of adaptive and/or autonomous traffic control
CN105574543A (en) * 2015-12-16 2016-05-11 武汉烽火众智数字技术有限责任公司 Vehicle brand and model identifying method and system based on deep learning
CN105787437A (en) * 2016-02-03 2016-07-20 东南大学 Vehicle brand type identification method based on cascading integrated classifier
CN105957356A (en) * 2016-05-30 2016-09-21 曹莲英 Traffic control system and method based on number of pedestrians
CN106056934A (en) * 2016-08-04 2016-10-26 杭州普乐科技有限公司 Intelligent active traffic signal controller and control method
CN106355885A (en) * 2016-11-24 2017-01-25 深圳市永达电子信息股份有限公司 Traffic signal dynamic control method and system based on big data analysis platform

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130176146A1 (en) 2010-06-15 2013-07-11 The Provost, Fellows And Scholars Of The College Of The Holy & Undivided Trinity Of Queen Elizabeth Decentralised Autonomic System and Method for Use in an Urban Traffic Control Environment

Patent Citations (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0454166A2 (en) * 1990-04-27 1991-10-30 Hitachi, Ltd. Traffic flow measuring method and apparatus
US6339383B1 (en) * 1999-11-05 2002-01-15 Sumitomo Electric Industries, Ltd. Traffic signal control apparatus optimizing signal control parameter by rolling horizon scheme
CN101017080A (en) * 2002-09-30 2007-08-15 石川岛播磨重工业株式会社 Device of measuring object
CN101038700A (en) * 2007-04-20 2007-09-19 东南大学 Mixed controlling method of single dot signal controlling crossing
CN101281685A (en) * 2008-01-30 2008-10-08 吉林大学 Coordination control method for area mixed traffic self-adaption signal
CN101751782A (en) * 2009-12-30 2010-06-23 北京大学深圳研究生院 Crossroad traffic event automatic detection system based on multi-source information fusion
DE102011018821A1 (en) * 2010-06-28 2011-12-29 Deutsches Zentrum für Luft- und Raumfahrt e.V. Traffic-dependent controlling method for traffic light for controlling traffic streams at intersection, involves distributing phase-specific release times or phase-specific extension times in dependence of total release time
CN102156822A (en) * 2011-04-28 2011-08-17 北京市劳动保护科学研究所 Pedestrian traffic data assembly multi-step forecasting method
CN103473926A (en) * 2013-09-11 2013-12-25 无锡加视诚智能科技有限公司 Gun-ball linkage road traffic parameter collection and rule breaking snapshooting system
US20160027299A1 (en) * 2014-07-28 2016-01-28 Econolite Group, Inc. Self-configuring traffic signal controller
WO2016022108A1 (en) * 2014-08-06 2016-02-11 Robinson Kurt B Systems and methods involving features of adaptive and/or autonomous traffic control
CN104200680A (en) * 2014-08-18 2014-12-10 浙江大学 Traffic signal coordination control method of arterial street under super saturation traffic state
CN104463194A (en) * 2014-11-04 2015-03-25 深圳市华尊科技有限公司 Driver-vehicle classification method and device
CN105574543A (en) * 2015-12-16 2016-05-11 武汉烽火众智数字技术有限责任公司 Vehicle brand and model identifying method and system based on deep learning
CN105787437A (en) * 2016-02-03 2016-07-20 东南大学 Vehicle brand type identification method based on cascading integrated classifier
CN105957356A (en) * 2016-05-30 2016-09-21 曹莲英 Traffic control system and method based on number of pedestrians
CN106056934A (en) * 2016-08-04 2016-10-26 杭州普乐科技有限公司 Intelligent active traffic signal controller and control method
CN106355885A (en) * 2016-11-24 2017-01-25 深圳市永达电子信息股份有限公司 Traffic signal dynamic control method and system based on big data analysis platform

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110969871A (en) * 2019-12-18 2020-04-07 浙江大学 Intelligent traffic light control system and control method
CN115280391A (en) * 2020-03-27 2022-11-01 大众汽车股份公司 Computer program, device, method for a vehicle for determining a traffic density from at least one motion curve of the vehicle, and vehicle

Also Published As

Publication number Publication date
WO2018141403A1 (en) 2018-08-09

Similar Documents

Publication Publication Date Title
CN110494902A (en) For managing the system, apparatus and method of the traffic in geographical location
US11755018B2 (en) End-to-end interpretable motion planner for autonomous vehicles
JP7105305B2 (en) Multistage image-based object detection and recognition
US20200302662A1 (en) System and Methods for Generating High Definition Maps Using Machine-Learned Models to Analyze Topology Data Gathered From Sensors
US11157010B1 (en) Method and system for deterministic trajectory selection based on uncertainty estimation for an autonomous agent
US11565716B2 (en) Method and system for dynamically curating autonomous vehicle policies
US11537127B2 (en) Systems and methods for vehicle motion planning based on uncertainty
US11960290B2 (en) Systems and methods for end-to-end trajectory prediction using radar, LIDAR, and maps
CN111699519A (en) System, device and method for detecting abnormal traffic events in a geographic location
US12013457B2 (en) Systems and methods for integrating radar data for improved object detection in autonomous vehicles
CN114418468A (en) Smart city traffic scheduling strategy control method and Internet of things system
CA3170637A1 (en) System and method for autonomous vehicle systems simulation
US11620838B2 (en) Systems and methods for answering region specific questions
US11891087B2 (en) Systems and methods for generating behavioral predictions in reaction to autonomous vehicle movement
EP4303793A2 (en) Path segment risk regression system for on-demand transportation services
DE102021112349A1 (en) VEHICLE OPERATION USING A DYNAMIC ALLOCATION GRID
US20210278852A1 (en) Systems and Methods for Using Attention Masks to Improve Motion Planning
US11256926B2 (en) Method and system for analyzing the movement of bodies in a traffic system
US12023812B2 (en) Systems and methods for sensor data packet processing and spatial memory updating for robotic platforms
CN110843789A (en) Vehicle lane change intention prediction method based on time sequence convolution network
CN113257016B (en) Traffic signal control method and device and readable storage medium
US11887324B2 (en) Cross-modality active learning for object detection
Patil Applications of deep learning in traffic management: A review
CN104915628A (en) Pedestrian movement prediction method and device by carrying out scene modeling based on vehicle-mounted camera
CN117387647A (en) Road planning method integrating vehicle-mounted sensor data and road sensor data

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
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20220215

Address after: Munich, Germany

Applicant after: Yunnix Co.,Ltd.

Address before: Munich, Germany

Applicant before: Siemens Transportation Co.,Ltd.

RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20191122