Method and device for controlling signals at intersections

Abstract

A method for controlling a signal at an intersection according to one embodiment of the present disclosure includes the steps of: acquiring input data through a traffic condition collection device; predicting traffic condition information at a predetermined time point from the input data through a trained traffic condition prediction model - the predicted traffic condition information being related to at least one of a maximum number of waiting vehicles or a passing traffic volume -; calculating an optimal aspect of the target intersection based on the predicted traffic condition information; and controlling a signal at the target intersection based on the calculated optimal aspect.

Description

[Technical field]

[0001] The present disclosure relates to a technology for controlling traffic signals at an intersection. Specifically, the present disclosure relates to a technology for predicting traffic condition information using artificial intelligence and calculating an optimal signal phase based on the prediction result. Furthermore, the present disclosure relates to a technology for controlling a traffic signal at an intersection based on the calculated optimal signal phase. [Background technology]

[0002] As artificial intelligence technology develops, it is being used in a variety of industrial fields. In particular, research on predicting future traffic information based on collected traffic information using deep learning models and controlling traffic signals has been attracting attention recently.

[0003] Previously, traffic signals were controlled by manually allocating traffic signals to each intersection based on traffic volume information collected by on-site investigators, who would then directly determine which time periods had the most vehicles. However, the conventional signal control method had limitations in that it could not reflect real-time traffic information and could not alleviate traffic congestion because it controlled signal distribution in the same way until the next investigation. Furthermore, the conventional signal control method had limitations in that it was inefficient and could not guarantee accuracy because signals were allocated manually by investigators.

[0004] Therefore, there is a demand for the development of new methods for predicting traffic condition information, methods for controlling signals at intersections, and devices for carrying out these methods. Summary of the Invention [Problem to be solved by the invention]

[0005] One problem to be solved by the present disclosure is to provide an intersection signal control method for predicting future traffic conditions at an intersection in real time and automatically and optimally controlling intersection signals based on the prediction results, and an electronic device for carrying out the same.

[0006] The problems to be solved by the present disclosure are not limited to the problems described above, and problems not mentioned will be clearly understood by a person having ordinary skill in the art to which the present disclosure pertains from this specification and the attached drawings. [Means for solving the problem]

[0007] A method for controlling a signal at an intersection according to one embodiment of the present disclosure may include a step of acquiring input data through a traffic condition collection device; a step of predicting traffic condition information at a predetermined time point from the input data through a trained traffic condition prediction model - the predicted traffic condition information being related to at least one of a maximum number of waiting vehicles or a passing traffic volume - a step of calculating an optimal aspect of the target intersection based on the predicted traffic condition information; and a step of controlling a signal at the target intersection based on the calculated optimal aspect.

[0008] An electronic device according to an embodiment of the present disclosure may include a processor configured to acquire input data through a traffic condition collection device, predict traffic condition information at a predetermined time point from the input data through a trained traffic condition prediction model, calculate an optimal phase for the target intersection based on the predicted traffic condition information, and control a signal at the intersection based on the calculated optimal phase.

[0009] The solutions to the problems of the present disclosure are not limited to the solutions described above, and solutions not mentioned will be clearly understood by a person having ordinary skill in the art to which the present disclosure pertains from this specification and the attached drawings. Effect of the Invention

[0010] According to an embodiment of the present disclosure, a method for controlling a traffic signal at an intersection and an electronic device for performing the same can provide an effect of predicting traffic condition information at a predetermined time point after an analysis time point through a trained traffic condition prediction model.

[0011] According to an embodiment of the present disclosure, a method for controlling traffic signals at an intersection and an electronic device for performing the same can provide an effect of controlling traffic signals in a direction that reduces congestion sections by distributing signals in consideration of the main flow direction of traffic at the intersection.

[0012] According to an embodiment of the present disclosure, a method for controlling intersection signals and an electronic device for performing the same can provide an effect of increasing user convenience by allowing a user to visually check, through a user interface, information on predicted results using a traffic condition prediction model and simulated traffic conditions based on a calculated optimal aspect.

[0013] The effects of the present disclosure are not limited to the effects described above, and effects not mentioned will be clearly understood by a person having ordinary skill in the art to which the present disclosure pertains from this specification and the attached drawings. [Brief description of the drawings]

[0014] [Figure 1] 1 is a schematic diagram of an intersection signal control system according to one embodiment of the present disclosure. [Diagram 2] 1 is a diagram illustrating an operation of an electronic device according to an embodiment of the present disclosure. [Diagram 3] 1 is a flowchart showing a method for controlling a signal at an intersection according to an embodiment of the present disclosure. [Figure 4] 1 is a diagram illustrating an aspect of predicting traffic condition information using a trained traffic condition prediction model according to an embodiment of the present disclosure. [Diagram 5] 1 is a diagram illustrating an example of input variables and output variables of a traffic condition prediction model that has completed training in accordance with an embodiment of the present disclosure. [Figure 6] 1 is a diagram illustrating one aspect of training a traffic condition prediction model according to an embodiment of the present disclosure. [Figure 7] 4 is a flowchart embodying steps for computing an optimal phase for a target intersection according to one embodiment of the present disclosure. [Figure 8]1 is a diagram for explaining one aspect of calculating an optimal current signal for a target intersection according to an embodiment of the present disclosure; [Figure 9] 1 is a diagram illustrating one aspect of a user interface. [Figure 10] 1 is a diagram illustrating one aspect of a user interface. [Figure 11] 1 is a diagram illustrating one aspect of a user interface. [Figure 12] 1 is a diagram illustrating one aspect of a user interface. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] The above-mentioned objects, features and advantages of the present disclosure will become more apparent through the following detailed description taken in conjunction with the accompanying drawings. However, since the present disclosure can be modified in various ways and can have various embodiments, the following will describe in detail a specific embodiment by way of example in the drawings.

[0016] In principle, the same reference numerals denote the same elements throughout the specification, and elements having the same functions within the same concept shown in the drawings of each embodiment will be described using the same reference numerals, and redundant description thereof will be omitted.

[0017] If it is determined that a detailed description of a known function or configuration related to the present disclosure may unnecessarily obscure the gist of the present disclosure, the detailed description will be omitted. In addition, numbers (e.g., 1, 2, etc.) used in the description of this specification are merely identification symbols for distinguishing one component from another.

[0018] In addition, the suffixes "module" and "part" for components used in the following examples are given or mixed for the sole consideration of ease of writing the specification, and do not have any distinct meanings or roles in themselves.

[0019] In the following examples, the singular expression includes the plural expression unless the context clearly indicates otherwise.

[0020] In the following examples, terms such as "comprise" or "have" mean the presence of the features or components described in the specification, but do not preclude the possibility that one or more other features or components may be added.

[0021] In the drawings, the size of components may be exaggerated or reduced for the convenience of explanation. For example, the size and thickness of each component shown in the drawings are arbitrarily shown for the convenience of explanation, and the present disclosure is not necessarily limited to those shown in the drawings.

[0022] When an embodiment is alternatively implemented, the order of certain processes may be performed differently than described. For example, two processes described in succession may be performed substantially simultaneously or may be performed in the reverse order from that described.

[0023] In the following embodiments, when it is stated that elements are connected, this includes not only the case where the elements are directly connected, but also the case where the elements are indirectly connected via an intervening element between the elements.

[0024] For example, when it is stated in this specification that components are electrically connected, this includes not only cases where the components are directly electrically connected, but also cases where the components are indirectly electrically connected through an intervening component.

[0025] A method for controlling a signal at an intersection according to one embodiment of the present disclosure may include a step of acquiring input data through a traffic condition collection device; a step of predicting traffic condition information at a predetermined time point from the input data through a trained traffic condition prediction model - the predicted traffic condition information being related to at least one of a maximum number of waiting vehicles or a passing traffic volume - a step of calculating an optimal aspect of the target intersection based on the predicted traffic condition information; and a step of controlling a signal at the target intersection based on the calculated optimal aspect.

[0026] According to one embodiment of the present disclosure, the traffic condition prediction model is trained based on a learning dataset consisting of traffic condition information at a first time point and traffic condition information at a second time point a predetermined time after the first time point, and the traffic condition information in the learning dataset may include at least one of a maximum number of waiting vehicles, a vehicle movement speed, a demand traffic volume, and a through traffic volume, as well as time information.

[0027] According to one embodiment of the present disclosure, the traffic condition prediction model is configured to output a prediction value related to a maximum number of waiting vehicles or a passing traffic volume based on the traffic condition information at the first time point, and parameters of the traffic condition prediction model can be updated and trained based on a difference between the prediction value and the traffic condition information at the second time point in the learning dataset.

[0028] According to one embodiment of the present disclosure, the step of calculating the optimal current time for the target intersection may further include a step of acquiring main movement direction information of the target intersection by comparing a predicted maximum number of waiting vehicles in the straight-on direction of the target intersection with a predicted maximum number of waiting vehicles in the turning direction of the target intersection; a step of calculating an average number of waiting vehicles between the predicted maximum number of waiting vehicles in the straight-on direction and the predicted maximum number of waiting vehicles in the turning direction; and a step of allocating the current time allocated to the turning direction to the current time for the straight-on direction if the predicted maximum number of waiting vehicles in the straight-on direction is greater than the calculated average number of waiting vehicles.

[0029] According to an embodiment of the present disclosure, the input data may include at least one of a maximum number of waiting vehicles, a vehicle movement speed, a demand traffic volume, and a through traffic volume, and time information.

[0030] According to an embodiment of the present disclosure, a computer-readable recording medium having a program recorded thereon for executing the intersection signal control method may be provided.

[0031] An electronic device according to an embodiment of the present disclosure may include a processor configured to acquire input data through a traffic condition collection device, predict traffic condition information at a predetermined time point from the input data through a trained traffic condition prediction model, calculate an optimal phase for the target intersection based on the predicted traffic condition information, and control a signal at the intersection based on the calculated optimal phase.

[0032] Hereinafter, with reference to Figures 1 to 12, a description will be given of an intersection signal control method according to one embodiment of the present disclosure, an electronic device (or server, hereinafter referred to as an electronic device) for performing the same, and / or an intersection signal control system.

[0033] FIG. 1 is a schematic diagram of an intersection signal control system 10 according to one embodiment of the present disclosure.

[0034] The intersection signal control system 10 according to an embodiment of the present disclosure may include a traffic condition collection device 100, a client terminal 200, a traffic signal controller 300, and / or an electronic device 1000.

[0035] The traffic condition collecting device 100 according to an embodiment of the present disclosure may include a device for collecting any data related to an intersection, including a camera and / or an optical signal sensor. For example, the camera of the traffic condition collecting device 100 may be embodied to acquire a traffic image of a target intersection and transmit the acquired traffic image to the electronic device 1000 through any transceiver. For example, the optical signal sensor of the traffic condition collecting device 100 may be embodied to acquire optical signal data for the target intersection and transmit the acquired optical signal data to the electronic device 1000 through any transceiver. For example, the traffic condition collecting device 100 may be embodied to calculate at least one of the maximum waiting vehicles by time, the vehicle movement speed, the passing traffic volume, the demand traffic volume, the maximum waiting vehicles, and / or the remaining waiting vehicles of the intersection based on the traffic image data and the optical signal data, and transmit the calculated data to the electronic device 1000 through any transceiver.

[0036] The electronic device 1000 according to an embodiment of the present disclosure may predict a traffic condition at a target intersection at a specific time after traffic image data and / or sensor signal data for the target intersection are acquired. Specifically, the electronic device 1000 may be configured to predict traffic condition information at the target intersection using a trained traffic condition prediction model in the form of a deep learning neural network, and to calculate an optimal phase based on the predicted traffic condition information to control a signal at the target intersection.

[0037] The client terminal 200 according to an embodiment of the present disclosure may receive predicted traffic condition information, calculated optimal phase information, and / or signal control information of the target intersection from the electronic device 1000 through any transceiver. Furthermore, the client terminal 200 may display the predicted traffic condition information and / or signal control information of the target intersection through any output unit (e.g., a display, etc.). Furthermore, the client terminal 200 may receive a request to instruct prediction of traffic condition information and / or a request to calculate the optimal phase of the target intersection through any input unit (e.g., a touchpad, a mouse, a keyboard, etc.). Meanwhile, the electronic device 1000 may be configured to perform an operation of predicting traffic condition information of the target intersection or calculating the optimal phase of the target intersection in response to a request from the client terminal 200.

[0038] The traffic signal controller 300 according to an embodiment of the present disclosure may receive a signal for controlling a traffic signal of a target intersection from the electronic device 1000 through an optional transceiver unit. In this case, the traffic signal controller 300 may perform an operation of controlling a signal of the target intersection based on the received signal.

[0039] The electronic device 1000 according to an embodiment of the present disclosure may include a transceiver unit 1100, a memory 1200, and a processor 1300.

[0040] The transceiver 1100 of the electronic device 1000 may communicate with any external device or external server. For example, the electronic device 1000 may obtain traffic image data of a target intersection and / or sensor signal data related to the target intersection from the traffic condition collecting device 100 through the transceiver 1100. For example, the electronic device 1000 may obtain traffic image data of a target intersection from a camera of the traffic condition collecting device 100 through the transceiver 1100. For example, the electronic device 1000 may obtain sensor signal data related to the target intersection from a sensor signal sensor of the traffic condition collecting device 100 through the transceiver 1100. For example, the electronic device 1000 may obtain traffic image data of the traffic condition collecting device 100 and / or input data calculated from the sensor signal sensor through the transceiver 1100. As an example, the electronic device 1000 may receive execution data for properly executing the trained traffic condition prediction model, including structure information, hierarchical information, calculation information, and / or parameter information of the trained traffic condition prediction model, through the transceiver 1100. As an example, the electronic device 1000 may transmit the calculated optimal current information to any external device including the client terminal 200 or any external server through the transceiver 1100. As an example, the electronic device 1000 may transmit a signal for traffic signal control to any external device including the traffic signal controller 300 or any external server through the transceiver 1100.

[0041] The electronic device 1000 can connect to a network through the transceiver 1100 to transmit and receive various data. The transceiver 1100 can be broadly classified into a wired type and a wireless type. Since the wired type and the wireless type have their respective advantages and disadvantages, the electronic device 1000 may be provided with both a wired type and a wireless type depending on the case. Here, in the case of the wireless type, a communication method of the WLAN (Wireless Local Area Network) series such as Wi-Fi can be mainly used. Alternatively, in the case of the wireless type, a communication method of the cellular communication, for example, a LTE or 5G series can be used. However, the wireless communication protocol is not limited to the above-mentioned examples, and any appropriate wireless type communication method can be used. In the case of the wired type, LAN (Local Area Network) and USB (Universal Serial Bus) communication are representative examples, and other methods are also possible.

[0042] The memory 1200 of the electronic device 1000 may store various information. Various data may be temporarily or semi-permanently stored in the memory 1200. Examples of the memory 1200 include a hard disk drive (HDD), a solid state drive (SSD), a flash memory, a read-only memory (ROM), and a random access memory (RAM). The memory 1200 may be provided in a form built into the electronic device 1000 or in a removable form. The memory 1200 may store various data necessary for the operation of the electronic device 1000, including an operating system (OS) for driving the electronic device 1000 and programs for operating each component of the electronic device 1000.

[0043] The processor 1300 may control the overall operation of the electronic device 1000. For example, the processor 1300 may control the overall operation of the electronic device 1000, including an operation of acquiring input data, an operation of predicting traffic condition information at a time when a predetermined time has elapsed from the input data through a trained traffic condition prediction model, an operation of calculating an optimal aspect of a target intersection based on the predicted traffic condition information, and / or an operation of controlling a signal of a target intersection. Specifically, the processor 1300 may load and execute a program for the overall operation of the electronic device 1000 from the memory 1200. The processor 1300 may be embodied as an AP (Application Processor), a CPU (Central Processing Unit), an MCU (Microcontroller Unit), or a similar device using hardware, software, or a combination thereof. In this case, the hardware may be provided in the form of an electronic circuit that processes an electrical signal to perform a control function, and the software may be provided in the form of a program or code that drives a hardware circuit.

[0044] FIG. 2 is a diagram illustrating the operation of the electronic device 1000 according to an embodiment of the present disclosure.

[0045] The electronic device 1000 according to an embodiment of the present disclosure may acquire input data calculated from traffic image data of a target intersection captured by a camera of the traffic condition collecting device 100 and sensory signal data sensed by a sensory signal sensor of the traffic condition collecting device 100 through the transceiver 1100. Specifically, the electronic device 1000 may acquire input data related to at least one of the maximum waiting vehicles by time, vehicle movement speed, through traffic volume, demand traffic volume, maximum waiting vehicles, and / or remaining waiting vehicles of the target intersection acquired by analyzing the traffic image data and the sensory signal data. Here, the through traffic volume may mean the number of vehicles passing during one aspect. The demand traffic volume may mean the number of vehicles obtained by adding up the through traffic volume and the number of remaining waiting vehicles. The maximum waiting vehicles may mean the number of vehicles in a lane with the largest number of waiting vehicles before the signal changes from red to green. The remaining waiting vehicles may mean the number of vehicles remaining without being able to pass immediately after the signal changes from green to red. Meanwhile, the aspect may refer to a time when a green light is displayed, indicating that a traffic flow that can pass or a group of vehicles or people that can pass at the same time has priority. In the above, the electronic device 1000 has been described as acquiring the calculated input data through the transceiver 1100. However, this is merely an example, and the electronic device 1000 may be embodied to analyze the traffic image data and the sensor signal data using any algorithm and directly calculate the input data.

[0046] The electronic device 1000 according to an embodiment of the present disclosure may acquire execution data of the trained traffic condition prediction model through the transceiver 1100. Here, the execution data may include any data for appropriately executing the trained traffic condition prediction model, including structure information, hierarchical information, operation information, and / or parameter information of the trained traffic condition prediction model.

[0047] The electronic device 1000 may perform an operation of predicting traffic condition information at a time when a predetermined time has elapsed from the input data through the model. More specifically, the electronic device 1000 may input input data including at least one of a maximum number of waiting vehicles, a vehicle movement speed, a demand traffic volume, and a passing traffic volume, and time information, to an input layer of the trained traffic condition prediction model. At this time, the trained traffic condition prediction model may predict traffic condition information at a time when a predetermined time has elapsed from a time corresponding to the input data. At this time, the electronic device 1000 may obtain predicted traffic condition information through an output layer of the trained traffic condition prediction model.

[0048] Meanwhile, the traffic situation prediction model may be trained based on a learning data set composed of traffic situation information (e.g., maximum number of waiting vehicles, vehicle movement speed, demand traffic volume, and / or through traffic volume) at a first time point and traffic situation information (e.g., maximum number of waiting vehicles, vehicle movement speed, demand traffic volume, and / or through traffic volume) at a second time point at which a predetermined time has elapsed since the first time point. Specifically, the traffic situation prediction model may be configured to output a predicted value related to the maximum number of waiting vehicles or through traffic volume based on the traffic situation information at the first time point, and may be trained by updating parameters of the traffic situation prediction model based on a difference between the predicted value and the traffic situation information at the second time point of the learning data set. For example, the traffic situation prediction model may be trained by updating parameters of the traffic situation prediction model so as to output a predicted value approximated to the traffic situation information at the second time point using the traffic situation information at the second time point as ground-truth. Thus, the traffic situation prediction model after completion of training may perform a task of predicting traffic situation information at a time point at which a predetermined time has elapsed since the time point corresponding to the input data based on the input data.

[0049] The method of training and / or inference of the traffic condition prediction model will be described in more detail with reference to FIGS.

[0050] The electronic device 1000 according to an embodiment of the present disclosure may perform an operation of calculating an optimal aspect of a target intersection based on predicted traffic condition information. More specifically, the electronic device 1000 may obtain main movement direction information of the target intersection by comparing predicted traffic condition information of a first direction (e.g., straight direction) of the target intersection with predicted traffic condition information of a second direction (e.g., turning direction) of the target intersection. Furthermore, the electronic device 1000 may calculate average traffic condition information of the predicted traffic condition information of the first direction (e.g., straight direction) and the predicted traffic condition information of the second direction. At this time, the electronic device 1000 may allocate an actual time of the target intersection by comparing the predicted traffic condition information of the first direction (e.g., straight direction) with the calculated average traffic condition information. The content of calculating the optimal aspect of the target intersection will be described in more detail with reference to FIGS. 7 to 8.

[0051] The electronic device 1000 according to an embodiment of the present disclosure may perform an operation of controlling a traffic light at a target intersection based on the calculated optimal timing. For example, the electronic device 1000 may be embodied to control a traffic light at a target intersection based on a timing allocated to a first direction (e.g., a straight direction) and a second direction (e.g., a turning direction) of the target intersection.

[0052] Furthermore, the electronic device 1000 according to an embodiment of the present disclosure may be configured to transmit a signal for traffic signal control based on the optimal phase calculated by the traffic signal controller 300 through the transceiver 1100. The traffic signal controller may receive the signal for traffic signal control and control the signal of the target intersection based on the received signal.

[0053] Furthermore, the electronic device 1000 according to an embodiment of the present disclosure may be configured to transmit any information related to a traffic condition prediction result, including the calculated optimal current information, to the client terminal 200 via the transceiver 1100. Specifically, the electronic device 1000 may be configured to provide any information related to a traffic condition prediction result to a user through a user interface. The user interface will be described in more detail with reference to FIGS. 9 to 12.

[0054] Hereinafter, the operation of the electronic device 1000 according to an embodiment of the present disclosure and a method for controlling a traffic light at an intersection performed by the electronic device 1000 will be described in more detail with reference to Figures 3 to 12. Meanwhile, in describing the method for controlling a traffic light at an intersection, some embodiments that overlap with the contents previously described with reference to Figure 2 may be omitted. However, this is merely for convenience of description and should not be construed as being restrictive.

[0055] FIG. 3 is a flowchart illustrating a method for controlling a signal at an intersection according to an embodiment of the present disclosure.

[0056] A method for controlling a traffic signal at an intersection according to one embodiment of the present disclosure may include a step of acquiring input data through a traffic condition collection device (S1000), a step of predicting traffic condition information from the input data at a predetermined time point through a trained traffic condition prediction model (S2000), a step of calculating an optimal phase for the target intersection based on the predicted traffic condition information (S3000), and / or a step of controlling a traffic signal at the target intersection based on the calculated optimal phase (S4000).

[0057] In the step of acquiring input data through the traffic condition collecting device (S1000), the electronic device 1000 can acquire input data related to at least one of the maximum waiting vehicles by time, vehicle movement speed, passing traffic volume, demand traffic volume, maximum waiting vehicles, and / or remaining waiting vehicles at the target intersection from the traffic condition collecting device 100 through the transceiver 1100.

[0058] In the step of predicting traffic condition information at a time when a predetermined time has elapsed from the input data through the trained traffic condition prediction model (S2000), the electronic device 1000 may acquire execution data of the trained traffic condition prediction model through the transceiver unit 1100. Here, the execution data may include any data for appropriately executing the trained traffic condition prediction model, including structure information, hierarchical information, calculation information, and / or parameter information of the trained traffic condition prediction model.

[0059] Furthermore, in the step of predicting traffic condition information at a predetermined time point from the input data through the trained traffic condition prediction model (S2000), the electronic device 1000 can perform an operation of predicting traffic condition information at a predetermined time point from the input data through the trained traffic condition prediction model.

[0060] 4 is a diagram for explaining how traffic condition information is predicted using a trained traffic condition prediction model according to an embodiment of the present disclosure. Specifically, FIG. 4(a) is a diagram for explaining how traffic condition information predicted through a trained traffic condition prediction model according to an embodiment of the present disclosure is obtained.

[0061] Specifically, the electronic device 1000 can input input data including at least one of the maximum number of waiting vehicles, the vehicle movement speed, the demand traffic volume, and the passing traffic volume, and time information, to an input layer of the traffic condition prediction model that has been trained. At this time, the traffic condition prediction model that has been trained can calculate traffic condition information related to the maximum waiting vehicles (or waiting queue) and / or the passing traffic volume at a time when a predetermined time has elapsed from a time corresponding to the input data. At this time, the electronic device 1000 can acquire predicted traffic condition information related to the maximum waiting vehicles and / or the passing traffic volume at a time when a predetermined time has elapsed from a time corresponding to the input data through an output layer of the traffic condition prediction model that has been trained.

[0062] FIG. 4(b) is a graph comparing traffic condition information predicted through a traffic condition prediction model that has completed training according to an embodiment of the present disclosure with actual traffic condition information.

[0063] 4(b), it can be seen that the traffic condition information predicted through the trained traffic condition prediction model according to one embodiment of the present disclosure (e.g., waiting queue prediction in FIG. 4(b)) and the actual traffic condition information (e.g., actual waiting queue (meters) in FIG. 4(b)) show similar trends. Furthermore, it can be seen that the through traffic volume predicted through the trained traffic condition prediction model according to one embodiment of the present disclosure (e.g., through traffic volume prediction in FIG. 4(b)) and the actual through traffic volume (e.g., actual through traffic volume (number of vehicles) in FIG. 4(b)) show similar trends.

[0064] FIG. 5 is a diagram illustrating an example of input variables and output variables of a traffic condition prediction model that has completed training according to an embodiment of the present disclosure.

[0065] Meanwhile, the input variables of the input data input to the traffic condition prediction model after training can be set to at least one of the maximum number of waiting vehicles, vehicle movement speed, demand traffic volume, and through traffic volume, and time information.

[0066] As an example, a traffic condition prediction model that has completed training may be configured to receive the maximum number of waiting vehicles, vehicle movement speed, demand traffic volume, through traffic volume, and / or a combination thereof, as well as time information (e.g., a first time point) as input variables, and to output the maximum waiting vehicles at a point in time after a predetermined time has elapsed (e.g., a second time point) as an output variable.

[0067] However, the input variables and output variables illustrated in FIG. 5 are merely examples, and the input variable and output variable forms may be set to any appropriate variables in order to achieve the purpose of more accurately predicting future traffic condition information.

[0068] FIG. 6 is a diagram illustrating one aspect of training a traffic condition prediction model according to an embodiment of the present disclosure.

[0069] A traffic condition prediction model according to an embodiment of the present disclosure may be trained based on a learning dataset including a plurality of data (e.g., D1, D2 to Dn in FIG. 6) including traffic condition information at a first time point and traffic condition information at a second time point after a predetermined time has elapsed since the first time point. Here, the traffic condition information at the first time point of the learning dataset may include the maximum number of waiting vehicles, vehicle movement speed, demand traffic volume, through traffic volume, and / or a combination thereof at the intersection at the first time point corresponding to the input variables of the traffic condition prediction model, and time information at the first time point. Furthermore, the traffic condition information at the second time point of the learning dataset may include the maximum number of waiting vehicles, vehicle movement speed, demand traffic volume, through traffic volume, and / or a combination thereof at the intersection at the second time point of the traffic condition prediction model, and time information at the second time point.

[0070] More specifically, the traffic situation prediction model may be configured to output a predicted value related to the maximum number of waiting vehicles or a passing traffic volume based on traffic situation information at a first time point. At this time, the traffic situation prediction model may be trained by updating parameters of the traffic situation prediction model based on a difference between the predicted value and traffic situation information at a second time point of the learning dataset. For example, the traffic situation prediction model may be trained by updating parameters of the traffic situation prediction model to output a predicted value that is approximated to the actual maximum number of waiting vehicles based on a difference between a predicted value related to the maximum number of waiting vehicles and an actual maximum number of waiting vehicles at a second time point of the learning dataset (first ground truth). For example, the traffic situation prediction model may be trained by updating parameters of the traffic situation prediction model to output a predicted value that is approximated to the actual passing traffic volume based on a difference between a predicted value related to the passing traffic volume and an actual passing traffic volume at a second time point of the learning dataset (second ground truth).

[0071] Meanwhile, although not shown in Figures 4 to 6, a traffic condition prediction model that has completed training in one embodiment of the present disclosure may be composed of a first model for predicting traffic condition information for the straight-through direction of the intersection and a second model for predicting traffic condition information for the turning direction of the intersection.

[0072] For example, the first model may be trained in a similar manner to that described above using a learning data set including traffic condition information for a straight-through direction at an intersection at a first time point and traffic condition information for a straight-through direction at a second time point a predetermined time after the first time point. In this case, the first model that has completed training may perform a task of predicting a maximum waiting vehicle and / or a passing traffic volume for a straight-through direction based on input data for a target intersection in the straight-through direction.

[0073] For example, the second model may be trained in a similar manner to that described above using a learning data set including traffic condition information for a turning direction (e.g., left turning direction) at a first time point of an intersection and traffic condition information for a turning direction (e.g., left turning direction) at a second time point a predetermined time after the first time point. In this case, the second model that has completed training may perform a task of predicting the maximum number of waiting vehicles and / or passing traffic volume for a turning direction (e.g., left turning direction) based on input data for the turning direction (e.g., left turning direction) of a target intersection.

[0074] In the step of calculating an optimal aspect of a target intersection based on the predicted traffic condition information (S3000), the electronic device 1000 may perform an operation of calculating an optimal aspect of a target intersection based on the traffic condition information predicted through step S2000. For example, the electronic device 1000 may perform an operation of distributing an aspect time of a first time point of the target intersection based on the predicted maximum number of waiting vehicles (or a predicted waiting queue) and / or a predicted passing traffic volume through step S2000, and calculating an optimal aspect of a second time point.

[0075] Hereinafter, the step of calculating the optimal current aspect of a target intersection (S3000) according to an embodiment of the present disclosure will be described in more detail with reference to Figures 7 and 8. Figure 7 is a flowchart embodying the step of calculating the optimal current aspect of a target intersection according to an embodiment of the present disclosure (S3000). Figure 8 is a diagram for explaining one aspect of calculating the optimal current aspect of a target intersection according to an embodiment of the present disclosure.

[0076] The step of calculating the optimal current time for a target intersection according to one embodiment of the present disclosure (S3000) may further include a step of acquiring main movement flow direction information for the target intersection (S3100), a step of calculating an average number of waiting vehicles between a predicted maximum number of waiting vehicles in the straight-line direction and a predicted maximum number of waiting vehicles in the turning direction (S3200), and a step of distributing the current time allocated in the turning direction to the current time in the straight-line direction if the predicted maximum number of waiting vehicles in the straight-line direction is greater than the calculated average number of waiting vehicles (S3300).

[0077] In the step of acquiring main movement flow direction information of the target intersection (S3100), the electronic device 1000 may acquire main movement direction information of the target intersection by comparing the predicted traffic condition information for the straight direction of the target intersection acquired through the step S2000 with the predicted traffic condition information for the turning direction of the target intersection acquired through the step S2000. For example, the electronic device 1000 may compare the predicted maximum number of waiting vehicles for the straight direction of the target intersection acquired through the step S2000 (e.g., 19 vehicles in FIG. 8(b)) with the predicted maximum number of waiting vehicles for the turning direction of the target intersection acquired through the step S2000 (e.g., 6 vehicles in FIG. 8(b)), and acquire the direction showing the larger maximum number of waiting vehicles (e.g., the straight direction in FIG. 8) as the main movement flow direction based on the comparison result. Meanwhile, in FIG. 8, it has been described that the main movement flow direction information is acquired based on the maximum number of waiting vehicles. However, this is merely an example, and the electronic device 1000 may be embodied to determine the main traffic flow direction of the target intersection using any data, including predicted through traffic volume, existing signal times, and / or demand traffic volume through step S2000.

[0078] In the step of calculating an average number of waiting vehicles of the predicted maximum number of waiting vehicles in the straight direction and the predicted maximum number of waiting vehicles in the turning direction (S3200), the electronic device 1000 may be embodied to calculate an average number of waiting vehicles of the target intersection based on the predicted maximum number of waiting vehicles in the straight direction of the target intersection and the predicted maximum number of waiting vehicles in the turning direction of the target intersection. For example, the electronic device 1000 may be embodied to calculate an average number of waiting vehicles (e.g., 12.5 vehicles in FIG. 8(b)) of the target intersection based on the predicted maximum number of waiting vehicles in the straight direction of the target intersection (e.g., 19 vehicles in FIG. 8(b)) and the predicted maximum number of waiting vehicles in the turning direction of the target intersection (e.g., 6 vehicles in FIG. 8(b)).

[0079] In the step S3300 of distributing the current time allocated in the rotation direction to the current time in the straight direction when the predicted maximum number of waiting vehicles in the straight direction is greater than the calculated average number of waiting vehicles, the electronic device 1000 may be embodied to calculate an optimal current time for the target intersection by distributing the current time allocated in the rotation direction to the current time in the straight direction when the predicted maximum number of waiting vehicles in the straight direction is greater than the average number of waiting vehicles calculated through step S3200. For example, the electronic device 1000 may be embodied to calculate an optimal current time for the target intersection by distributing a part of the current time allocated in the rotation direction to the current time in the straight direction (for example, distributing 6 seconds of the current time in the rotation direction to the current time in the straight direction in FIG. 8(b)) when the predicted maximum number of waiting vehicles in the straight direction (e.g., 19 vehicles in FIG. 8(b)) is greater than the average number of waiting vehicles in the target intersection calculated through step S3200 (e.g., 12.5 vehicles in FIG. 8(b)). In this case, the current indication time to be distributed (e.g., 6 seconds) may be determined based on the difference (e.g., difference of 6.5 vehicles) between the maximum number of waiting vehicles predicted through step S2000 (e.g., straight ahead (19 vehicles) or turning direction (6 vehicles)) and the average maximum number of vehicles at the target intersection calculated through step S3200 (e.g., 12.5 vehicles). For example, the current indication time to be distributed may be determined as a value proportional to the difference (e.g., difference of 6.5 vehicles) between the maximum number of waiting vehicles predicted through step S2000 (e.g., straight ahead (19 vehicles) or turning direction (6 vehicles)) and the average maximum number of vehicles at the target intersection calculated through step S3200 (e.g., 12.5 vehicles). However, this is merely an example, and the current indication time to be distributed to the current indication time for the straight direction may be calculated using any appropriate method.

[0080] Referring again to FIG. 3, the intersection signal control method according to the embodiment of the present disclosure may include controlling the signal of the target intersection based on the calculated optimal phase (S4000). In controlling the signal of the target intersection based on the calculated optimal phase (S4000), the electronic device 1000 may be embodied to control the signal of the target intersection based on the optimal phase calculated through step S3000. For example, the electronic device 1000 may control the signal of the target intersection at the first phase (i.e., 94 seconds (existing phase (88 seconds) + proposed phase (6 seconds)) for the straight direction and at the second phase (i.e., 29 seconds (existing phase (35 seconds) + proposed phase (-6 seconds)) for the turning direction based on the optimal phase calculated through step S3000.

[0081] 3 to 8, the electronic device 1000 has been described as controlling the signal of the target intersection. However, this is merely for convenience of explanation, and the electronic device 1000 may be configured to transmit the calculated optimal aspect to the traffic signal controller 300, and the traffic signal controller 300 may control the signal of the target intersection. Alternatively, the electronic device 1000 may be configured to transmit the calculated optimal aspect to the client terminal 200, and the traffic signal controller 300 may control the signal of the target intersection based on a signal control request from the client.

[0082] Hereinafter, a user interface that may be provided to a client terminal 200 according to an embodiment of the present disclosure will be described with reference to Figures 9 to 12. Figures 9 to 12 are diagrams for explaining one aspect of a user interface.

[0083] 9, a user interface according to an embodiment of the present disclosure may receive an input for selecting a type of traffic condition prediction model (e.g., the STGCN model, GraphWaveNet (GWNET) model, and STGCN-Cov model in FIG. 9) and input variables of the traffic condition prediction model through any input unit (e.g., a mouse, a touchpad, a keyboard, etc.) of the client terminal 200. In addition, the user interface may receive an input for selecting a time to be predicted (e.g., 5 minutes later, 10 minutes later, 15 minutes later, 20 minutes later, (omitted) 60 minutes later, etc.) through any input unit (e.g., a mouse, a touchpad, a keyboard, etc.) of the client terminal 200.

[0084] At this time, the user interface may be provided to output information on a prediction result based on a selected prediction model, input variables, and a time to be predicted to a user through an arbitrary output unit (e.g., a display, etc.) of the client terminal 200. For example, the user interface may be provided to output a prediction result for a prediction target (e.g., vehicle movement speed, passing traffic volume, waiting queue, etc. in FIG. 9) to a user through an arbitrary output unit of the client terminal 200. For example, the user interface may be provided to output information on learning data used in the prediction model to a user through an arbitrary output unit of the client terminal 200. For example, the user interface may be provided to output information on indexes indicating the performance of the prediction model (e.g., MAE value, RMSE value in FIG. 9) to a user through an arbitrary output unit of the client terminal 200.

[0085] Referring to FIG. 10, a user interface according to an embodiment of the present disclosure may be provided to visualize and output result information predicted through a traffic condition prediction model to a user through an arbitrary output unit of the client terminal 200. For example, the user interface may be provided to visualize and output result information predicted through a traffic condition prediction model on a map to a user. More specifically, the user interface may display an intersection icon (e.g., M1 in FIG. 10) representing an intersection on a map, and display the degree of congestion at the intersection based on the prediction result on the map (e.g., a road with a smooth intersection is displayed in a first color (e.g., green), a road with a congested intersection is displayed in a second color (e.g., red), or a road with a normal intersection is displayed in a third color (e.g., yellow)), and may be provided to a user. For example, the user interface may be provided to output result information predicted through a traffic condition prediction model (e.g., average waiting queue prediction value, average moving speed prediction value in FIG. 10) and prediction accuracy of the predicted result information (e.g., average waiting queue prediction accuracy, average moving speed prediction accuracy in FIG. 10) to a user.

[0086] 11, a user interface according to an embodiment of the present disclosure may be provided to a user in the form of a table showing information on the results of prediction through a traffic condition prediction model through an arbitrary output unit of the client terminal 200. For example, the user interface may be provided to a user in the form of a table showing analysis date and time, intersection name, time (e.g., day of the week), entrance direction to the intersection, actual speed, predicted speed value, prediction accuracy, actual waiting queue, and / or predicted waiting queue value, etc., predicted through a traffic condition prediction model through an arbitrary output unit of the client terminal 200.

[0087] 12, a user interface according to an embodiment of the present disclosure may be provided to a user to output a first map displaying a current traffic situation and a second map displaying a traffic situation simulated based on a proposed current indication (e.g., traffic situation when the proposed current indication is applied 30 minutes or 60 minutes later) through an arbitrary output unit of the client terminal 200. In this case, the user interface may be provided to a user to output information on the proposed current indication, the through traffic volume (or waiting queue) before and after the application of the simulation, and / or the intersection improvement effect through an arbitrary output unit of the client terminal 200.

[0088] However, the information illustrated in Figures 9 to 12 is merely exemplary, and the user interface may be configured to provide any appropriate information to the user in any appropriate form as needed.

[0089] According to an embodiment of the present disclosure, a method for controlling a traffic signal at an intersection and an electronic device for performing the same can provide an effect of predicting traffic condition information at a predetermined time point after an analysis time point through a trained traffic condition prediction model.

[0090] According to an embodiment of the present disclosure, a method for controlling traffic signals at an intersection and an electronic device for performing the same can provide an effect of controlling traffic signals in a direction that reduces congestion sections by distributing signals in consideration of the main flow direction of traffic at the intersection.

[0091] According to an embodiment of the present disclosure, a method for controlling intersection signals and an electronic device for performing the same can provide an effect of increasing user convenience by allowing a user to visually check, through a user interface, information on predicted results using a traffic condition prediction model and traffic conditions resulting from a calculated optimal aspect.

[0092] Various operations of the electronic device 1000 described above may be stored in a memory 1200 of the electronic device 1000 , and a processor 1300 of the electronic device 1000 may be configured to perform the operations stored in the memory 1200 .

[0093] The features, structures, effects, etc. described in the above embodiments are included in at least one embodiment of the present disclosure and are not necessarily limited to only one embodiment. Furthermore, the features, structures, effects, etc. exemplified in each embodiment can be combined or modified in other embodiments by a person having ordinary knowledge in the field to which the embodiment belongs. Therefore, the contents related to such combinations and modifications should be interpreted as being included in the scope of the present disclosure.

[0094] In addition, although the above description focuses on the embodiment, this is merely an example and does not limit the present disclosure. A person having ordinary skill in the art to which the present disclosure belongs will understand that various modifications and applications not exemplified above are possible within the scope of the essential characteristics of the present embodiment. In other words, each component specifically illustrated in the embodiment can be modified and implemented. Differences related to such modifications and applications should be interpreted as being included in the scope of the present disclosure defined in the appended claims. [Industrial Applicability]

[0095] The above-described intersection signal control method and apparatus for performing the same can be applied to the fields of traffic situation prediction and traffic signal control.

Claims

1. 1. A method for controlling a traffic light at an intersection in which an electronic device controls a traffic light, comprising: acquiring input data through a traffic condition collection device; a step of predicting traffic condition information at the selected time based on the selected traffic condition prediction model, the selected input variables, and the selected time when a type of a traffic condition prediction model used to predict traffic condition information among the trained traffic condition prediction models, an input variable of the input data to be input to the traffic condition prediction model, and a time to be predicted through the traffic condition prediction model are selected by a client terminal - the input variables include a combination of a maximum number of waiting vehicles, a vehicle movement speed, a demand traffic volume, and a through traffic volume, and time information, and the predicted traffic condition information is related to at least one of the maximum number of waiting vehicles and the through traffic volume -; calculating an optimal timing for the target intersection based on the predicted traffic condition information; receiving a signal control request from the client terminal on which a first map displaying the calculated optimal aspect and a current traffic situation and a second map displaying a traffic situation at the selected time simulated based on the calculated optimal aspect are provided through a user interface; and controlling a signal at the target intersection based on the calculated optimal aspect in response to the received signal control request; The step of calculating an optimal phase of the target intersection comprises: acquiring main movement direction information of the target intersection by comparing a predicted maximum number of waiting vehicles in a straight-through direction of the target intersection with a predicted maximum number of waiting vehicles in a turning direction of the target intersection; calculating an average number of waiting vehicles between the predicted maximum number of waiting vehicles in the straight-ahead direction and the predicted maximum number of waiting vehicles in the turning direction; and The present signal control method for an intersection further includes distributing a current time allocated to the rotation direction to a current time in the straight direction when the predicted maximum number of waiting vehicles in the straight direction is greater than the calculated average number of waiting vehicles.

2. The traffic situation prediction model is The system is trained based on a learning data set including traffic condition information at a first time point and traffic condition information at a second time point a predetermined time after the first time point, The intersection signal control method according to claim 1 , wherein the traffic condition information in the learning data set includes at least one of a maximum number of waiting vehicles, a vehicle movement speed, a demand traffic volume, and a through traffic volume, and time information.

3. The traffic situation prediction model is a prediction value related to a maximum number of waiting vehicles or a passing traffic volume based on the traffic condition information at the first time point; The intersection signal control method according to claim 2 , wherein parameters of the traffic condition prediction model are updated and trained based on a difference between the predicted value and the traffic condition information at the second time point in the learning data set.

4. A computer-readable recording medium having recorded thereon a program for causing a computer to execute the method according to any one of claims 1 to 3.

5. 1. An electronic device for controlling traffic lights at an intersection, comprising: Obtain input data through a traffic condition collector, When a type of a traffic situation prediction model used to predict traffic situation information among the trained traffic situation prediction models, input variables of the input data to be input to the traffic situation prediction model, and a time to be predicted through the traffic situation prediction model are selected at a client terminal, traffic situation information at the selected time is predicted based on the selected traffic situation prediction model, the selected input variables, and the selected time - the input variables include a combination of a maximum number of waiting vehicles, a vehicle movement speed, a demand traffic volume, and a through traffic volume, and time information, and the predicted traffic situation information is related to at least one of the maximum number of waiting vehicles and the through traffic volume -; Calculating an optimal timing for the target intersection based on the predicted traffic condition information; receiving a signal control request from the client terminal, in which a first map displaying the calculated optimal aspect and a current traffic situation, and a second map displaying a traffic situation at the selected time simulated based on the calculated optimal aspect are provided through a user interface; a processor configured to control a signal at the target intersection based on the calculated optimal phase in response to the received signal control request; In calculating the optimal current aspect of the target intersection, the processor acquiring main movement direction information of the target intersection by comparing the predicted maximum number of waiting vehicles in the straight-through direction of the target intersection with the predicted maximum number of waiting vehicles in the turning direction of the target intersection; Calculating an average number of waiting vehicles between the predicted maximum number of waiting vehicles in the straight-ahead direction and the predicted maximum number of waiting vehicles in the turning direction; an electronic device configured to distribute the current time allocated to the rotation direction to the current time in the straight ahead direction when the predicted maximum number of waiting vehicles in the straight ahead direction is greater than the calculated average number of waiting vehicles.

Images