ES2608911A1 - System and method of control for auto-regulated traffic lights (Machine-translation by Google Translate, not legally binding) - Google Patents

Abstract

System and control method for self-regulated traffic lights. A method for the control of self-regulated traffic lights, executed in a system comprising a first data acquisition subsystem (100); a second subsystem of optimization of phases and cycle of the traffic lights (200) that regulate an intersection between two roads of terrestrial transport; and a third wireless communication subsystem (300) between intersections; wherein said method is characterized in that it comprises: a first data acquisition process in the first subsystem (100); a second process of optimization of phases and cycle of the traffic lights in the second subsystem (200); and a third wireless communication process between intersections of an urban area through the third subsystem (300). (Machine-translation by Google Translate, not legally binding)

Description

DESCRIPTION

CONTROL SYSTEM AND METHOD FOR SELF-REGULATED SEMAPHORS

SECTOR OF THE TECHNIQUE 5

The present invention is framed within intelligent transport systems.

STATE OF THE TECHNIQUE

 10

Currently there are different systems to regulate the times that configure the operation of traffic lights, mainly cycle times and phases. Specifically, for those acting on the basis of traffic data, these systems have traditionally required complex computational algorithms (eg, neural networks or fuzzy logic) to process all traffic information collected in the area of interest 15 and, subsequently, (re) configure traffic lights to improve their effectiveness.

Currently, two of the most used systems are SCOOT (Split Cycle and Offset Optimization Technique) and SCATS (Sydney Coordinated Adaptive Traffic). The main problem presented by these systems is their high cost, both computational and the necessary infrastructure for their deployment. All data collected from vehicular traffic must be sent through an extensive communications network to the central servers, which process the information to obtain optimal system regulation.

 25

Due to this complexity, distributed traffic control systems have emerged. This paradigm consists of the installation of different autonomous traffic control agents or nodes, which coordinate among themselves without human intervention or with minimal intervention in case of emergency or for greater security. This coordination can be done explicitly, by exchanging messages, or implicitly, 30 by self-coordination.

DESCRIPTION OF THE INVENTION

The object of the present invention is to improve the fluidity of vehicle traffic at 35 intersections regulated by traffic lights. For this, the present invention comprises a

self-manageable system, capable of making decisions for itself, without using a centralized computing system. By means of the present invention it is possible to reduce the average waiting time at the intersection by the conductors, as well as to reduce the emission of polluting gases into the atmosphere.

 5

The system object of the present invention is an autonomous system, capable of estimating the real state of traffic at an intersection and, without external help or additional data, managing the traffic light times that regulate said intersection.

More specifically, the control system for self-regulated traffic lights comprises: 10

a) a first traffic data acquisition subsystem comprising means for knowing the number of vehicles and vehicle occupants, so that not only traffic density is known but the density of people in real-time traffic, which can greatly affect and is a great differentiator 15 compared to other proposals;

b) a second optimization subsystem that appropriately adjusts the cycle times and the different phases that configure the traffic lights that regulate the intersection by means of a probabilistic method capable of predicting if there is an incipient congestion and acting autonomously; and 20

c) a third subsystem of wireless communication between intersections, giving rise to a system not only autonomous but also distributed for the regulation of traffic in a city.

With this self-regulation of the traffic phase phases, a more fluid circulation is achieved, allowing shorter waiting times for drivers, as well as reducing emissions of polluting gases into the atmosphere. Among the advantages compared to other alternatives, there is a much lower processing load than other proposals with equal or greater efficiency, which facilitates its implementation and deployment in real systems, and the capacity for self-adjustment without the need for human intervention, or with minimal intervention from an operations center.

Throughout the description and the claims the word "comprises" and its variants are not intended to exclude other technical characteristics, additives, components or steps. For those skilled in the art, other objects, advantages and features of the invention will be derived partly from the description and partly from the practice of the invention. The

The following examples and drawings are provided by way of illustration, and are not intended to restrict the present invention. In addition, the present invention covers all possible combinations of particular and preferred embodiments indicated herein.

BRIEF DESCRIPTION OF THE FIGURES 5

A series of drawings that help to better understand the invention and that expressly relate to an embodiment of said invention which is presented as a non-limiting example thereof is described very briefly below.

 10

 FIG. 1 shows a flow chart of the control method executed by the control system for self-regulated traffic lights object of the present invention.

 FIG. 2 shows a schematic figure of an application example of the control system for self-regulated traffic lights object of the present invention.

 fifteen

PREFERRED EMBODIMENT OF THE INVENTION

The present invention relates to a new control method and a new control system capable of preventing traffic congestion regulated by traffic lights and acting before it occurs. For this, the system that composes it is divided into three 20 subsystems:

1) a first data acquisition subsystem (100) which in turn is divided into:

1.a) a first image processing module (101);

1.b) a second signal processing module (102); 25

As can be seen in Figure 1, the first data acquisition subsystem (100) in turn executes a first image processing stage in a first image processing module (101). This module or image processing stage (101) is configured to obtain statistics on the actual traffic of vehicles on a road using an image processing algorithm based on visual events. As an example of a processing algorithm, you could use background substraction (M. Piccardi, "Background subtraction techniques: a review", IEEE Intl Conf Systems, Man & Cybernetics, 2004) or improve this technique by including a preprocessing (for example with gray conversion, resolution adjustment and histogram equalization) and a global processing after 35 background substraction using a temporal mean of contrast and detection of

events by regions using Bayesian networks (“Foundations of Artificial Intelligence”, Handbook of Knowledge Representation, Edited by Frank van Harmelen, Vladimir Lifschitz and Bruce Porter, Vol. 3, pp. 1-1006, 2008).

On the other hand, the second signal processing module (102) is configured to estimate the number of occupants of the vehicles based on a people counting algorithm based on the cellular telephone signal of mobile devices captured by a communication subsystem (300) similar to how it is done for example with WiFi signals (Wei Xi et al., "Electronic Frog Eye: Counting crowd using WiFi", in Proc. IEEE Conference on Computer Communications, pp. 361-369 , 2014). Therefore, according to the present invention the number of occupants will correspond to the number of occupants carrying a mobile device turned on but not necessarily active, that is, it is not necessary that the mobile device of a vehicle occupant is sending or receiving mobile (cellular) voice or data traffic to be detected (counted). fifteen

With this information about the occupants of the vehicles it is possible to compensate the data obtained in the traffic of vehicles, being able to have many vehicles in one way, but with few travelers, while in the other way we can meet with few vehicles and a larger number of occupants, then adequately compensating the signaling of the traffic light.

The second phase and cycle optimization subsystem (200) is intended to (re) configure both the phases of the traffic lights that control an intersection (i.e. times in red, green and yellow, as well as the emptying interval, i.e. , when all traffic lights 25 are in red to empty the intersection) as the cycle, understood as the aggregate sum of all phases of a traffic light.

For this, the second subsystem (200) with the data obtained in the first subsystem (100) begins with a stage of a state calculation (201), where 30 is detected if there is a congestion (202). If there is no congestion, the intervals and cycles (203) are maintained at the traffic lights and a communications counter to know if the third communications subsystem (300) is activated or not is set to zero (204), closing the cycle with the continuous acquisition of data in the first subsystem (100).

 35

On the contrary, if in the state calculation (201) it is detected that there is congestion (202),

then it is evaluated whether congestion is incipient (205) taking into account both the number of vehicles and the number of occupants of the vehicles. To determine if congestion is incipient or if there is congestion directly, we compare the state of the route that is calculated from the data collected with a pre-established range (maximum and minimum value) so that if the calculated state (201) is within the range it is assumed that the congestion is incipient and a probabilistic reconfiguration stage (206) of the intervals and the cycle is carried out and if it is outside that range, by excess, it is assumed that there is congestion and it is passed to a deterministic reconfiguration stage (207) of the intervals and cycles.

 10

In either of the two reconfigurations (206, 207) the value of the communications counter (208) is increased to subsequently establish if said communications counter is greater than a configured value, in order to determine when the third subsystem must be activated (300) of communications.

 fifteen

Thus, for example, if it is assumed that congestion is in an incipient state, then the reconfiguration is probabilistic, that is, it may or may not be reconfigured depending on the probability "p". This probability function should be greater as it is closer to the upper limit of the range, and may be a linear, logarithmic or exponential probability function, among others. On the contrary, if we assume that there is congestion then phase 20 in green is always increased, that is, with probability 1.

The communications subsystem (300) includes several communication interfaces to (i) allow the capture of mobile signal that will then be used by the first data acquisition subsystem (100) and (ii) allow the sending and receiving of messages between 25 intersections of an urban area. Optionally, communication with a control center can be established for security reasons.

Initially, each intersection is self-regulating as indicated in the operation of the second subsystem (200), sending the reconfiguration to the control systems 30 of the intersections themselves. This means that, for example, several intersections that make up a road are also self-regulating from the point of view of the road, that is, the correct synchrony between intersections to facilitate the flow of traffic along that road.

 35

However, the third communication subsystem (300) is designed to include a

communication protocol between intersections composed of several messages whose purpose is:

a) report on the state of each intersection, for example, if you are applying the optimization of probabilistic time, if the cycle that regulates the intersection has been modified, if you have entered into modification of deterministic phases or any other incident.

b) Order actions that are executed by other intersections, such as requesting the participation and execution of a selection algorithm or a distributed coordination algorithm or ordering a cycle change. 10

In addition, the communication protocol also includes confirmation messages for receiving messages, as well as forwarding.

Operating Example 15

The present invention is further illustrated by the application example shown in FIG. 2, which is not intended to be limiting of its scope. Suppose an urban artery composed of two intersections (intersection 1 and intersection 2) where N indicates north, S indicates south, E indicates east and O indicates west. Each intersection starts with cycle times 20 and equal phases, for example cycle-time = 60 seconds, which is divided into green-time = 30 seconds, red-time = 20 seconds, yellow-time = 5 seconds, empty time = 5 seconds (the cycle time must be equal to the sum of the times of the different states).

 25

For each intersection, the range that defines whether the system is in probabilistic or deterministic mode is established. The range must be established based on urban transport policy. For example, if we only looked at the number of vehicles waiting, the lower limit of the range could be 50% of the vehicles that the traffic light can serve when they are green and the upper limit 80% of that value. 30

If we lower these percentages, the algorithm reacts before, that is, it will begin to increase the green times before. On the contrary, if they are very high, many vehicles in queue will be necessary for the method to start. By also taking into account the number of travelers, we could set those limits based on the density of travelers 35 on the road or even set those limits in terms of minimum and maximum pollution or

a weighted average of passenger density and pollution.

Both intersections with East-West green phase time are shown in FIG.2a, while both intersections in the emptying time are shown in FIG.2b, that is, at the time when the state of each intersection corresponding to 5 the address that will then turn green (ie the NS address).

The system monitors traffic data in real time and based on these calculates a value called <<state>>. The frequency with which the state is obtained can be, for example, every time the intersection is at a time of emptying, as in Fig. 2b. A state value is obtained for the traffic direction of an intersection whose traffic lights are paired and to which it corresponds then to enter the green time phase. For example, suppose that in the figure the east-west traffic lights are green, then yellow and then enter the emptying time. After the emptying time, the north-south traffic lights will turn green and the east-west will turn red. During the emptying time, a state value is calculated for the north-south direction of intersection 1 and another state value for intersection 2. In the example shown in FIG.2b the state of intersection 1 NS: Intersection_ State1_NS = weight1 * (travelers_detected_N + travelers_detected_S) + weight2 * (vehicles_in_waiting_N + vehicles_in_waiting_S). This state is boxed in Figure 2b.

Thus, for example, weights could be assigned weight1 = 0.2 and weight2 = 0.8, therefore giving priority to reducing pollution caused by vehicles. If we opt for the opposite scenario, weight1 = 0.8 and weight2 = 0.2, we would be giving priority to the travel time 25 of the occupants of the vehicles, this being useful for example on roads defined as traveling to workplaces.

Each state is compared with its corresponding predefined range. If a state is within the range then it will probabilistically apply an increase in green time, and 30 the green time of the north-south traffic lights will go from 30 seconds to 35 seconds with a probability p. Note that an increase in the north-south green phase also means that the east-west red phase also increases the same amount of time since the cycle remains constant in this example. If the state is out of range, in excess, then the increase is made deterministically. If the status is outside the range, below the start value, the phase in green is not modified. Supposing

an intersection detects a consecutive number of green phase increments in the same direction, the cycle will be increased, which is communicated to the adjacent intersection / s so that they also modify it. In the same way in the event that an intersection detects a consecutive number of non-increments of the green phase in the same direction, the cycle will be reduced. 5

Claims (6)

1. A method for the control of self-regulated traffic lights, executed in a system comprising a first data acquisition subsystem (100); 5 a second subsystem of optimization of phases and cycle of traffic lights (200) that regulate an intersection between two land transport routes; Y a third wireless communication subsystem (300) between intersections; wherein said method is characterized in that it comprises: to. a first data acquisition process in the first subsystem (100) which in turn comprises: i. a stage of obtaining statistics on the actual traffic of a land transport route; Y ii. a stage for estimating the number of occupants of a vehicle; b. a second process of optimization of phases and cycle of traffic lights in the second second subsystem (200), which in turn includes the stages of: i. detect congestion based on the data collected by the first data acquisition subsystem (100); ii. modify the intervals and / or intersection traffic light cycle in a probabilistic or deterministic manner depending on the degree of congestion; Y C. a third process of wireless communication between intersections of an urban area by means of the third subsystem (300) which includes, in turn, the stages of: i. allow mobile signal capture that will then be used by the data acquisition subsystem (100); Y ii. allow the sending and receiving of messages between intersections or with a control center. 2. The method of claim 1 wherein the step of obtaining statistics on the actual traffic of vehicles on a road comprises an image processing stage based on detection of visual events. 3. The method according to any of claims 1-2 wherein the step of estimating the number of vehicle occupants comprises a stage of counting 35 people based on the signal of active mobile devices captured by the communication subsystem (300). 4. The method according to any of claims 1-3 wherein at the stage of modification of the intervals and / or the cycle of the traffic lights of the intersection two values are determined which determine when the (re) configuration of the traffic light is 5 probabilistic, within the range determined by these values and when it is deterministic, outside that range. 5. The method according to any of the steps 1-4 where the wireless communication process comprises a communication protocol between 10 intersections consisting of several messages whose mission is: a) to report the state of each intersection and b) order actions to be executed by other intersections. 6. A system for the control of self-regulated traffic lights, executed in a system 15 comprising: a) a first data acquisition subsystem (100) comprising, in turn: to. a first module (101) configured to obtain the statistics on the actual traffic of vehicles on a road and comprising image processing means based on visual event detection; Y b. a second module (102) configured to estimate the number of vehicle occupants and comprising means of counting people based on the signal of active mobile devices captured by a communication subsystem (300); 25 b) a second phase and cycle optimization subsystem of traffic lights (200) that regulate an intersection between two land transport routes; Y c) a third wireless communication subsystem (300) between intersections; wherein said system is characterized in that it comprises means for executing the method according to any one of claims 1 to 5.