Computer-implemented method for minimizing negative externalities in road traffic network segments and associated system, computer programs and reading medium

Abstract

A computer-implemented method is disclosed comprising data collection steps, by means of environmental sensors (20) and traffic sensors (21) relating to a traffic segment (10), to identify vulnerabilities and negative externalities by first training a model for applying adaptive behaviors. The method includes sending recommendations to connected automated vehicles (CAVs) (60) circulating in an area (11) upstream of the segment (10), with the adaptive behaviors that minimize negative externalities, e.g., pollution or noise for pedestrians (80) or critical buildings (A), considering the non-autonomous conventional vehicles (CVs) (70) present on the road. The previously trained model continues to be trained during the application of the invention on a segment (10). The system that implements the method comprises computational modules for control and evaluation (30) and decision, selection and transmission (40), which through transmission units (50) launch recommendations to the CAVs (60) with adaptive behaviors.

Description

TECHNICAL FIELD

[0001] The subject matter of the present application falls within the scope of computer-implemented inventions. More specifically, this application relates to a computer-implemented method and a system for minimizing negative externalities in road traffic segments.BACKGROUND

[0002] In 2019, around 307,000 people died prematurely in the European union (EU) due to exposure to air pollution. It is almost universally agreed that with the introduction of connected vehicles and automated driving systems (CAV), an improvement in road safety can be expected. However, academic studies show that, although still uncertain, CAVs can have a beneficial environmental performance on the traffic network and optimize the performance of other vehicles if they are induced to follow a specific operational behavior [1].

[0003] On the other hand, by assuming a standard response behavior independent of the characteristics of the road infrastructure and traffic conditions, they can contribute, in certain situations, to a serious deterioration in traffic performance, safety, congestion, pollutant emissions and air quality [1].

[0004] Optimizing the performance of CAVs can be done using driving algorithms incorporated into the vehicle itself [2]. However, the objective of minimizing the negative externalities of a traffic flow in which CAVs and conventional vehicles (CVs) coexist is more complex than a scenario consisting only of CAVs or only CVs. In summary, the following difficulties can be identified: 1) The contribution of CAVs to traffic flow performance and consequent impacts varies depending on their market penetration rate and the level of technology achieved; 2) The environmental impact of the traffic flow varies according to the composition of the mixed fleet, i.e. percentage of CAVs and percentage of CVs, in particular the type of propulsion technology and percentage of vehicles with pollution levels belonging to the various pollutant emission standards (Euro 1, Euro 2, ... Euro 6); 3) Minimizing the different externalities of road traffic can involve different strategies. For example, minimizing gases such as nitrogen oxides (NOx) may imply smoother acceleration, while minimizing greenhouse gases such as carbon dioxide (CO 2 ) may imply greater acceleration and more aggressive behavior in order to reduce congestion and travel time [3]; 4) Even assuming high computational power, the real-time or near real-time optimization of the driving behavior of a CAV or set of CAVs to minimize the negative externalities of a mixed traffic flow, in a set of network segments for scenarios of varying demand levels and fleet compositions implies highly complex calculations and long computation times. This determines the suitability and feasibility of the optimization process itself.

[0005] In the case of negative externalities related to pollution or air quality, these also vary according to atmospheric conditions

[0006] On days of high atmospheric stability and low wind speed, for example, air pollutants tend to accumulate in a critical zone adjacent to the road, promoting a degradation of air quality. On rainy and windy days, air pollution tends to disperse. On the other hand, air quality depends on the daily behavior of traffic (through its emissions), and it is by acting on this behavior that the present invention is intended to reduce negative externalities, as will be discussed later in this application. For example, during peak traffic hours there is significant congestion resulting in high emission values, which in turn translate into equally high pollutant concentrations and therefore, more pronounced negative externalities. As another example, during the day there are moderate levels of congestion and air pollution, but higher noise levels due to the increased speed of vehicles. In these periods, there is still some margin for vehicles to slightly alter their driving behavior. At night, with less traffic flow, there is a greater degree of freedom for vehicles to change their driving profile. In addition, the relative contribution of a given class of vehicle to air pollution is also dynamic. During the day, heavy passenger vehicles make a large contribution to pollutant emissions, but at night it is light vehicles that contribute almost exclusively to negative externalities.PRIOR ART

[0007] US patent no. US 11,360,236 B1, by Prathamesh Khedekar, Downers Grove, IL (USA), published on June 14, 2022, discloses a system comprising a plurality of autonomous units within a geographical region, each configured with a set of sensors and a cognitive emissions and air pollutant mapping module that allows them to map the surrounding environment and detect and overlay pollutant and emissions data on the aforementioned map, including cameras and object detection algorithms to track and photograph pollutant sources. Each unit securely transmits the merged map and pollutant source data to one or more servers that compile a complete 3D map of the geographic area overlaid with pollution data that is updated in real time, and also notifies relevant third parties of the action of pollutant sources within the area.

[0008] Although the above-mentioned state-of-the-art document identifies and monitors polluting agents on a road segment, it does not provide information on how to solve the problem of acting in real time on real traffic, namely on the autonomous vehicles circulating on the road, while also considering non-autonomous vehicles, in order to reduce the negative externalities that can be air pollution, noise pollution, congestion or the safety of pedestrians and other road users.

[0009] Chinese patent application no. CN111405522, by Suzhou Kunpeng Intelligent Network Tech Co Ltd, published on July 10, 2020, discloses an online simulation system and device of V2X testing based on vehicle-road cooperation, and a method comprises the following steps: 1, building a real V2X test environment and verifying communication between vehicles, between vehicles and infrastructure, between vehicles and pedestrians; 2, testing a real communication environment signal and monitoring the main condition of a communication signal in a test area in real time; 3, testing V2X signal shielding in a tunnel and analyzing the tunnel communication test environment of a closed test area; and 4, building a V2X communication environment to test and verify the system capability and signal processing mechanism of a networked intelligent automobile in various communication environments. According to the V2X test online simulation system and device based on vehicle-road cooperation, the vehicle connected to the intelligent network is accelerated to go on the road, and traffic safety and efficiency are improved. The system and device are important in reducing emissions.

[0010] Although the above document discloses a system and a method for simulating communications between vehicles and between vehicles and infrastructures with a view to reducing emissions and also safety, it does not solve the technical problem of not knowing traffic patterns on certain road segments, nor does it simultaneously solve the problems of safety, air pollution and noise, in accordance with applicable regulations, and it also does not consider non-autonomous vehicles in its assessment.

[0011] Thus, there is currently no sufficiently complete technological solution capable of efficiently minimizing, through the use of adaptive behaviors recommended for autonomous vehicles circulating in a traffic segment, the negative externalities identified in that traffic segment, taking into account traffic profiles and priorities.

[0012] The need therefore remains for a solution to minimize negative externalities in road traffic segments.SUMMARY OF THE INVENTION

[0013] The present invention relates to a computer-implemented method for information management for connected and automated driving vehicles (CAVs), wherein the method comprises steps for collecting and processing information regarding a given road traffic segment, in particular regarding negative externalities such as air pollution, noise, congestion and / or safety, in order to identify and prioritize the critical problems in a given road traffic segment at different time periods and to provide the CAVs with instructions comprising adaptive behaviors aimed at minimizing externalities such as air pollution, noise, congestion and / or safety, noise, congestion and / or safety, in order to identify and prioritize the critical problems on a given road traffic segment in different time periods and to provide CAVs with instructions comprising adaptive behaviors aimed at minimizing negative externalities identified on the traffic segment under consideration. The adaptive behaviors are determined based on the characteristics of the traffic present on an upstream section of the specific road traffic segment.

[0014] The present invention also relates to a system that executes the aforementioned computer-implemented method. The technical solution provided by the present invention provides support for automated driving by efficiently and quickly instructing CAVs to contribute to minimizing an externality or set of negative externalities associated with traffic flow, with the necessary information requirements being the level of demand, the composition of fleets, and the negative externalities to be minimized in a given period and in a given area.

[0015] In a first aspect, the present invention relates to a computer-implemented method for minimizing negative externalities in road traffic segments, wherein the method comprises the following four main steps: (a) determination of vulnerabilities in a delimited road traffic segment, (b) determination of traffic profiles, negative externalities and priorities in the aforementioned delimited road traffic segment, (c) definition of scenarios that relate the different traffic profiles determined in step (b) with adaptive behaviors for the delimited road traffic segment through initial training of a model, and (d) recognition of traffic profile patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model.

[0016] In a second aspect, the present invention relates to a system configured to implement the computer-implemented method according to the first aspect of the invention.

[0017] In a third and fourth aspect, the present invention relates to computer programs comprising logical instructions capable of enabling computational units comprised in the system according to the second aspect of the present invention to execute the steps of the method according to the first aspect of the invention.

[0018] In a fifth aspect, the present invention relates to a means of reading by a computer apparatus, comprising the installation of computer programs according to any of the third or fourth aspects of the invention.SOLUTION TO THE PROBLEM

[0019] Connected and autonomous vehicles (CAVs) are already a reality and it is expected that in the coming decades there will be a growing number of these vehicles circulating in urban environments with built-in autonomous driving systems. At the most advanced levels of autonomy, it is assumed that human intervention in the act of driving will become negligible or only necessary to act in emergency situations. At the same time, it is expected that a considerable percentage of conventional vehicles (CVs) powered by internal combustion engines will continue to circulate on road networks. These play a key role in generating air and noise pollution, as well as contributing to the increased risk of other negative externalities such as road accidents, most of which are due to the human factor, in particular the adoption of inappropriate driving behavior. Given the specific characteristics of each urban area, the degree of vulnerability to the negative externalities caused by road traffic will vary.

[0020] For example, consider an urban artery with several educational and residential establishments. Given this profile with different uses throughout the day, there are also different critical problems associated with the negative externalities caused by road traffic. In the morning (rush hour) one of the critical problems will be safety and pedestrians and children crossing the road. During the day, the most critical problem concerns exposure to air pollutants and noise from children in the playground. From 6pm onwards, the most critical problem will no longer be associated with local negative externalities because there is no longer a vulnerable population exposed, but with global negative externalities such as climate change or noise at night. The main benefits of this invention are improvements in terms of social and environmental impacts and energy consumption. The present invention assumes that centralized and automated management of vehicle operating behavior will be more effective than human action. This invention is based on the possibility for CAVs to change their behavior taking into account detailed information regarding traffic conditions and critical vulnerabilities in a given period of time for a given area of the network. The determination of critical factors can be done manually by the infrastructure manager (e.g. traffic management center) or in an automated way.

[0021] Thus, the present invention solves the problems of the prior art through a technological solution that makes it possible to analyze traffic conditions in real time, identify situations of critical vulnerability, optimize the driving behavior of CAVs based on a broad set of simulation and impact data, and provide operational indications to CAVs so that they are active agents in promoting more sustainable road flow.

[0022] For the present invention, it is assumed that a CAV is an autonomous vehicle with an automation level higher than 3 according to the SAE (Society of Automotive Engineers).ADVANTAGEOUS EFFECTS OF THE INVENTION

[0023] The method and system according to the present invention allow a prior definition of vulnerabilities, negative externalities and priorities, through the priorities and respective effective adaptive behavior for reducing negative externalities, making it possible, in an operational phase, to immediately associate an appropriate adaptive behavior. This is done through real-time recognition of automated adaptive driving behavior in order to minimize the negative externalities identified. Notwithstanding this advantage, the method and system of the invention also allow continuous learning of the model that is applied, using deep learning mechanisms, and possibly ensemble methods based on bagging, boosting and random forests for problems of classification and categorization of typologies of behavior.

[0024] Taking into account the forecasts for the penetration of connected and autonomous vehicles on the market, it is expected that from 2030 onwards, in certain European cities, there will be a significant percentage of these vehicles so that the present invention will be an advantage for the cities that implement it, as well as being economically advantageous. European regulations and strategies such as the zero pollution action plan, in line with the European Commission's road safety policy 2021-2030, the Stockholm Declaration and the Safe System approach, are guidelines against which the incorporation of active safety systems with a holistic and integrative vision of the main externalities of road traffic (safety, noise, air quality and greenhouse gases) is highly desirable.BRIEF DESCRIPTION OF THE FIGURES

[0025] In order to promote an understanding of the principles according to the embodiments of the present invention, reference will be made to the embodiments illustrated in the figures and the language used to describe them. In any case, it should be understood that there is no intention to limit the scope of the present invention to the content of the figures. Any subsequent changes or modifications to the inventive features illustrated herein, as well as any additional applications of the principles and embodiments of the invention illustrated, which would normally occur to an expert in the art being in possession of this description, are considered within the scope of the claimed invention. Figure 1 - illustrates a diagram of the four main steps of the method according to the invention; Figures 1a, 1b, 1c and 1d - illustrate flowcharts of an embodiment of the sub-steps of each main step a), b), c) and d), respectively, of the method according to the invention; Figure 2 - illustrates an embodiment of the system according to the invention applied to a road traffic segment; Figure 3 - graph illustrating an example of the relative importance of priorities of negative externalities to be minimized with the method and system according to the present invention; Figure 4 - schematically illustrates the communication systems of an autonomous vehicle comprised in embodiments of the system according to the present invention; Figure 5 - illustrates an embodiment of the system according to the invention applied in a road traffic segment, wherein the system is issuing adaptive behavior recommendations to autonomous vehicles; Figure 6 - illustrates an embodiment of the system according to the invention applied in a road traffic segment, wherein the system is issuing recommendations of other adaptive behaviors to autonomous vehicles; Figure 7 - illustrates an embodiment of the system according to the invention applied to a road traffic segment, wherein the system is issuing recommendations of yet other adaptive behaviors to autonomous vehicles. DESCRIPTION OF EMBODIMENTS

[0026] The technological solution of the present invention relates to a computer-implemented method for minimizing the negative externalities present in a given road traffic segment and also relates to a system comprising, among others, sensors and computational units that execute the method according to the invention.

[0027] The computer-implemented method according to the invention comprises four main steps: a) determining vulnerabilities in a defined road traffic segment; b) determining traffic profiles, negative externalities and priorities in the road traffic segment; c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile -patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model.

[0028] Figure 1 illustrates a diagram of the four main steps of the method according to the invention. In step a) of determining vulnerabilities in a delimited road traffic segment, environmental sensors installed in the aforementioned road traffic segment first collect data relating to environmental parameters, such as air pollution data, e.g. emissions of CO 2 and / or NOx, particulate matter (PM), noise levels and / or relating to the presence of people or users in the aforementioned delimited road traffic segment or in areas adjacent to traffic circulation, whereby these first data will make it possible to determine the vulnerabilities existing in the aforementioned traffic segment. In addition, other data on the vulnerabilities associated with the considered traffic segment can be retrieved from a database or entered manually by a system operator or manager (not shown). Figure 1a shows a flowchart of the sub-steps of the main step a). In step a1 shown in figure 1a, the first data mentioned above is collected by the environmental sensors. In sub-step a2, these first data are sent by the environmental sensors to a control and evaluation computer module. This data is sent using means of communication known in the art, either wired or wireless. For example, the data sent by the environmental sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known in the art. In sub-step a3, the control and evaluation computer module checks a database 1 relating to the weather conditions present on the delimited traffic segment and / or at the time of data collection in sub-step a1, wherein the weather conditions may include at least one of the group consisting of relative humidity, local atmospheric pressure, wind speed, temperature, precipitation, among others. Then, in sub-step a4, the control and evaluation computational module assigns maximum or minimum threshold values to parameters related to the first data collected in sub-step a1, wherein the maximum or minimum threshold values can be established, for example, by law, by standards, or by another entity, for the parameters associated with the first data collected by the environmental sensors. The control and assessment computer module then compares, in sub-step a5, the first data collected with the threshold values and, based on this comparison, determines, in sub-step a6, the presence of vulnerabilities relating to the parameters wherein the first data is more or less close to, or even exceeds, the corresponding threshold values and quantifies these vulnerabilities, according to the range of the difference between the first data collected and the threshold values. Optionally, in sub-step a6, the control and assessment computational module also considers other data associated with vulnerabilities which it retrieves from a database 1 within the control and assessment computational module's non-volatile memory and / or, alternatively or in addition, receives, via input from a network operator or manager, other data relating to vulnerabilities through a user interface of the control and assessment computational module. In sub-step a7, the control and assessment computer module keeps the vulnerabilities obtained in sub-step a6 and their quantification in the aforementioned database 1.

[0029] Referring again to figure 1, in step b) of determining traffic profiles, negative externalities and priorities in the aforementioned delimited road traffic segment, second data is collected by means of traffic sensors installed in an area upstream of the aforementioned delimited road traffic segment, the second data being related to second parameters concerning the traffic circulating in the aforementioned area upstream of the road traffic segment. This data can include, for example, the number of vehicles with a level of automation equal to or greater than 3 that circulate per hour of the day or per day of the week, the number of conventional vehicles that circulate per hour of the day or per day of the week, the composition of the fleet, i.e. the relative percentage of the presence of connected and / or autonomous vehicles (CAV) and conventional vehicles (CV). Figure 1b illustrates one way of carrying out the sub-steps included in step b) of the method according to the invention. In sub-step b1, a second dataset is collected by traffic sensors installed at suitable points upstream of the delimited traffic segment in question. In sub-step b2, this second set of data is sent to the control and evaluation computer module. This is done by means of some communication protocol or type, either wired or wireless. For example, the data sent by the traffic sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known as available. In sub-step b3, the control and evaluation computer module performs the traffic profiles modeling and groups them by type of traffic profile, by at least one of, without limitation, vehicle type, percentage of heavy vehicles, percentage of autonomous vehicles, by date / time, by day of the week, by day of the month, by weather conditions, for example. Traffic profile (TP) modeling includes the clustering of sets of TPs and their negative externalities (without imposing adaptive behavior) and are generated according to the number of vehicles, fleet composition and other external variables such as weather conditions. Each TP group is simulated and validated experimentally. An example of a traffic profile might be as follows: traffic flow - 2500 vehicles per hour (vph), 25% heavy vehicles, 15% CAVs, average speed 20 km / h, carbon dioxide emissions (CO 2 ) - 500 kg / h, nitrogen oxides (NOx) 300 g / h, particulate matter (PM) 20 g / h.

[0030] Next, in sub-step b4, the control and evaluation computer module calculates the negative externalities, taking into account the traffic profiles and vulnerabilities and their quantification obtained in step a). Based on the negative externalities resulting from the calculations in the previous sub-step, priorities are determined in sub-step b5 with a view to reducing or minimizing the negative externalities in the traffic segment in question. Finally, in sub-step b6, the results obtained in sub-steps b3, b4 and b5 are stored in a database 1 housed in the aforementioned non-volatile memory of the control and evaluation computer module. Below, with reference to figure 3, an example of the results of the negative externalities prioritization calculation carried out in sub-step b5 will be discussed.

[0031] In step c) according to the method of the invention, as further illustrated in figure 1, the control and evaluation computer module defines scenarios for the different traffic profiles determined in step b) and taking into account the negative externalities and priorities calculated in step b). To do this, the control and evaluation computer module runs a microsimulation to train a model before applying it to the traffic segment in question. To train this model, the input data includes: adaptive behaviors obtained from database 2, wherein adaptive behaviors can be at least one of longitudinal control of a CAV, lateral control of a CAV, wherein longitudinal control of a CAV can comprise decreasing the speed or increasing the speed at which the CAV drives, or a combination of these, e.g. a set of adaptive behaviors can be: use of left lane for forward movement at traffic circle + distance to vehicle in front reduced by 12 %, acceleration between 10 and 20 km / h increased by 3 %); negative externalities calculated in step b); traffic profiles obtained in step b); priorities calculated in step b); wherein the training levels include the application of one or more adaptive behaviors for each set of input data; and wherein the output data are scenarios that relate traffic profiles and priorities of negative externalities with the most appropriate adaptive behaviors to send to the CAVs in a real situation. Specifically, figure 1c illustrates a flowchart of the sub-steps that step c) comprises according to the method of the invention. In sub-step c1, input data for the model to be trained is read from database 1, i.e. negative externalities, traffic profiles and priorities of the negative externalities calculated in step b). In sub-step c2, adaptive behaviors are read from database 2, wherein the adaptive behaviors can be at least one of the group consisting of: longitudinal control of an autonomous vehicle, lateral control of an autonomous vehicle, wherein the longitudinal control of an autonomous vehicle can include decreasing the speed or increasing the speed at which the autonomous vehicle travels, or a combination of these. Sub-step c3 performs the model's training level, wherein one or more adaptive behaviors are selected and their application to a set of input data is simulated, in order to obtain, in sub-step c4, scenarios that relate the input data to one or more appropriate adaptive behaviors. Finally, in sub-step c5, the results, i.e. the scenarios obtained in the previous sub-step, are stored in database 2.

[0032] According to certain embodiments of the method of the invention, step c) can be performed by the control and evaluation computer module. According to other embodiments of the invention, step c) can be performed by a decision, selection and transmission computational module. According to yet other embodiments of the method of the invention, step c) can be performed by a computational unit comprising both the control and evaluation and the decision, selection and transmission computational modules. Both the control and evaluation computational module and the decision, selection and transmission computational module comprise processing means suitable for executing supervised learning algorithms or sets of logical and sequential instructions for classification problems and / or comprise processing means suitable for executing unsupervised learning algorithms or sets of logical and sequential instructions for clustering problems.

[0033] In step d) according to the method of the invention, as further illustrated in figure 1, the model trained in step c) is applied in real time. Step d) is based on a communication network including a device in the at least one autonomous vehicle or CAV, a radio communication network between the decision, selection and transmission module and the transmission unit and the CAV vehicle to transmit the best adaptive behavior in a given time frame depending on the vulnerabilities, traffic profiles and negative externalities to be minimized. In particular, in step d), the decision, selection and transmission module transmits a recommendation with one or more adaptive behaviors selected according to the scenarios learned in step c), to one or more connected and autonomous vehicles circulating in real time in the traffic segment or in an area upstream of it. Subsequently, new first data collected in real time by the environmental sensors is compared with previous first data and the decision, selection and transmission module concludes on the reduction of at least one of the negative externalities identified in step b), wherein the sequence of these sub-steps is repeated a plurality of times in a continuous training of the initially trained model. The reduction of at least one of the negative externalities is considered a positive impact.

[0034] According to embodiments of the invention, step d) is carried out a plurality of times, for example, more than 10 times, more than 20 times or more than 50 times, and step d) is carried out in a processing cycle lasting from 1 second to 60 minutes, preferably from 3 seconds to 30 minutes and more preferably the processing cycle lasts approximately 15 minutes.

[0035] Figure 1d illustrates a flowchart of the sub-steps comprised in step d) according to the invention. Sub-step d1 involves the collection of second data in real time by the traffic sensors, second data which is sent in sub-step d2 to the decision, selection and transmission module. This is done by some means of communication protocols or types, either wired or wireless. For example, the data sent by the traffic sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known in the art. The second traffic data includes the identification of the connected and / or autonomous vehicles. This identification can be done, for example, 1) by means of a video camera that identifies the vehicle's license plate or 2) by means of wireless communication technologies, such as the vehicle-to-vehicle communication system (V2V) and the vehicle-to-infrastructure communication system (V2I), 3) the information is transmitted whenever a vehicle passes and if it is a CAV it receives the information, otherwise it does not, the information is ignored, and in this way it identifies a conventional vehicle (CV). Next, in sub-step d3, the decision, selection and transmission module retrieves scenarios stored in database 2, after which it compares, in sub-step d4, the second real-time traffic data with the traffic profiles in the scenarios retrieved from database 2 and retrieves a set of adaptive behaviors from the same database. In sub-step d5, the decision, selection and transmission module decides which at least one adaptive behavior best suits the second real-time traffic data collected in sub-step d1 and sends, in sub-step d6, a recommendation with the at least one adaptive behavior to a transmission unit. This transmission is done through some communication means, with or without wires. For example, the recommendation with at least one adaptive behavior can be sent to the transmission unit via the Internet, Wi-Fi, Bluetooth or any other means known to the technique. In sub-step d7, the transmission unit transmits a recommendation with at least one adaptive behavior to at least one connected or autonomous vehicle that is driving in real time on the traffic segment or in an area upstream of it. The transmission of this recommendation is done by some wireless communication means. For example, this recommendation can be transmitted via the Internet, or via Wi-Fi, Bluetooth or any other means known in the prior art. For example, the communication can be made by means of the aforementioned Vehicle-to-Vehicle Communication System (V2V), Vehicle-to-Infrastructure Communication System (V2I), such as Dedicated Short-Range Communications (DSRC) and Cellular Vehicle-to-Everything (C-V2X). Sub-step d7 may include a sub-step d71 (not shown in figure 1d) to obtain confirmation that the autonomous vehicle (60) has carried out the recommendation sent in sub-step d7). This sub-step d71) can obtain confirmation that the recommendation has been carried out in one or more of the following ways: collection of third data g n+1 , by the traffic sensors (21) with subsequent transmission, by the traffic sensors (21), of the third data g n+1 collected, to the computational decision, selection and transmission module (40); transmission, by one or more autonomous vehicles (60), of an acceptance signal of the recommendation to the transmission unit (50), wherein subsequently the transmission unit (50) sends the aforementioned one or more acceptance signals to the computational decision, selection, and transmission module (40); storage, in a memory of the one or more connected or autonomous vehicles (60), of the acceptance of the recommendation received in sub-step d7. The aforementioned collection of third data g n+1 , by the traffic sensors (21) is carried out after the recommendation has been issued, wherein the third data includes data, for example, images of the CAV (60) that received the recommendation in order to confirm or not the change in behavior on the part of that CAV, namely through video camera, photographic cameras, for example. The conclusion of the evaluation of these data, whether or not it includes the CAV (60) following up on the recommendation, will be taken into account in the processing of continuous learning.

[0036] Sub-step d71) is followed by a sub-step d72) of processing of the aforementioned third data and / or of the one or more acceptance signals of the recommendation received in sub-step d71), wherein said processing of third data includes a comparative evaluation with third data collected in previous time instants, in order to make it possible to interpret the first data received in real time more accurately, since, for example, the failure to reduce negative externalities may not be due to inadequate adaptive behavior recommended to one or more CAVs (60), but rather to the fact that at least one CAV (60) did not follow the recommended adaptive behavior. Sub-step d8 comprises the collection of the first data by the environmental sensors, wherein the first data is collected in real time, after which, in sub-step d9, the environmental sensors send the real-time first data to the decision, selection and transmission computational module. This data is sent using some means of communication protocols or type, either wired and wireless. For example, the data sent by the environmental sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known in the art. Next, the decision, selection and transmission computational module, in sub-step d10, performs continuous training, in real time, of the model initially trained in step c). To do this, the decision, selection and transmission computer module uses the following input data: first data received in real time first data received previously, compares the first data received in real time from the environmental sensors with the values of the first data previously received and stored in memory, deciding on the effectiveness of the at least one adaptive behavior recommended in sub-step d7 in minimizing, or reducing, at least one negative externality identified in step b) and, if it concludes that the adaptive behavior was not effective for at least one negative externality associated with the first data collected in real time, it adjusts the adaptive behavior in the at least one scenario used, wherein the adjustment comprises a change in the correspondence between negative externalities and adaptive behaviors, thus continuing to train the model. The results of sub-step d10, i.e. at least one scenario trained or retrained, or adjusted, in sub-step d10, are stored in database 2, in sub-step d11. Step d) is carried out continuously, a plurality of times, preferably more than 10 times, more than 20 times and more preferably more than 50 times, thus obtaining continuous training of the artificial intelligence model, namely deep learning and ensemble methods based on bagging, "boosting" and "random forests" for classification problems for the minimization of negative externalities in a delimited road traffic segment wherein the method of the invention is applied.

[0037] Referring now to the remaining figures, potential approaches and technicalities of realizing the system that implements the method according to the invention will be described, and the components of the aforementioned system will be discussed.

[0038] Figure 2 illustrates the application of the method and system according to one embodiment of the present invention. Figure 2 illustrates a delimited traffic segment (10) considered critical, i.e. typically an urban road section with specific problems or vulnerabilities for which the aim is to minimize problems related to, for example, air pollution, noise, safety, congestion. Next to the traffic segment (10) there may be pedestrian crossings and / or buildings (A) to which vulnerabilities can be associated, such as a school, kindergarten or hospital. For example, in an area wherein there is a hospital or a school, it may be necessary to comply with pollution and noise requirements, and maximum permissible limits are set for that area.

[0039] Points representing pedestrians (80) are shown crossing the pedestrian crossing or walking alongside the segment (10). As can be seen in figure 2, the traffic segment (10) is preceded by an upstream area or zone (11) and, according to the present invention, both the segment (10) and the area (11) are subject to monitoring and action, aimed at reducing or minimizing negative externalities associated with the vulnerabilities of the segment (10). Installed in the vicinity of the traffic segment (10), or even in the traffic segment itself, are environmental sensors (20). Environmental sensors (20) can include, but are not limited to, noise sensors, air pollution sensors, image capture devices, people counting sensors and / or vehicle counting sensors, video cameras, photographic cameras, pollutant gas concentration analyzers, sound level meters, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, sensing coils, thermometers, precipitation sensors, air humidity sensors, wind speed sensors and combinations thereof, for example, wherein the air pollution sensors may be, for example, sensors for the concentration of certain gases, such as NOx, CO 2 , carbon monoxide (CO) or other gases potentially harmful to health and / or the environment. The set of sensors (20) can also include meteorological data sensors, such as temperature, precipitation, wind speed, air humidity. There may also be image sensors (20), or video cameras, installed on the road traffic segment (10), which collect safety-related data, for example, they collect information on the flow of pedestrians (80) on the roadway of the segment (10) or in adjacent areas. In certain embodiments of the present invention, safety-related data can be obtained from existing databases (not shown) with information on accident history and / or density variability of pedestrians, cyclists or other vulnerable users. In the upstream area or zone (11) of the segment (10), as illustrated in figure 2, traffic sensors (21) are installed to monitor traffic and obtain second data, i.e. data relating to, without limitation, the percentage of autonomous and connected vehicles (CAV), the percentage of conventional vehicles (CV), the percentage of heavy vehicles, by time of day, by day of the week, etc. These traffic sensors (21) can be at least one of the group consisting of video cameras, photographic cameras, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, detection coils and combinations of these. Figure 2 also illustrates the presence of autonomous vehicles (CAVs) (60) and conventional vehicles (70), wherein the CAVs (60) are equipped with means of bidirectional vehicle-to-infrastructure communication, or V2I / I2V, as known to a skilled person in the art. Figure 2 also shows the control and evaluation computer module (30) and the decision, selection and transmission computer module (40) according to the present invention. These two computational modules (30,40) can belong to different computational units or to the same computational unit (not shown). Preferably, the computational modules (30,40) belong to the same computational unit. Taking into account that the analysis and evaluation of the current, predicted and controlled state based on informed decision control relies on the processing and analysis of large volumes of data (sometimes from different sources), at least one of the computing modules (30,40) may be an artificial intelligence (AI) computing device, with dedicated, high-performance resources for higher-level planning. At least the decision, selection and transmission module (40) can be equipped with graphics processors from video cards, since these show potential for accelerating and optimizing calculations needed in deep learning, and ensemble methods based on bagging, boosting and random forests for classification problems. Computational performance is greatly increased by the possibility of running multiple, simultaneous calculations, allowing training processes to be distributed and significantly speeding up learning operations. In this way, it is possible to instantly (in 1-3 seconds) provide at least one CAV with a suitable adaptive behavior recommendation considering the upstream (11) or approach zone, as well as the traffic segment (10).

[0040] The control and evaluation computer module (30) comprises a non-volatile memory (not shown) in which database 1 is housed. Database 1 contains first data captured by the environmental sensors (20) and second data captured by the traffic sensors (21), i.e. database 1 contains data related to negative externalities (air pollution, e.g., NOx and PM concentration, CO 2 concentration, noise), traffic profile data (fleet volume and composition) and priorities of the negative externalities calculated in step b). The decision, selection and transmission computational module (40) comprises a non-volatile memory (not shown) in which a database 2 is housed. Both computational modules (30) and (40) involve communication features configured to receive data sent by the sensors (20,21). When the modules (30,40) are not part of the same computing unit, they are both equipped with means to communicate with each other. These means of communication can be wired or wireless and can be of a type known to an expert in the field, for example, they can communicate via any OTA (over-the-air) communication system, such as wireless WAN networks or another wireless network. The decision, selection and transmission computer module (40) is further equipped with wired or wireless communication means for communicating with transmission units (50), discussed below. The system according to the invention includes communication means comprising a device in the CAV (60), a radio communication network (not shown) between the decision, selection and transmission module (40) and a transmission unit (50) and the CAV (60) to transmit the best adaptive behavior to achieve a positive impact, i.e. reduce at least one negative externality.

[0041] As further illustrated in figure 2, there are transmission units (50) configured to communicate with the CAVs (60) by issuing recommendations with adaptive behaviors according to the present invention. The transmission units (50) are installed in the vicinity of the segment (10) and / or the upstream zone (11) and comprise communication means configured to receive recommendations / communication from a decision, selection and transmission computational module (40) and transmit the aforementioned recommendations to at least one CAV (60) circulating in the vicinity, in order to cause a change in behavior on the part of the aforementioned at least one CAV (60) and thereby contribute to reducing at least one negative externality. In general terms, figure 2 illustrates the system implementing the computer-implemented method according to the invention, namely steps a), b) and c) described above. In particular, figure 2 illustrates the circulation of CAVs (60) and CVs (70) in an upstream zone (11) of the traffic segment (10) and on the segment itself (10). Environmental sensors (20) collect data associated with the traffic segment (10), or first data, and send it to the control and evaluation computer module (30). Traffic sensors (21) also collect data associated with the composition of traffic, or second data, and send it to the control and evaluation computer module (30). The full arrows drawn on the sidewalk of the traffic segment (10) indicate the adaptive behaviors that can be received by the CAVs in a recommendation sent by the decision, selection and transmission computational module (40) and transmitted by the transmission unit (50). These adaptive behaviors can consist of a recommendation to increase speed, to reduce speed, to change lanes, or to change direction.

[0042] Figure 3 shows a graph illustrating an example of the relative importance of priorities of negative externalities to be minimized with the method and system according to the present invention. In particular, according to the method of the invention and as described above, in step b) priorities are identified according to the negative externalities and vulnerabilities detected in step a) of the method according to the invention. The methodology employed for this purpose can be based on the use of time series of data collected by environmental and weather sensors (air quality, traffic, noise, temperature, humidity, wind, precipitation, safety, exposed population, congestion) in order to group typologies of negative externalities according to the level of priority. Figure 3 graphically exemplifies how such negative externalities can be prioritized and grouped over 24 hours. Referring to figure 3, the following periods over the course of a day were grouped together: I1 - night rest period - this period has a low density of pedestrians and is dominated by the prioritization of minimizing noise (E2) and greenhouse gases (E3); I2 - period of high concentration of pedestrians and young children crossing the road - during this period top priority is given to the safety and protection of pedestrians (E7); I9 - period with moderate pedestrian concentration and air pollution below the limits - in this period there is an even distribution of the various negative externalities to be minimized; I4 - period in which the concentration of ozone (E4) approaches the limits for human health - in this period there is an increase in the level of importance of air quality and emissions of NOx (E7) and PM (E6); I7 - late afternoon rush hour period - a period when priority is given to relieving congestion (E1) and ensuring the safety of pedestrians (E8).

[0043] Step b) of determining traffic profiles (TP), negative externalities and priorities on the aforementioned delimited road traffic segment (10) comprises monitoring the traffic composition throughout the day, namely the flow (vehicles per hour) upstream of the segment to be optimized, and the category of vehicles in circulation and the weather conditions. Counting can be carried out using the traffic sensors (21) described above, wherein vehicles are counted and their category identified (e.g. CAV, CV, heavy vehicles, etc.). The data is collected, sent to the control and evaluation module (30) and stored in a database 1. Traffic monitoring can be carried out in an upstream area (11) of the traffic segment (10), allowing the necessary time for the identification of the traffic profile, selection of negative externalities to be minimized, choice of adaptive behavior and transmission via means of communication between infrastructures, or the decision, selection and transmission module (40) and the CAV vehicles, according to step d) of the method of the invention.Example of a set of data collected by the system according to the invention

[0044] Table 1 shows an example of the structure of the data collected by the environmental sensors (20) and traffic sensors (21) in an embodiment according to the invention, wherein this data is stored in database 1, as already described. Table 1 Structure of variables to be monitored and stored in the database Traffic Profile 8h33Precipitation 0 mm / hTemperature 8°CWind 34 km / hHumidity 85%Ice NoLevel of automation of the vehicle fleetL0L1L2L3L4L520%28%22%11%10%9%DieselPetrolBEVHEVPHEV20%10%30%30%10%CategoryLightHeavyMotorcycleBicycles70%15%10%5%

[0045] Wherein BEV refers to "Battery Electric Vehicle", HEV refers to "Hybrid Electric Vehicle", PHEV refers to "Plug-in Hybrid Electric Vehicle", i.e. a hybrid electric vehicle that can be recharged from an external source.

[0046] As mentioned above, in step c) of defining scenarios for the different traffic profiles and priorities determined in step b) for the delimited road traffic segment (10), the traffic impacts for different levels of vehicle composition and different adaptive behaviors for the CAVs are estimated using the longitudinal and lateral control systems (see figure 4).

[0047] Figure 4 illustrates a CAV (60) according to the invention, wherein the means of communication that a CAV typically has are indicated and are as follows: LIDAR system (1), which is a system that makes it possible to measure the distances to various objects in the vicinity of the CAV (60) and also makes it possible to create three-dimensional images of the detected objects, mapping the environment around the CAV (60); RADAR system (2) consisting of a set of short- and long-range radar sensors installed around the vehicle for assistance, distance maintenance control and braking assistance; three-dimensional image cameras (3) which automatically detect objects (other vehicles, pedestrians, cyclists, road signs, road markings, bridges, among others), classify them, and determine the distances between them and the vehicle; satellite navigation system (GNSS) (4), which allows geographical location with numerical coordinates (e.g. latitude, longitude) representing their physical location in space. They can also navigate by combining GPS coordinates in real time with other digital mapping data (e.g. via Google Maps); and communication module with the transmission unit (5), i.e. a vehicle-to-infrastructure communication unit (V2I), which is a control unit for receiving instructions wirelessly via the transmission unit (50) or communication protocol (V2X) on adaptive behavior and incorporating the recommendations sent by the transmission unit (50) into the CAV's planning and decision algorithms (60).

[0048] The longitudinal and lateral control systems in CAVs ensure safe movement in a lane. Longitudinal control, maintained by the cruise controls, controls vehicle speeds using the brakes or accelerator and maintains a safe distance between two vehicles. Cooperative adaptive cruise control (CACC) systems integrate dedicated short-range communication (DSRC) technologies and / or 5G technology to establish vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication. It should be noted that GNSS technology can have decimal-level accuracy to keep CAVs in their lane and maintain a safe distance from other vehicles. The longitudinal control strategy in CAVs keeps the distance between two vehicles and the speed stabilized. In various situations, this strategy can be adjusted slightly to achieve a certain objective from the network manager's point of view. For example, in situations of moderate traffic, the speed can be optimized to influence the speed of other traffic and minimize the emission of noise or a given pollutant. In situations of severe congestion, the distance between two vehicles can be slightly reduced to increase traffic flow capacity on the critical segment of the network. In terms of lateral behavior, it can be advantageous to use traffic lanes to increase the capacity of the roadway from the start or to better manage the distribution of pollutants emitted in traffic segments (10). Figure 4 shows the communication systems of a CAV that could potentially be used in the present invention. Figure 4 also illustrates a CV (70) circulating on the track, wherein the CAV (60) detects the distance to the CV (70) and can control this distance by means of adaptive behavior recommendations received from the transmission unit (50). As an example, the recommendations received by one or more CAVs (60) from the transmission unit (50) can include adaptive behaviors such as adjusting the speed at free speed, so that the average speed of the group of vehicles circulating, and taking into account their composition and atmospheric pollutants, can minimize a critical pollutant, according to the priorities established as described above. In addition, CAVs can act as moderating agents by promoting traffic calming actions to minimize the risk of accidents or collisions during critical periods, also in accordance with the priorities established as described above.

[0049] Figure 5 illustrates a recommendation sent to the CAVs (60), comprising adaptive behavior to adjust the optimum speed at free regime (38 km / h) to minimize CO 2 emissions and noise in an integrated manner. Figure 5 also illustrates part of the traffic segment (10) next to a critical building (A), wherein CAV (60) and CV (70) circulate, wherein the CAVs (60) receive recommendations from the transmission unit (50), the result of processing the data sent by the sensors (20,21), as described above, to the control and evaluation (30) and decision, selection and transmission (40) computational modules, as also described above, in which case the process of identifying the optimum speed to respond to a set of priorities (I1) (see figure 3) and an upstream traffic profile (TP2) is exemplified. This information is transmitted to the CAVs (60) in an area (11) upstream of the traffic segment (10). In the case illustrated in figure 5, the adaptive behavior selected by the decision, selection and transmission computational module is the one that has the greatest influence on reducing CO 2 emissions and noise, taking into account the traffic profile. For example, in figure 5 it can be seen that the CAV (60) that is still circulating outside the traffic segment (10) is circulating at a speed of 53 km / h, still without system intervention according to the invention, while the CAV (60) that is circulating inside the traffic segment (10) is already circulating at the speed recommended by the transmission unit (50), the optimum speed calculated by the decision, selection and transmission module (40), 38 km / h. The graph shown in figure 5 indicates the optimum speed of 38 km / h for observing the priorities detected for the traffic profile in question and taking into account the negative externalities present, as shown in figure 3.

[0050] Table 2 exemplifies a consolidated data structure facilitating the process of selecting adaptive behaviors to send to one or more CAVs. Table 2 - Example of a data structure for selecting adaptive behaviors Traffic ProfileAdaptive behaviorNoiseNOxCO 2 Accident riskDelayTPAB1, AB2- 4 dB4%3.16%0+2%AB1- 4 dB12%9.48%0+15%AB2- 8 dB-3%-2.37%03%AB3+ 4 dB-4%-3.16%4%2%ABN- 3 dB7%5.53%-20%1%..................

[0051] Wherein ABN is a hypothetical adaptive behavior of order N that reduces traffic flow noise, for example, by 3 dB, as shown in Table 2.

[0052] The adaptive behavior of order N, ABN, exemplified in Table 2, also reduces the risk of an accident by 20%, although it increases the emission of NOx and CO 2 which, for the scenario exemplified, will not be such negative externalities as noise or the risk of an accident.

[0053] For example, at moderate speeds, vehicles may tend to move more evenly, which can reduce turbulence and noise emissions. Any of these changes in speed has an influence on travel time and therefore delay.

[0054] Figure 6 represents yet another case in which the CAVs (60-A, 60-B) circulating in the vicinity of segment (10), through the application of the method and system according to the invention, receive recommendations (M1, M2) to change lanes in order to reduce the emission of pollutants, taking into account the pedestrians (80) present on segment (10). In this case, the CAV (60-A) is powered by electricity and is therefore considerably less polluting and receives the recommendation (M1) to change lanes to the left, while the CAV (60-B), which is powered by internal combustion and emits a higher level of pollution, receives the recommendation (M2) to change lanes to the right. In this way, the level of pollution to which pedestrians (80) are exposed is reduced.

[0055] Figure 7 illustrates another case of application of the method and system of the present invention. In figure 7, the priority to be observed is to relieve congestion (I4, as shown in figure 3). In this case, the decision, selection and transmission module has calculated an optimum distance of 0.7 m to help relieve traffic congestion. The recommendation given to the CAVs (60) present is to maintain this distance from the vehicle in front of them. Vehicle density (k), representing the number of vehicles per kilometer of track, can be optimized using parameters such as the average distance between vehicles when stationary in a queue or the increase in the minimum following distance due to vehicles moving. Figure 7 exemplifies the transmission of an adaptive behavior on the optimized distance to the vehicle in front, taking into account a specific traffic profile (TP3) to minimize a negative externality with priority, i.e. in this case congestion. Traffic flow (q) represents the number of vehicles per hour. In the fundamental traffic diagram shown in figure 7, as traffic density (k) increases, traffic flow (q) also increases until it reaches a peak or maximum value. From that point on, traffic flow (q) begins to decrease as traffic density (k) continues to increase. This happens because, at very high densities, vehicles start to move more slowly, reducing the flow (volume of cars passing through a section).

[0056] In a first aspect, the invention relates to a computer-implemented method for minimizing negative externalities in road traffic segments comprising the following steps: a) determination of vulnerabilities in a defined road traffic segment (10); b) determination of traffic profiles (TP), negative externalities and priorities in the aforementioned delimited road traffic segment (10); c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment (10), taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model; wherein step a) of determination of vulnerabilities in a delimited road traffic segment (10) comprises the following sub-steps: a1) collection of first data dn, by means of a set of environmental sensors (20) installed on or near the aforementioned delimited road traffic segment (10), the first data dn being related to first parameters of the delimited road traffic segment (10); a2) transmission of the first data collected in sub-step a1), by each of the environmental sensors (20), to a control and evaluation computer module (30), via wired or wireless means of communication; a3) consultation, by the control and evaluation computer module (30), of a weather condition database; a4) assignment of threshold values by the control and evaluation computer module (30), which can be maximum or minimum threshold values xn for the first data dn collected in sub-step a1); a5) comparison, by the control and evaluation computer module (30), of each of the first data dn collected in sub-step a1), with the respective threshold values xn assigned in sub-step a4), for each of the parameters of the delimited traffic segment (10); a6) determination, by the control and evaluation computer module (30), of the presence of vm vulnerabilities, relative to the parameters wherein the first data dn and quantification of these vm vulnerabilities, according to the range of the difference between the first data dn and the threshold xn values and considering the meteorological conditions obtained in sub-step a3); a7) storage, by the control and assessment computer module (30), of the vm vulnerabilities and their quantification, i.e. the results obtained in sub-step a6) in a database 1 housed in a non-volatile computer memory of the aforementioned control and assessment computer module (30); wherein, optionally, in sub-step a6), the control and assessment computer module (30) also considers other data associated with vulnerabilities retrieved from the aforementioned database 1 and / or receives, via input by a network operator or manager, other data relating to vulnerabilities through a user interface of the control and assessment computer module (30). wherein step b) of determination of traffic profiles, negative externalities and priorities in the aforementioned segment of road traffic delimited (10) comprises the following sub-steps: b1) collection of second data f n , by means of a set of traffic sensors (21) installed in an area (11) upstream of the aforementioned delimited road traffic segment (10), the second data f n being relative to the composition of the vehicles circulating in the area (11); b2) sending the second data collected in sub-step b1), by each of the traffic sensors (21) to a control and evaluation computer module (30), via wired or wireless means of communication; b3) based on the second data collected in step b1), the control and evaluation computer module (30) draws up a model of the traffic profiles, wherein the modeling includes grouping the data by at least one of vehicle types, percentage of heavy vehicles, percentage of connected, autonomous vehicles, by date / time, by day of the week, by day of the month, by weather conditions; b4) based on the traffic profile modeling carried out in the previous sub-step, and considering the vulnerabilities and their quantification calculated in step a), the control and evaluation computer module (30) calculates the negative externalities associated with the traffic segment (10); b5) prioritization, by the control and evaluation computer module (30), of the negative externalities associated with the traffic segment (10); (b6) storage by the control and evaluation computer module (30) of the results of sub-steps (b3), (b4) and (b5) in a database 1 housed in the computer memory of the aforementioned control and evaluation computer module (30); wherein step c) of definition of scenarios relates the traffic profiles determined in step b) with adaptive behavior for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model and comprises the following sub-steps carried out by the control and evaluation computer module (30): c1) reading input data to train a model, starting from database 1, wherein the input data comprises: negative externalities, traffic profiles, and / or priorities of the negative externalities calculated in step b); c2) reading of adaptive behaviors ABp, from database 2, wherein the aforementioned adaptive behaviors ABp are adequate to minimize negative externalities and wherein the adaptive behaviors ABp are at least one of the group constituted by: longitudinal control of an autonomous vehicle (60), lateral control of an autonomous vehicle (60), wherein the longitudinal control of an autonomous vehicle (60) may comprise decreasing the speed or increasing the speed at which the autonomous vehicle (60) travels, or a combination of these instructions. c3) selection of one or more adaptive ABp behaviors and simulation of their application to a specific set of input data read in sub-steps c1) and c2); c4) determining ci scenarios that relate a specific set of input data to one or more adaptive ABp behaviors, resulting from the training of the model carried out in sub-step c3); c5) storing the results of sub-step c4) in a database 2 housed in the computer memory of the aforementioned control and evaluation computer module (30); wherein step d) of recognition of traffic profile patterns, identifies and communicates adaptive behaviors, and performs continuous training of the initially trained model, step d) comprising the following sub-steps: d1) collection of second data f n+1 , by means of the set of traffic sensors (21) installed in the zone (11) upstream of the aforementioned delimited road traffic segment (10), the second data f n+1 being relative to the composition of the vehicles circulating in the zone (11), and including information on which autonomous vehicles (60) are connected; d2) sending the second data f n+1 collected in sub-step d1), by each of the traffic sensors (21) to a decision, selection and transmission computer module (40), via some wired or wireless means of communication; d3) retrieval, by the decision, selection and transmission computer module (40), of scenarios stored in database 2; d4) comparison, by the decision, selection and transmission computational module (40), of the second data f n+1 with the traffic profiles of the scenarios retrieved in d3), and retrieval of the adaptive behaviors stored in database 2; d5) by recognizing traffic profiles defined by the second data f n+1 , deciding on at least one adaptive behavior to be applied to at least one connected autonomous vehicle circulating in the upstream zone (11) of the traffic segment (10) or in the traffic segment itself (10); d6) sending at least one adaptive behavior to be applied, by wired or wireless communication means from the decision, selection and transmission module (40), to one or more transmission units (50); d7) sending a recommendation with at least one adaptive behavior to be applied, by one or more transmission units (50), via wireless communication means, to at least one connected autonomous vehicle (60) circulating in the area (11) upstream or in the traffic segment (10) identified in the second data f n+1 collected in sub-step d1); d8) collection of the first data d n+1 , by means of the environmental sensors (20) installed on or near the delimited road traffic segment (10), the first data d n+1 relating to the first real-time parameters of the delimited road traffic segment (10) and consultation, by the decision, selection and transmission computer module (40), of the aforementioned weather conditions database; d9) sending the first data collected in sub-step d8), by each of the environmental sensors (20), to the decision, selection and transmission computer module (40), via wired or wireless means of communication; d10) the decision, selection and transmission computer module (40), based on the following input data: first data received in real time d n+1 , first data received previously d n , makes a comparison between these two sets of first data d n+1 and d n , wherein, if this difference reveals a reduction of at least one of the negative externalities identified in step b), the module concludes on the effectiveness of the application of at least one adaptive behavior sent in sub-step d7) and wherein, if this difference reveals a non-reduction of at least one of the negative externalities identified in step b), the module concludes on the ineffectiveness of the application of at least one adaptive behavior sent in sub-step d7) and inserts an adjustment in at least one of the scenarios determined in step c), wherein the adjustment comprises a change in the correspondence between negative externalities and adaptive behaviors, resulting in at least one adjusted scenario; d11) storing the results of sub-step d10, i.e. at least one scenario considered effective and / or at least one adjusted scenario; wherein step d) is performed n times, wherein n is greater than or equal to one, preferably n is greater than 50, and wherein step d) is performed in a processing cycle lasting from 1 second to 60 minutes.

[0057] In an embodiment of the first aspect of the invention, in sub-step a1) of collecting first data, the aforementioned first parameters are at least one of the group consisting of pedestrian concentration rate, vehicle concentration rate, gas concentration rate, noise, safety level, congestion or a combination of these.

[0058] In another embodiment of the first aspect of the invention, in sub-step a4), the assignment of threshold values is made for different periods of time, for example, the assignment of threshold values is made by day of the week, by hour of the day of the week, by season, or by a combination of these.

[0059] In another embodiment of the first aspect of the invention, in sub-step b5), the determination of priorities is based on the proximity between the first data d n and the corresponding threshold values, period of the day, week, month, type of population living nearby, type of buildings in the vicinity of the delimited traffic segment (10).

[0060] In another embodiment of the first aspect of the invention, sub-step d7) comprises the following sub-steps: d71) obtaining confirmation that the autonomous vehicle (60) has carried out the recommendation sent in sub-step d7) in at least one of the ways belonging to the group consisting of: collection of third data g n+1 , by the traffic sensors (21) with subsequent sending, by the traffic sensors (21), of the third data g n+1 collected, to the computational decision, selection and transmission module (40); the one or more autonomous vehicles (60) send a recommendation acceptance signal to the transmission unit (50), whereupon the transmission unit (50) sends these one or more acceptance signals to the decision, selection and transmission computational module (40); storing, in a memory of the one or more autonomous vehicles (60), the acceptance of the recommendation received in sub-step d7); d72) processing, by the decision, selection and transmission computational module (40), of the third data g n+1 and / or of the one or more recommendation acceptance signals received in sub-step d71), wherein the aforementioned processing of the third data g n+1 includes a comparative evaluation with other third data g n collected at previous time instants; wherein the third data g n , g n+1 includes data from the one or more autonomous vehicles (60) that received one or more recommendations; and in the comparison step d10), the third data g n+1 and / or the one or more recommendation acceptance signals sent by the one or more autonomous vehicles (60) are additionally included in the input data.

[0061] In another embodiment of the first aspect of the invention, in sub-step d10), the effectiveness of the at least one adaptive behavior applied is concluded on the basis of at least one of: the number of people present, the length of time people are present, the decrease in exposure to polluting gases such as NOx, PM, CO 2 , the decrease in noise, the increase in safety, the decrease in congestion.

[0062] In another embodiment of the first aspect of the invention, in sub-step c) of definition of scenarios for the delimited road traffic segment (10), the definition of scenarios is carried out for different periods of time, for example, the definition of scenarios is carried out by day of the week, by hour of the day of the week, by season, or by a combination of these.

[0063] In another embodiment of the first aspect of the invention, in the sending recommendation sub-steps a2), b2), d2), d6), d9) the transmission of instructions is carried out by wired or wireless means of communication; and in the sending sub-step d7) the transmission is carried out by wireless means of communication.

[0064] In another embodiment of the first aspect of the invention, step c) is optionally carried out by the decision, selection and transmission module (40).

[0065] In a second aspect, the present invention relates to a system configured to implement the computer-implemented method as defined above in relation to the first aspect of the invention, wherein the system comprises: a set of environmental sensors (20) installed on and / or near a delimited road traffic segment (10), wherein the set of environmental sensors (20) comprises at least one of the group consisting of video cameras, photographic cameras, pollutant gas concentration analyzers, sound level meters, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, sensing coils, thermometers, precipitation sensors, air humidity sensors, wind speed sensors, and combinations thereof; a set of traffic sensors (21) installed in an upstream area (11) of the aforementioned delimited road traffic segment (10), wherein the set of traffic sensors (21) comprises at least one of the group consisting of video cameras, photographic cameras, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, detection coils and combinations of these; a control and evaluation computing module (30), wherein the computing module (30) comprises control and evaluation computing module communication means, control and evaluation computing module database storage means, and control and evaluation computing module processing means; a decision, selection and transmission computing module (40), wherein a decision, selection and transmission computing module (40) involves decision, selection and transmission computing module communication means, decision, selection and transmission computing module database storage means and decision, selection and transmission computing module processing means; at least one CAV (60) which is configured to drive partially or fully self-controlled on the delimited road traffic segment (10), in its vicinity or in the upstream area (11) of the aforementioned traffic segment (10) and comprises communication means suitable for receiving recommendations from a transmission unit (50); at least one transmission unit (50) configured to communicate with the at least one CAV (60); optionally, at least one conventional vehicle CV (70) which is configured to circulate with a driver and without being partially or totally self-controlled in the delimited road traffic segment (10), in its vicinity or in the upstream area (11) of the aforementioned traffic segment (10); wherein the control and evaluation computational module (30) and the decision, selection and transmission computational module (40) are included in the same computational unit or in different computational units.

[0066] In an embodiment of the second aspect of the invention, the aforementioned control and evaluation computer module (30) and the aforementioned decision, selection and transmission computer module (40) are selected from the group consisting of a local computer, a remote computer, a server, a tablet, a smart phone, such as a smartphone.

[0067] In a third aspect, the present invention relates to a computer program comprising logical instructions to enable the control and evaluation computer module (30), comprised in the system as defined in relation to the second aspect of the present invention, to carry out the steps of the method as defined in relation to the first aspect of the present invention.

[0068] In a fourth aspect, the present invention relates to a computer program comprising logical instructions to enable the decision, selection and transmission computer module (40), comprised in the system as defined with regard to the second aspect of the present invention, to carry out the steps of the method as defined with regard to the first aspect of the present invention.

[0069] In a fifth aspect, the present invention relates to a means of reading by a computer apparatus comprising the installation of the computer program as defined in any of the third or fourth aspects of the present invention.DEFINITIONS

[0070] As used in this application, the term "delimited traffic segment" or "delimited road traffic segment" refers to a section of roadway and adjacent areas in which vulnerabilities exist, or potentially exist, and to which the present invention is applicable.

[0071] As used in this request, the term "vulnerability" refers to a characteristic associated with the site, i.e. the delimited traffic segment , and is related to human activity, namely the socio-demographic profile of the population, the occupation of the public road and adjacent areas, and the degree of exposure to negative externalities associated with traffic, for example, and without limitation, noise, air pollution, safety, congestion. In this application, vulnerability is measured in terms of proximity or even exceeding established or desirable thresholds for characteristics associated with the location, such as noise levels, concentration levels of certain harmful gases, speed associated with the safety of those present in the traffic segment, among others.

[0072] As used in this request, the term "negative externality" refers to a negative effect that traffic has on the vulnerabilities of the delimited road segment, i.e. it refers to the worsening of the measured values associated with one or more vulnerabilities identified in a specific road traffic segment, for example, approaching or exceeding established or desirable thresholds.

[0073] As used in this request, the term "connected autonomous vehicle or CAV" refers to motor vehicles that have the means for at least partial autonomous driving and fall within one of the levels of automation greater than or equal to 3 according to the SAE (Society of Automotive Engineers), i.e. vehicles with a degree of autonomy of level SAE3, SAE4 or SAE5, with or without a driver, and which incorporate a planning module capable of interpreting information received from an I2V / V2I transmission unit (infrastructure-to-vehicle / vehicle-to-infrastructure), adapting speed, lane choice, acceleration or deceleration rate, distance to the vehicle in front. The planning module also takes into account the mechanical, dynamic and kinetic characteristics of the vehicle.

[0074] As used in this request, the term "conventional vehicle or CV " refers to vehicles with a degree of autonomy less than or equal to SAE2 level, with a driver.

[0075] As used in this application, the term "adaptive behavior" refers to a behavior that is recommended to an autonomous vehicle by changing the expected behavior that autonomous vehicle would have, in order to contribute to causing a positive impact or impact, i.e. decreasing one or more negative externalities in the traffic segment where it is, or will be, circulating. Adaptive behavior can be an increase in speed, a decrease in speed, a change of lane, among others.

[0076] As used in this request, the term "positive impact" or "impact" refers to the positive effect that adaptive behaviors have on one or more vulnerabilities, translating into a reduction in negative externalities, i.e. a departure from the established or desirable thresholds for a specific delimited traffic segment.

[0077] As used in this request, the term "traffic profile" or "TP" refers to a set of aggregated data that characterizes the type of vehicles circulating in the delimited traffic segment, by hour of the day, by day of the week, by type of vehicle, among others. Traffic data and its negative externalities are grouped using a measure of (dis)similarity between objects, requiring the definition of the number of groups (k = 1, 2, 3, ..., n).

[0078] As used in this application, the term "deep learning" refers to a training process belonging to the concept of "machine learning", in which e.g. artificial neural networks, i.e. algorithms or sets of logical instructions modeled to function like the human brain, learn from large amounts of data.

[0079] As used in this description, the expressions "about" and "approximately" refer to a range of values of plus or minus 10% the specified number.

[0080] As used throughout this patent application, the term "or" is used in the inclusive sense rather than the exclusive sense, unless the exclusive sense is clearly defined in a specific situation. In this context, a phrase such as "X uses A or B" should be interpreted as including all relevant inclusive combinations, for example "X uses A", "X uses B" and "X uses A and B".

[0081] As used throughout this patent application, the indefinite articles "one" or "a" should generally be interpreted as "one or more" and "one or more", unless the meaning of a singular embodiment is clearly defined in a specific situation.

[0082] As presented in this description, terms related to examples should be interpreted with the purpose of illustrating an example of something and not indicating a preference.

[0083] The subject-matter described above is provided as an illustration of the present invention and should not be construed to limit it. Terminology used for the purpose of describing specific embodiments in accordance with the present invention should not be construed to limit the invention. As used in the description, the definite and indefinite articles, in their singular form, are intended to be interpreted as also including the plural forms, unless the context of the description explicitly indicates otherwise. It will be understood that the terms "comprise" and "include", when used in this description, specify the presence of the related features, elements, components, steps and operations, but do not exclude the possibility that other features, elements, components, steps and operations are also covered.

[0084] All changes, provided they do not modify the essential characteristics of the following claims, are to be considered within the scope of protection of the present invention.LIST OF REFERENCE SIGNS

[0085] a1 - collecting first data a2 - sending the first data a3 - consulting the weather database a4 - assigning threshold values a5 - comparison of the first data with the threshold values a6 - determining vulnerabilities and quantifying them a7 - storing the results of a6 b1 - collecting second data b2 - sending the second data b3 - traffic profile modeling and clustering b4 - calculating negative externalities b5 - determining priorities b6 - storing the results obtained in b4 c1 - reading priorities and traffic profiles c2 - reading adaptive behaviors (AB) c3 - selecting a AB suitable for the scenario and simulating the application of the AB in the scenario c4 - determining the relationship between scenarios and appropriate ABs to be applied in the trained scenarios c5 - storing the trained model in database 2 d1 - collecting second data d2 - sending the second data d3 - scenario recovery d4 - comparing real-time traffic profiles with AB scenarios and recovery d5 - decision on AB to be applied d6 - sending the AB d7 - transmission from AB to CAV d8 - first data Dn+1 collection d9 - sending the first data collected d10 - comparing and deciding which AB to keep or change d11 - storing the new scenarios learned 1 - LIDAR system in a CAV 2 - RADAR system in a CAV 3 - Three-dimensional imaging cameras in a CAV 4 - GNSS satellite navigation system in a CAV 5 - communication module with a transmission unit 10 - delimited road traffic segment 11- area or zone upstream of the delimited road traffic segment 20 - environmental sensors 21 - traffic sensors 30 - control and evaluation computer module 40 - decision, selection and transmission computer module 50 - transmission unit 60 - connected and / or autonomous vehicle or CAV 70 - conventional vehicle or CV 80 - pedestrians LIST OF CITATIONS

[0086] Here is the list of citations:PATENT LITERATURE

[0087] US Patent No. US 11,360,236 B1, by Prathamesh Khedekar, Downers Grove, IL (USA), published on June 14, 2022; Chinese patent application No. CN111405522, by Suzhou Kunpeng Intelligent Network Tech Co Ltd, published on July 10, 2020. NON-PATENT LITERATURE

[0088] [1] Bandeira, Jorge M., Eloisa Macedo, Paulo Fernandes, Monica Rodrigues, Mario Andrade, and Margarida C. Coelho. 2021. "Potential Pollutant Emission Effects of Connected and Automated Vehicles in a Mixed Traffic Flow Context for Different Road Types." IEEE Open Journal of Intelligent Transportation Systems 2: 364-83. [2] Mohebifard, Rasool, and Ali Hajbabaie. 2020. "Effects of Automated Vehicles on Traffic Operations at Roundabouts." In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), 1-6. IEEE. [3] Rafael, Sandra, Paulo Fernandes, Diogo Lopes, Micael Rebelo, Jorge Bandeira, Eloisa Macedo, Mónica Rodrigues, Margarida C. Coelho, Carlos Borrego, and Ana I. Miranda. 2022. "How Can the Built Environment Affect the Impact of Autonomous Vehicles' Operational Behavior on Air Quality?" Journal of Environmental Management 315 (August): 115154.

Examples

Embodiment Construction

[0026]The technological solution of the present invention relates to a computer-implemented method for minimizing the negative externalities present in a given road traffic segment and also relates to a system comprising, among others, sensors and computational units that execute the method according to the invention.

[0027]The computer-implemented method according to the invention comprises four main steps:

a) determining vulnerabilities in a defined road traffic segment; b) determining traffic profiles, negative externalities and priorities in the road traffic segment; c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile -patterns, identification and communication of adaptive behaviors, and continued training of the in...

TECHNICAL FIELD

[0001] The subject matter of the present application falls within the scope of computer-implemented inventions. More specifically, this application relates to a computer-implemented method and a system for minimizing negative externalities in road traffic segments.BACKGROUND

[0002] In 2019, around 307,000 people died prematurely in the European union (EU) due to exposure to air pollution. It is almost universally agreed that with the introduction of connected vehicles and automated driving systems (CAV), an improvement in road safety can be expected. However, academic studies show that, although still uncertain, CAVs can have a beneficial environmental performance on the traffic network and optimize the performance of other vehicles if they are induced to follow a specific operational behavior [1].

[0003] On the other hand, by assuming a standard response behavior independent of the characteristics of the road infrastructure and traffic conditions, they can contribute, in certain situations, to a serious deterioration in traffic performance, safety, congestion, pollutant emissions and air quality [1].

[0004] Optimizing the performance of CAVs can be done using driving algorithms incorporated into the vehicle itself [2]. However, the objective of minimizing the negative externalities of a traffic flow in which CAVs and conventional vehicles (CVs) coexist is more complex than a scenario consisting only of CAVs or only CVs. In summary, the following difficulties can be identified: 1) The contribution of CAVs to traffic flow performance and consequent impacts varies depending on their market penetration rate and the level of technology achieved; 2) The environmental impact of the traffic flow varies according to the composition of the mixed fleet, i.e. percentage of CAVs and percentage of CVs, in particular the type of propulsion technology and percentage of vehicles with pollution levels belonging to the various pollutant emission standards (Euro 1, Euro 2, ... Euro 6); 3) Minimizing the different externalities of road traffic can involve different strategies. For example, minimizing gases such as nitrogen oxides (NOx) may imply smoother acceleration, while minimizing greenhouse gases such as carbon dioxide (CO 2 ) may imply greater acceleration and more aggressive behavior in order to reduce congestion and travel time [3]; 4) Even assuming high computational power, the real-time or near real-time optimization of the driving behavior of a CAV or set of CAVs to minimize the negative externalities of a mixed traffic flow, in a set of network segments for scenarios of varying demand levels and fleet compositions implies highly complex calculations and long computation times. This determines the suitability and feasibility of the optimization process itself.

[0005] In the case of negative externalities related to pollution or air quality, these also vary according to atmospheric conditions

[0006] On days of high atmospheric stability and low wind speed, for example, air pollutants tend to accumulate in a critical zone adjacent to the road, promoting a degradation of air quality. On rainy and windy days, air pollution tends to disperse. On the other hand, air quality depends on the daily behavior of traffic (through its emissions), and it is by acting on this behavior that the present invention is intended to reduce negative externalities, as will be discussed later in this application. For example, during peak traffic hours there is significant congestion resulting in high emission values, which in turn translate into equally high pollutant concentrations and therefore, more pronounced negative externalities. As another example, during the day there are moderate levels of congestion and air pollution, but higher noise levels due to the increased speed of vehicles. In these periods, there is still some margin for vehicles to slightly alter their driving behavior. At night, with less traffic flow, there is a greater degree of freedom for vehicles to change their driving profile. In addition, the relative contribution of a given class of vehicle to air pollution is also dynamic. During the day, heavy passenger vehicles make a large contribution to pollutant emissions, but at night it is light vehicles that contribute almost exclusively to negative externalities.PRIOR ART

[0007] US patent no. US 11,360,236 B1, by Prathamesh Khedekar, Downers Grove, IL (USA), published on June 14, 2022, discloses a system comprising a plurality of autonomous units within a geographical region, each configured with a set of sensors and a cognitive emissions and air pollutant mapping module that allows them to map the surrounding environment and detect and overlay pollutant and emissions data on the aforementioned map, including cameras and object detection algorithms to track and photograph pollutant sources. Each unit securely transmits the merged map and pollutant source data to one or more servers that compile a complete 3D map of the geographic area overlaid with pollution data that is updated in real time, and also notifies relevant third parties of the action of pollutant sources within the area.

[0008] Although the above-mentioned state-of-the-art document identifies and monitors polluting agents on a road segment, it does not provide information on how to solve the problem of acting in real time on real traffic, namely on the autonomous vehicles circulating on the road, while also considering non-autonomous vehicles, in order to reduce the negative externalities that can be air pollution, noise pollution, congestion or the safety of pedestrians and other road users.

[0009] Chinese patent application no. CN111405522, by Suzhou Kunpeng Intelligent Network Tech Co Ltd, published on July 10, 2020, discloses an online simulation system and device of V2X testing based on vehicle-road cooperation, and a method comprises the following steps: 1, building a real V2X test environment and verifying communication between vehicles, between vehicles and infrastructure, between vehicles and pedestrians; 2, testing a real communication environment signal and monitoring the main condition of a communication signal in a test area in real time; 3, testing V2X signal shielding in a tunnel and analyzing the tunnel communication test environment of a closed test area; and 4, building a V2X communication environment to test and verify the system capability and signal processing mechanism of a networked intelligent automobile in various communication environments. According to the V2X test online simulation system and device based on vehicle-road cooperation, the vehicle connected to the intelligent network is accelerated to go on the road, and traffic safety and efficiency are improved. The system and device are important in reducing emissions.

[0010] Although the above document discloses a system and a method for simulating communications between vehicles and between vehicles and infrastructures with a view to reducing emissions and also safety, it does not solve the technical problem of not knowing traffic patterns on certain road segments, nor does it simultaneously solve the problems of safety, air pollution and noise, in accordance with applicable regulations, and it also does not consider non-autonomous vehicles in its assessment.

[0011] Thus, there is currently no sufficiently complete technological solution capable of efficiently minimizing, through the use of adaptive behaviors recommended for autonomous vehicles circulating in a traffic segment, the negative externalities identified in that traffic segment, taking into account traffic profiles and priorities.

[0012] The need therefore remains for a solution to minimize negative externalities in road traffic segments.SUMMARY OF THE INVENTION

[0013] The present invention relates to a computer-implemented method for information management for connected and automated driving vehicles (CAVs), wherein the method comprises steps for collecting and processing information regarding a given road traffic segment, in particular regarding negative externalities such as air pollution, noise, congestion and / or safety, in order to identify and prioritize the critical problems in a given road traffic segment at different time periods and to provide the CAVs with instructions comprising adaptive behaviors aimed at minimizing externalities such as air pollution, noise, congestion and / or safety, noise, congestion and / or safety, in order to identify and prioritize the critical problems on a given road traffic segment in different time periods and to provide CAVs with instructions comprising adaptive behaviors aimed at minimizing negative externalities identified on the traffic segment under consideration. The adaptive behaviors are determined based on the characteristics of the traffic present on an upstream section of the specific road traffic segment.

[0014] The present invention also relates to a system that executes the aforementioned computer-implemented method. The technical solution provided by the present invention provides support for automated driving by efficiently and quickly instructing CAVs to contribute to minimizing an externality or set of negative externalities associated with traffic flow, with the necessary information requirements being the level of demand, the composition of fleets, and the negative externalities to be minimized in a given period and in a given area.

[0015] In a first aspect, the present invention relates to a computer-implemented method for minimizing negative externalities in road traffic segments, wherein the method comprises the following four main steps: (a) determination of vulnerabilities in a delimited road traffic segment, (b) determination of traffic profiles, negative externalities and priorities in the aforementioned delimited road traffic segment, (c) definition of scenarios that relate the different traffic profiles determined in step (b) with adaptive behaviors for the delimited road traffic segment through initial training of a model, and (d) recognition of traffic profile patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model.

[0016] In a second aspect, the present invention relates to a system configured to implement the computer-implemented method according to the first aspect of the invention.

[0017] In a third and fourth aspect, the present invention relates to computer programs comprising logical instructions capable of enabling computational units comprised in the system according to the second aspect of the present invention to execute the steps of the method according to the first aspect of the invention.

[0018] In a fifth aspect, the present invention relates to a means of reading by a computer apparatus, comprising the installation of computer programs according to any of the third or fourth aspects of the invention.SOLUTION TO THE PROBLEM

[0019] Connected and autonomous vehicles (CAVs) are already a reality and it is expected that in the coming decades there will be a growing number of these vehicles circulating in urban environments with built-in autonomous driving systems. At the most advanced levels of autonomy, it is assumed that human intervention in the act of driving will become negligible or only necessary to act in emergency situations. At the same time, it is expected that a considerable percentage of conventional vehicles (CVs) powered by internal combustion engines will continue to circulate on road networks. These play a key role in generating air and noise pollution, as well as contributing to the increased risk of other negative externalities such as road accidents, most of which are due to the human factor, in particular the adoption of inappropriate driving behavior. Given the specific characteristics of each urban area, the degree of vulnerability to the negative externalities caused by road traffic will vary.

[0020] For example, consider an urban artery with several educational and residential establishments. Given this profile with different uses throughout the day, there are also different critical problems associated with the negative externalities caused by road traffic. In the morning (rush hour) one of the critical problems will be safety and pedestrians and children crossing the road. During the day, the most critical problem concerns exposure to air pollutants and noise from children in the playground. From 6pm onwards, the most critical problem will no longer be associated with local negative externalities because there is no longer a vulnerable population exposed, but with global negative externalities such as climate change or noise at night. The main benefits of this invention are improvements in terms of social and environmental impacts and energy consumption. The present invention assumes that centralized and automated management of vehicle operating behavior will be more effective than human action. This invention is based on the possibility for CAVs to change their behavior taking into account detailed information regarding traffic conditions and critical vulnerabilities in a given period of time for a given area of the network. The determination of critical factors can be done manually by the infrastructure manager (e.g. traffic management center) or in an automated way.

[0021] Thus, the present invention solves the problems of the prior art through a technological solution that makes it possible to analyze traffic conditions in real time, identify situations of critical vulnerability, optimize the driving behavior of CAVs based on a broad set of simulation and impact data, and provide operational indications to CAVs so that they are active agents in promoting more sustainable road flow.

[0022] For the present invention, it is assumed that a CAV is an autonomous vehicle with an automation level higher than 3 according to the SAE (Society of Automotive Engineers).ADVANTAGEOUS EFFECTS OF THE INVENTION

[0023] The method and system according to the present invention allow a prior definition of vulnerabilities, negative externalities and priorities, through the priorities and respective effective adaptive behavior for reducing negative externalities, making it possible, in an operational phase, to immediately associate an appropriate adaptive behavior. This is done through real-time recognition of automated adaptive driving behavior in order to minimize the negative externalities identified. Notwithstanding this advantage, the method and system of the invention also allow continuous learning of the model that is applied, using deep learning mechanisms, and possibly ensemble methods based on bagging, boosting and random forests for problems of classification and categorization of typologies of behavior.

[0024] Taking into account the forecasts for the penetration of connected and autonomous vehicles on the market, it is expected that from 2030 onwards, in certain European cities, there will be a significant percentage of these vehicles so that the present invention will be an advantage for the cities that implement it, as well as being economically advantageous. European regulations and strategies such as the zero pollution action plan, in line with the European Commission's road safety policy 2021-2030, the Stockholm Declaration and the Safe System approach, are guidelines against which the incorporation of active safety systems with a holistic and integrative vision of the main externalities of road traffic (safety, noise, air quality and greenhouse gases) is highly desirable.BRIEF DESCRIPTION OF THE FIGURES

[0025] In order to promote an understanding of the principles according to the embodiments of the present invention, reference will be made to the embodiments illustrated in the figures and the language used to describe them. In any case, it should be understood that there is no intention to limit the scope of the present invention to the content of the figures. Any subsequent changes or modifications to the inventive features illustrated herein, as well as any additional applications of the principles and embodiments of the invention illustrated, which would normally occur to an expert in the art being in possession of this description, are considered within the scope of the claimed invention. Figure 1 - illustrates a diagram of the four main steps of the method according to the invention; Figures 1a, 1b, 1c and 1d - illustrate flowcharts of an embodiment of the sub-steps of each main step a), b), c) and d), respectively, of the method according to the invention; Figure 2 - illustrates an embodiment of the system according to the invention applied to a road traffic segment; Figure 3 - graph illustrating an example of the relative importance of priorities of negative externalities to be minimized with the method and system according to the present invention; Figure 4 - schematically illustrates the communication systems of an autonomous vehicle comprised in embodiments of the system according to the present invention; Figure 5 - illustrates an embodiment of the system according to the invention applied in a road traffic segment, wherein the system is issuing adaptive behavior recommendations to autonomous vehicles; Figure 6 - illustrates an embodiment of the system according to the invention applied in a road traffic segment, wherein the system is issuing recommendations of other adaptive behaviors to autonomous vehicles; Figure 7 - illustrates an embodiment of the system according to the invention applied to a road traffic segment, wherein the system is issuing recommendations of yet other adaptive behaviors to autonomous vehicles. DESCRIPTION OF EMBODIMENTS

[0026] The technological solution of the present invention relates to a computer-implemented method for minimizing the negative externalities present in a given road traffic segment and also relates to a system comprising, among others, sensors and computational units that execute the method according to the invention.

[0027] The computer-implemented method according to the invention comprises four main steps: a) determining vulnerabilities in a defined road traffic segment; b) determining traffic profiles, negative externalities and priorities in the road traffic segment; c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile -patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model.

[0028] Figure 1 illustrates a diagram of the four main steps of the method according to the invention. In step a) of determining vulnerabilities in a delimited road traffic segment, environmental sensors installed in the aforementioned road traffic segment first collect data relating to environmental parameters, such as air pollution data, e.g. emissions of CO 2 and / or NOx, particulate matter (PM), noise levels and / or relating to the presence of people or users in the aforementioned delimited road traffic segment or in areas adjacent to traffic circulation, whereby these first data will make it possible to determine the vulnerabilities existing in the aforementioned traffic segment. In addition, other data on the vulnerabilities associated with the considered traffic segment can be retrieved from a database or entered manually by a system operator or manager (not shown). Figure 1a shows a flowchart of the sub-steps of the main step a). In step a1 shown in figure 1a, the first data mentioned above is collected by the environmental sensors. In sub-step a2, these first data are sent by the environmental sensors to a control and evaluation computer module. This data is sent using means of communication known in the art, either wired or wireless. For example, the data sent by the environmental sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known in the art. In sub-step a3, the control and evaluation computer module checks a database 1 relating to the weather conditions present on the delimited traffic segment and / or at the time of data collection in sub-step a1, wherein the weather conditions may include at least one of the group consisting of relative humidity, local atmospheric pressure, wind speed, temperature, precipitation, among others. Then, in sub-step a4, the control and evaluation computational module assigns maximum or minimum threshold values to parameters related to the first data collected in sub-step a1, wherein the maximum or minimum threshold values can be established, for example, by law, by standards, or by another entity, for the parameters associated with the first data collected by the environmental sensors. The control and assessment computer module then compares, in sub-step a5, the first data collected with the threshold values and, based on this comparison, determines, in sub-step a6, the presence of vulnerabilities relating to the parameters wherein the first data is more or less close to, or even exceeds, the corresponding threshold values and quantifies these vulnerabilities, according to the range of the difference between the first data collected and the threshold values. Optionally, in sub-step a6, the control and assessment computational module also considers other data associated with vulnerabilities which it retrieves from a database 1 within the control and assessment computational module's non-volatile memory and / or, alternatively or in addition, receives, via input from a network operator or manager, other data relating to vulnerabilities through a user interface of the control and assessment computational module. In sub-step a7, the control and assessment computer module keeps the vulnerabilities obtained in sub-step a6 and their quantification in the aforementioned database 1.

[0029] Referring again to figure 1, in step b) of determining traffic profiles, negative externalities and priorities in the aforementioned delimited road traffic segment, second data is collected by means of traffic sensors installed in an area upstream of the aforementioned delimited road traffic segment, the second data being related to second parameters concerning the traffic circulating in the aforementioned area upstream of the road traffic segment. This data can include, for example, the number of vehicles with a level of automation equal to or greater than 3 that circulate per hour of the day or per day of the week, the number of conventional vehicles that circulate per hour of the day or per day of the week, the composition of the fleet, i.e. the relative percentage of the presence of connected and / or autonomous vehicles (CAV) and conventional vehicles (CV). Figure 1b illustrates one way of carrying out the sub-steps included in step b) of the method according to the invention. In sub-step b1, a second dataset is collected by traffic sensors installed at suitable points upstream of the delimited traffic segment in question. In sub-step b2, this second set of data is sent to the control and evaluation computer module. This is done by means of some communication protocol or type, either wired or wireless. For example, the data sent by the traffic sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known as available. In sub-step b3, the control and evaluation computer module performs the traffic profiles modeling and groups them by type of traffic profile, by at least one of, without limitation, vehicle type, percentage of heavy vehicles, percentage of autonomous vehicles, by date / time, by day of the week, by day of the month, by weather conditions, for example. Traffic profile (TP) modeling includes the clustering of sets of TPs and their negative externalities (without imposing adaptive behavior) and are generated according to the number of vehicles, fleet composition and other external variables such as weather conditions. Each TP group is simulated and validated experimentally. An example of a traffic profile might be as follows: traffic flow - 2500 vehicles per hour (vph), 25% heavy vehicles, 15% CAVs, average speed 20 km / h, carbon dioxide emissions (CO 2 ) - 500 kg / h, nitrogen oxides (NOx) 300 g / h, particulate matter (PM) 20 g / h.

[0030] Next, in sub-step b4, the control and evaluation computer module calculates the negative externalities, taking into account the traffic profiles and vulnerabilities and their quantification obtained in step a). Based on the negative externalities resulting from the calculations in the previous sub-step, priorities are determined in sub-step b5 with a view to reducing or minimizing the negative externalities in the traffic segment in question. Finally, in sub-step b6, the results obtained in sub-steps b3, b4 and b5 are stored in a database 1 housed in the aforementioned non-volatile memory of the control and evaluation computer module. Below, with reference to figure 3, an example of the results of the negative externalities prioritization calculation carried out in sub-step b5 will be discussed.

[0031] In step c) according to the method of the invention, as further illustrated in figure 1, the control and evaluation computer module defines scenarios for the different traffic profiles determined in step b) and taking into account the negative externalities and priorities calculated in step b). To do this, the control and evaluation computer module runs a microsimulation to train a model before applying it to the traffic segment in question. To train this model, the input data includes: adaptive behaviors obtained from database 2, wherein adaptive behaviors can be at least one of longitudinal control of a CAV, lateral control of a CAV, wherein longitudinal control of a CAV can comprise decreasing the speed or increasing the speed at which the CAV drives, or a combination of these, e.g. a set of adaptive behaviors can be: use of left lane for forward movement at traffic circle + distance to vehicle in front reduced by 12 %, acceleration between 10 and 20 km / h increased by 3 %); negative externalities calculated in step b); traffic profiles obtained in step b); priorities calculated in step b); wherein the training levels include the application of one or more adaptive behaviors for each set of input data; and wherein the output data are scenarios that relate traffic profiles and priorities of negative externalities with the most appropriate adaptive behaviors to send to the CAVs in a real situation. Specifically, figure 1c illustrates a flowchart of the sub-steps that step c) comprises according to the method of the invention. In sub-step c1, input data for the model to be trained is read from database 1, i.e. negative externalities, traffic profiles and priorities of the negative externalities calculated in step b). In sub-step c2, adaptive behaviors are read from database 2, wherein the adaptive behaviors can be at least one of the group consisting of: longitudinal control of an autonomous vehicle, lateral control of an autonomous vehicle, wherein the longitudinal control of an autonomous vehicle can include decreasing the speed or increasing the speed at which the autonomous vehicle travels, or a combination of these. Sub-step c3 performs the model's training level, wherein one or more adaptive behaviors are selected and their application to a set of input data is simulated, in order to obtain, in sub-step c4, scenarios that relate the input data to one or more appropriate adaptive behaviors. Finally, in sub-step c5, the results, i.e. the scenarios obtained in the previous sub-step, are stored in database 2.

[0032] According to certain embodiments of the method of the invention, step c) can be performed by the control and evaluation computer module. According to other embodiments of the invention, step c) can be performed by a decision, selection and transmission computational module. According to yet other embodiments of the method of the invention, step c) can be performed by a computational unit comprising both the control and evaluation and the decision, selection and transmission computational modules. Both the control and evaluation computational module and the decision, selection and transmission computational module comprise processing means suitable for executing supervised learning algorithms or sets of logical and sequential instructions for classification problems and / or comprise processing means suitable for executing unsupervised learning algorithms or sets of logical and sequential instructions for clustering problems.

[0033] In step d) according to the method of the invention, as further illustrated in figure 1, the model trained in step c) is applied in real time. Step d) is based on a communication network including a device in the at least one autonomous vehicle or CAV, a radio communication network between the decision, selection and transmission module and the transmission unit and the CAV vehicle to transmit the best adaptive behavior in a given time frame depending on the vulnerabilities, traffic profiles and negative externalities to be minimized. In particular, in step d), the decision, selection and transmission module transmits a recommendation with one or more adaptive behaviors selected according to the scenarios learned in step c), to one or more connected and autonomous vehicles circulating in real time in the traffic segment or in an area upstream of it. Subsequently, new first data collected in real time by the environmental sensors is compared with previous first data and the decision, selection and transmission module concludes on the reduction of at least one of the negative externalities identified in step b), wherein the sequence of these sub-steps is repeated a plurality of times in a continuous training of the initially trained model. The reduction of at least one of the negative externalities is considered a positive impact.

[0034] According to embodiments of the invention, step d) is carried out a plurality of times, for example, more than 10 times, more than 20 times or more than 50 times, and step d) is carried out in a processing cycle lasting from 1 second to 60 minutes, preferably from 3 seconds to 30 minutes and more preferably the processing cycle lasts approximately 15 minutes.

[0035] Figure 1d illustrates a flowchart of the sub-steps comprised in step d) according to the invention. Sub-step d1 involves the collection of second data in real time by the traffic sensors, second data which is sent in sub-step d2 to the decision, selection and transmission module. This is done by some means of communication protocols or types, either wired or wireless. For example, the data sent by the traffic sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known in the art. The second traffic data includes the identification of the connected and / or autonomous vehicles. This identification can be done, for example, 1) by means of a video camera that identifies the vehicle's license plate or 2) by means of wireless communication technologies, such as the vehicle-to-vehicle communication system (V2V) and the vehicle-to-infrastructure communication system (V2I), 3) the information is transmitted whenever a vehicle passes and if it is a CAV it receives the information, otherwise it does not, the information is ignored, and in this way it identifies a conventional vehicle (CV). Next, in sub-step d3, the decision, selection and transmission module retrieves scenarios stored in database 2, after which it compares, in sub-step d4, the second real-time traffic data with the traffic profiles in the scenarios retrieved from database 2 and retrieves a set of adaptive behaviors from the same database. In sub-step d5, the decision, selection and transmission module decides which at least one adaptive behavior best suits the second real-time traffic data collected in sub-step d1 and sends, in sub-step d6, a recommendation with the at least one adaptive behavior to a transmission unit. This transmission is done through some communication means, with or without wires. For example, the recommendation with at least one adaptive behavior can be sent to the transmission unit via the Internet, Wi-Fi, Bluetooth or any other means known to the technique. In sub-step d7, the transmission unit transmits a recommendation with at least one adaptive behavior to at least one connected or autonomous vehicle that is driving in real time on the traffic segment or in an area upstream of it. The transmission of this recommendation is done by some wireless communication means. For example, this recommendation can be transmitted via the Internet, or via Wi-Fi, Bluetooth or any other means known in the prior art. For example, the communication can be made by means of the aforementioned Vehicle-to-Vehicle Communication System (V2V), Vehicle-to-Infrastructure Communication System (V2I), such as Dedicated Short-Range Communications (DSRC) and Cellular Vehicle-to-Everything (C-V2X). Sub-step d7 may include a sub-step d71 (not shown in figure 1d) to obtain confirmation that the autonomous vehicle (60) has carried out the recommendation sent in sub-step d7). This sub-step d71) can obtain confirmation that the recommendation has been carried out in one or more of the following ways: collection of third data g n+1 , by the traffic sensors (21) with subsequent transmission, by the traffic sensors (21), of the third data g n+1 collected, to the computational decision, selection and transmission module (40); transmission, by one or more autonomous vehicles (60), of an acceptance signal of the recommendation to the transmission unit (50), wherein subsequently the transmission unit (50) sends the aforementioned one or more acceptance signals to the computational decision, selection, and transmission module (40); storage, in a memory of the one or more connected or autonomous vehicles (60), of the acceptance of the recommendation received in sub-step d7. The aforementioned collection of third data g n+1 , by the traffic sensors (21) is carried out after the recommendation has been issued, wherein the third data includes data, for example, images of the CAV (60) that received the recommendation in order to confirm or not the change in behavior on the part of that CAV, namely through video camera, photographic cameras, for example. The conclusion of the evaluation of these data, whether or not it includes the CAV (60) following up on the recommendation, will be taken into account in the processing of continuous learning.

[0036] Sub-step d71) is followed by a sub-step d72) of processing of the aforementioned third data and / or of the one or more acceptance signals of the recommendation received in sub-step d71), wherein said processing of third data includes a comparative evaluation with third data collected in previous time instants, in order to make it possible to interpret the first data received in real time more accurately, since, for example, the failure to reduce negative externalities may not be due to inadequate adaptive behavior recommended to one or more CAVs (60), but rather to the fact that at least one CAV (60) did not follow the recommended adaptive behavior. Sub-step d8 comprises the collection of the first data by the environmental sensors, wherein the first data is collected in real time, after which, in sub-step d9, the environmental sensors send the real-time first data to the decision, selection and transmission computational module. This data is sent using some means of communication protocols or type, either wired and wireless. For example, the data sent by the environmental sensors can be sent via the Internet, or via Wi-Fi, Bluetooth or any other means known in the art. Next, the decision, selection and transmission computational module, in sub-step d10, performs continuous training, in real time, of the model initially trained in step c). To do this, the decision, selection and transmission computer module uses the following input data: first data received in real time first data received previously, compares the first data received in real time from the environmental sensors with the values of the first data previously received and stored in memory, deciding on the effectiveness of the at least one adaptive behavior recommended in sub-step d7 in minimizing, or reducing, at least one negative externality identified in step b) and, if it concludes that the adaptive behavior was not effective for at least one negative externality associated with the first data collected in real time, it adjusts the adaptive behavior in the at least one scenario used, wherein the adjustment comprises a change in the correspondence between negative externalities and adaptive behaviors, thus continuing to train the model. The results of sub-step d10, i.e. at least one scenario trained or retrained, or adjusted, in sub-step d10, are stored in database 2, in sub-step d11. Step d) is carried out continuously, a plurality of times, preferably more than 10 times, more than 20 times and more preferably more than 50 times, thus obtaining continuous training of the artificial intelligence model, namely deep learning and ensemble methods based on bagging, "boosting" and "random forests" for classification problems for the minimization of negative externalities in a delimited road traffic segment wherein the method of the invention is applied.

[0037] Referring now to the remaining figures, potential approaches and technicalities of realizing the system that implements the method according to the invention will be described, and the components of the aforementioned system will be discussed.

[0038] Figure 2 illustrates the application of the method and system according to one embodiment of the present invention. Figure 2 illustrates a delimited traffic segment (10) considered critical, i.e. typically an urban road section with specific problems or vulnerabilities for which the aim is to minimize problems related to, for example, air pollution, noise, safety, congestion. Next to the traffic segment (10) there may be pedestrian crossings and / or buildings (A) to which vulnerabilities can be associated, such as a school, kindergarten or hospital. For example, in an area wherein there is a hospital or a school, it may be necessary to comply with pollution and noise requirements, and maximum permissible limits are set for that area.

[0039] Points representing pedestrians (80) are shown crossing the pedestrian crossing or walking alongside the segment (10). As can be seen in figure 2, the traffic segment (10) is preceded by an upstream area or zone (11) and, according to the present invention, both the segment (10) and the area (11) are subject to monitoring and action, aimed at reducing or minimizing negative externalities associated with the vulnerabilities of the segment (10). Installed in the vicinity of the traffic segment (10), or even in the traffic segment itself, are environmental sensors (20). Environmental sensors (20) can include, but are not limited to, noise sensors, air pollution sensors, image capture devices, people counting sensors and / or vehicle counting sensors, video cameras, photographic cameras, pollutant gas concentration analyzers, sound level meters, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, sensing coils, thermometers, precipitation sensors, air humidity sensors, wind speed sensors and combinations thereof, for example, wherein the air pollution sensors may be, for example, sensors for the concentration of certain gases, such as NOx, CO 2 , carbon monoxide (CO) or other gases potentially harmful to health and / or the environment. The set of sensors (20) can also include meteorological data sensors, such as temperature, precipitation, wind speed, air humidity. There may also be image sensors (20), or video cameras, installed on the road traffic segment (10), which collect safety-related data, for example, they collect information on the flow of pedestrians (80) on the roadway of the segment (10) or in adjacent areas. In certain embodiments of the present invention, safety-related data can be obtained from existing databases (not shown) with information on accident history and / or density variability of pedestrians, cyclists or other vulnerable users. In the upstream area or zone (11) of the segment (10), as illustrated in figure 2, traffic sensors (21) are installed to monitor traffic and obtain second data, i.e. data relating to, without limitation, the percentage of autonomous and connected vehicles (CAV), the percentage of conventional vehicles (CV), the percentage of heavy vehicles, by time of day, by day of the week, etc. These traffic sensors (21) can be at least one of the group consisting of video cameras, photographic cameras, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, detection coils and combinations of these. Figure 2 also illustrates the presence of autonomous vehicles (CAVs) (60) and conventional vehicles (70), wherein the CAVs (60) are equipped with means of bidirectional vehicle-to-infrastructure communication, or V2I / I2V, as known to a skilled person in the art. Figure 2 also shows the control and evaluation computer module (30) and the decision, selection and transmission computer module (40) according to the present invention. These two computational modules (30,40) can belong to different computational units or to the same computational unit (not shown). Preferably, the computational modules (30,40) belong to the same computational unit. Taking into account that the analysis and evaluation of the current, predicted and controlled state based on informed decision control relies on the processing and analysis of large volumes of data (sometimes from different sources), at least one of the computing modules (30,40) may be an artificial intelligence (AI) computing device, with dedicated, high-performance resources for higher-level planning. At least the decision, selection and transmission module (40) can be equipped with graphics processors from video cards, since these show potential for accelerating and optimizing calculations needed in deep learning, and ensemble methods based on bagging, boosting and random forests for classification problems. Computational performance is greatly increased by the possibility of running multiple, simultaneous calculations, allowing training processes to be distributed and significantly speeding up learning operations. In this way, it is possible to instantly (in 1-3 seconds) provide at least one CAV with a suitable adaptive behavior recommendation considering the upstream (11) or approach zone, as well as the traffic segment (10).

[0040] The control and evaluation computer module (30) comprises a non-volatile memory (not shown) in which database 1 is housed. Database 1 contains first data captured by the environmental sensors (20) and second data captured by the traffic sensors (21), i.e. database 1 contains data related to negative externalities (air pollution, e.g., NOx and PM concentration, CO 2 concentration, noise), traffic profile data (fleet volume and composition) and priorities of the negative externalities calculated in step b). The decision, selection and transmission computational module (40) comprises a non-volatile memory (not shown) in which a database 2 is housed. Both computational modules (30) and (40) involve communication features configured to receive data sent by the sensors (20,21). When the modules (30,40) are not part of the same computing unit, they are both equipped with means to communicate with each other. These means of communication can be wired or wireless and can be of a type known to an expert in the field, for example, they can communicate via any OTA (over-the-air) communication system, such as wireless WAN networks or another wireless network. The decision, selection and transmission computer module (40) is further equipped with wired or wireless communication means for communicating with transmission units (50), discussed below. The system according to the invention includes communication means comprising a device in the CAV (60), a radio communication network (not shown) between the decision, selection and transmission module (40) and a transmission unit (50) and the CAV (60) to transmit the best adaptive behavior to achieve a positive impact, i.e. reduce at least one negative externality.

[0041] As further illustrated in figure 2, there are transmission units (50) configured to communicate with the CAVs (60) by issuing recommendations with adaptive behaviors according to the present invention. The transmission units (50) are installed in the vicinity of the segment (10) and / or the upstream zone (11) and comprise communication means configured to receive recommendations / communication from a decision, selection and transmission computational module (40) and transmit the aforementioned recommendations to at least one CAV (60) circulating in the vicinity, in order to cause a change in behavior on the part of the aforementioned at least one CAV (60) and thereby contribute to reducing at least one negative externality. In general terms, figure 2 illustrates the system implementing the computer-implemented method according to the invention, namely steps a), b) and c) described above. In particular, figure 2 illustrates the circulation of CAVs (60) and CVs (70) in an upstream zone (11) of the traffic segment (10) and on the segment itself (10). Environmental sensors (20) collect data associated with the traffic segment (10), or first data, and send it to the control and evaluation computer module (30). Traffic sensors (21) also collect data associated with the composition of traffic, or second data, and send it to the control and evaluation computer module (30). The full arrows drawn on the sidewalk of the traffic segment (10) indicate the adaptive behaviors that can be received by the CAVs in a recommendation sent by the decision, selection and transmission computational module (40) and transmitted by the transmission unit (50). These adaptive behaviors can consist of a recommendation to increase speed, to reduce speed, to change lanes, or to change direction.

[0042] Figure 3 shows a graph illustrating an example of the relative importance of priorities of negative externalities to be minimized with the method and system according to the present invention. In particular, according to the method of the invention and as described above, in step b) priorities are identified according to the negative externalities and vulnerabilities detected in step a) of the method according to the invention. The methodology employed for this purpose can be based on the use of time series of data collected by environmental and weather sensors (air quality, traffic, noise, temperature, humidity, wind, precipitation, safety, exposed population, congestion) in order to group typologies of negative externalities according to the level of priority. Figure 3 graphically exemplifies how such negative externalities can be prioritized and grouped over 24 hours. Referring to figure 3, the following periods over the course of a day were grouped together: I1 - night rest period - this period has a low density of pedestrians and is dominated by the prioritization of minimizing noise (E2) and greenhouse gases (E3); I2 - period of high concentration of pedestrians and young children crossing the road - during this period top priority is given to the safety and protection of pedestrians (E7); I9 - period with moderate pedestrian concentration and air pollution below the limits - in this period there is an even distribution of the various negative externalities to be minimized; I4 - period in which the concentration of ozone (E4) approaches the limits for human health - in this period there is an increase in the level of importance of air quality and emissions of NOx (E7) and PM (E6); I7 - late afternoon rush hour period - a period when priority is given to relieving congestion (E1) and ensuring the safety of pedestrians (E8).

[0043] Step b) of determining traffic profiles (TP), negative externalities and priorities on the aforementioned delimited road traffic segment (10) comprises monitoring the traffic composition throughout the day, namely the flow (vehicles per hour) upstream of the segment to be optimized, and the category of vehicles in circulation and the weather conditions. Counting can be carried out using the traffic sensors (21) described above, wherein vehicles are counted and their category identified (e.g. CAV, CV, heavy vehicles, etc.). The data is collected, sent to the control and evaluation module (30) and stored in a database 1. Traffic monitoring can be carried out in an upstream area (11) of the traffic segment (10), allowing the necessary time for the identification of the traffic profile, selection of negative externalities to be minimized, choice of adaptive behavior and transmission via means of communication between infrastructures, or the decision, selection and transmission module (40) and the CAV vehicles, according to step d) of the method of the invention.Example of a set of data collected by the system according to the invention

[0044] Table 1 shows an example of the structure of the data collected by the environmental sensors (20) and traffic sensors (21) in an embodiment according to the invention, wherein this data is stored in database 1, as already described. Table 1 Structure of variables to be monitored and stored in the database Traffic Profile 8h33Precipitation 0 mm / hTemperature 8°CWind 34 km / hHumidity 85%Ice NoLevel of automation of the vehicle fleetL0L1L2L3L4L520%28%22%11%10%9%DieselPetrolBEVHEVPHEV20%10%30%30%10%CategoryLightHeavyMotorcycleBicycles70%15%10%5%

[0045] Wherein BEV refers to "Battery Electric Vehicle", HEV refers to "Hybrid Electric Vehicle", PHEV refers to "Plug-in Hybrid Electric Vehicle", i.e. a hybrid electric vehicle that can be recharged from an external source.

[0046] As mentioned above, in step c) of defining scenarios for the different traffic profiles and priorities determined in step b) for the delimited road traffic segment (10), the traffic impacts for different levels of vehicle composition and different adaptive behaviors for the CAVs are estimated using the longitudinal and lateral control systems (see figure 4).

[0047] Figure 4 illustrates a CAV (60) according to the invention, wherein the means of communication that a CAV typically has are indicated and are as follows: LIDAR system (1), which is a system that makes it possible to measure the distances to various objects in the vicinity of the CAV (60) and also makes it possible to create three-dimensional images of the detected objects, mapping the environment around the CAV (60); RADAR system (2) consisting of a set of short- and long-range radar sensors installed around the vehicle for assistance, distance maintenance control and braking assistance; three-dimensional image cameras (3) which automatically detect objects (other vehicles, pedestrians, cyclists, road signs, road markings, bridges, among others), classify them, and determine the distances between them and the vehicle; satellite navigation system (GNSS) (4), which allows geographical location with numerical coordinates (e.g. latitude, longitude) representing their physical location in space. They can also navigate by combining GPS coordinates in real time with other digital mapping data (e.g. via Google Maps); and communication module with the transmission unit (5), i.e. a vehicle-to-infrastructure communication unit (V2I), which is a control unit for receiving instructions wirelessly via the transmission unit (50) or communication protocol (V2X) on adaptive behavior and incorporating the recommendations sent by the transmission unit (50) into the CAV's planning and decision algorithms (60).

[0048] The longitudinal and lateral control systems in CAVs ensure safe movement in a lane. Longitudinal control, maintained by the cruise controls, controls vehicle speeds using the brakes or accelerator and maintains a safe distance between two vehicles. Cooperative adaptive cruise control (CACC) systems integrate dedicated short-range communication (DSRC) technologies and / or 5G technology to establish vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication. It should be noted that GNSS technology can have decimal-level accuracy to keep CAVs in their lane and maintain a safe distance from other vehicles. The longitudinal control strategy in CAVs keeps the distance between two vehicles and the speed stabilized. In various situations, this strategy can be adjusted slightly to achieve a certain objective from the network manager's point of view. For example, in situations of moderate traffic, the speed can be optimized to influence the speed of other traffic and minimize the emission of noise or a given pollutant. In situations of severe congestion, the distance between two vehicles can be slightly reduced to increase traffic flow capacity on the critical segment of the network. In terms of lateral behavior, it can be advantageous to use traffic lanes to increase the capacity of the roadway from the start or to better manage the distribution of pollutants emitted in traffic segments (10). Figure 4 shows the communication systems of a CAV that could potentially be used in the present invention. Figure 4 also illustrates a CV (70) circulating on the track, wherein the CAV (60) detects the distance to the CV (70) and can control this distance by means of adaptive behavior recommendations received from the transmission unit (50). As an example, the recommendations received by one or more CAVs (60) from the transmission unit (50) can include adaptive behaviors such as adjusting the speed at free speed, so that the average speed of the group of vehicles circulating, and taking into account their composition and atmospheric pollutants, can minimize a critical pollutant, according to the priorities established as described above. In addition, CAVs can act as moderating agents by promoting traffic calming actions to minimize the risk of accidents or collisions during critical periods, also in accordance with the priorities established as described above.

[0049] Figure 5 illustrates a recommendation sent to the CAVs (60), comprising adaptive behavior to adjust the optimum speed at free regime (38 km / h) to minimize CO 2 emissions and noise in an integrated manner. Figure 5 also illustrates part of the traffic segment (10) next to a critical building (A), wherein CAV (60) and CV (70) circulate, wherein the CAVs (60) receive recommendations from the transmission unit (50), the result of processing the data sent by the sensors (20,21), as described above, to the control and evaluation (30) and decision, selection and transmission (40) computational modules, as also described above, in which case the process of identifying the optimum speed to respond to a set of priorities (I1) (see figure 3) and an upstream traffic profile (TP2) is exemplified. This information is transmitted to the CAVs (60) in an area (11) upstream of the traffic segment (10). In the case illustrated in figure 5, the adaptive behavior selected by the decision, selection and transmission computational module is the one that has the greatest influence on reducing CO 2 emissions and noise, taking into account the traffic profile. For example, in figure 5 it can be seen that the CAV (60) that is still circulating outside the traffic segment (10) is circulating at a speed of 53 km / h, still without system intervention according to the invention, while the CAV (60) that is circulating inside the traffic segment (10) is already circulating at the speed recommended by the transmission unit (50), the optimum speed calculated by the decision, selection and transmission module (40), 38 km / h. The graph shown in figure 5 indicates the optimum speed of 38 km / h for observing the priorities detected for the traffic profile in question and taking into account the negative externalities present, as shown in figure 3.

[0050] Table 2 exemplifies a consolidated data structure facilitating the process of selecting adaptive behaviors to send to one or more CAVs. Table 2 - Example of a data structure for selecting adaptive behaviors Traffic ProfileAdaptive behaviorNoiseNOxCO 2 Accident riskDelayTPAB1, AB2- 4 dB4%3.16%0+2%AB1- 4 dB12%9.48%0+15%AB2- 8 dB-3%-2.37%03%AB3+ 4 dB-4%-3.16%4%2%ABN- 3 dB7%5.53%-20%1%..................

[0051] Wherein ABN is a hypothetical adaptive behavior of order N that reduces traffic flow noise, for example, by 3 dB, as shown in Table 2.

[0052] The adaptive behavior of order N, ABN, exemplified in Table 2, also reduces the risk of an accident by 20%, although it increases the emission of NOx and CO 2 which, for the scenario exemplified, will not be such negative externalities as noise or the risk of an accident.

[0053] For example, at moderate speeds, vehicles may tend to move more evenly, which can reduce turbulence and noise emissions. Any of these changes in speed has an influence on travel time and therefore delay.

[0054] Figure 6 represents yet another case in which the CAVs (60-A, 60-B) circulating in the vicinity of segment (10), through the application of the method and system according to the invention, receive recommendations (M1, M2) to change lanes in order to reduce the emission of pollutants, taking into account the pedestrians (80) present on segment (10). In this case, the CAV (60-A) is powered by electricity and is therefore considerably less polluting and receives the recommendation (M1) to change lanes to the left, while the CAV (60-B), which is powered by internal combustion and emits a higher level of pollution, receives the recommendation (M2) to change lanes to the right. In this way, the level of pollution to which pedestrians (80) are exposed is reduced.

[0055] Figure 7 illustrates another case of application of the method and system of the present invention. In figure 7, the priority to be observed is to relieve congestion (I4, as shown in figure 3). In this case, the decision, selection and transmission module has calculated an optimum distance of 0.7 m to help relieve traffic congestion. The recommendation given to the CAVs (60) present is to maintain this distance from the vehicle in front of them. Vehicle density (k), representing the number of vehicles per kilometer of track, can be optimized using parameters such as the average distance between vehicles when stationary in a queue or the increase in the minimum following distance due to vehicles moving. Figure 7 exemplifies the transmission of an adaptive behavior on the optimized distance to the vehicle in front, taking into account a specific traffic profile (TP3) to minimize a negative externality with priority, i.e. in this case congestion. Traffic flow (q) represents the number of vehicles per hour. In the fundamental traffic diagram shown in figure 7, as traffic density (k) increases, traffic flow (q) also increases until it reaches a peak or maximum value. From that point on, traffic flow (q) begins to decrease as traffic density (k) continues to increase. This happens because, at very high densities, vehicles start to move more slowly, reducing the flow (volume of cars passing through a section).

[0056] In a first aspect, the invention relates to a computer-implemented method for minimizing negative externalities in road traffic segments comprising the following steps: a) determination of vulnerabilities in a defined road traffic segment (10); b) determination of traffic profiles (TP), negative externalities and priorities in the aforementioned delimited road traffic segment (10); c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment (10), taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model; wherein step a) of determination of vulnerabilities in a delimited road traffic segment (10) comprises the following sub-steps: a1) collection of first data dn, by means of a set of environmental sensors (20) installed on or near the aforementioned delimited road traffic segment (10), the first data dn being related to first parameters of the delimited road traffic segment (10); a2) transmission of the first data collected in sub-step a1), by each of the environmental sensors (20), to a control and evaluation computer module (30), via wired or wireless means of communication; a3) consultation, by the control and evaluation computer module (30), of a weather condition database; a4) assignment of threshold values by the control and evaluation computer module (30), which can be maximum or minimum threshold values xn for the first data dn collected in sub-step a1); a5) comparison, by the control and evaluation computer module (30), of each of the first data dn collected in sub-step a1), with the respective threshold values xn assigned in sub-step a4), for each of the parameters of the delimited traffic segment (10); a6) determination, by the control and evaluation computer module (30), of the presence of vm vulnerabilities, relative to the parameters wherein the first data dn and quantification of these vm vulnerabilities, according to the range of the difference between the first data dn and the threshold xn values and considering the meteorological conditions obtained in sub-step a3); a7) storage, by the control and assessment computer module (30), of the vm vulnerabilities and their quantification, i.e. the results obtained in sub-step a6) in a database 1 housed in a non-volatile computer memory of the aforementioned control and assessment computer module (30); wherein, optionally, in sub-step a6), the control and assessment computer module (30) also considers other data associated with vulnerabilities retrieved from the aforementioned database 1 and / or receives, via input by a network operator or manager, other data relating to vulnerabilities through a user interface of the control and assessment computer module (30). wherein step b) of determination of traffic profiles, negative externalities and priorities in the aforementioned segment of road traffic delimited (10) comprises the following sub-steps: b1) collection of second data f n , by means of a set of traffic sensors (21) installed in an area (11) upstream of the aforementioned delimited road traffic segment (10), the second data f n being relative to the composition of the vehicles circulating in the area (11); b2) sending the second data collected in sub-step b1), by each of the traffic sensors (21) to a control and evaluation computer module (30), via wired or wireless means of communication; b3) based on the second data collected in step b1), the control and evaluation computer module (30) draws up a model of the traffic profiles, wherein the modeling includes grouping the data by at least one of vehicle types, percentage of heavy vehicles, percentage of connected, autonomous vehicles, by date / time, by day of the week, by day of the month, by weather conditions; b4) based on the traffic profile modeling carried out in the previous sub-step, and considering the vulnerabilities and their quantification calculated in step a), the control and evaluation computer module (30) calculates the negative externalities associated with the traffic segment (10); b5) prioritization, by the control and evaluation computer module (30), of the negative externalities associated with the traffic segment (10); (b6) storage by the control and evaluation computer module (30) of the results of sub-steps (b3), (b4) and (b5) in a database 1 housed in the computer memory of the aforementioned control and evaluation computer module (30); wherein step c) of definition of scenarios relates the traffic profiles determined in step b) with adaptive behavior for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model and comprises the following sub-steps carried out by the control and evaluation computer module (30): c1) reading input data to train a model, starting from database 1, wherein the input data comprises: negative externalities, traffic profiles, and / or priorities of the negative externalities calculated in step b); c2) reading of adaptive behaviors ABp, from database 2, wherein the aforementioned adaptive behaviors ABp are adequate to minimize negative externalities and wherein the adaptive behaviors ABp are at least one of the group constituted by: longitudinal control of an autonomous vehicle (60), lateral control of an autonomous vehicle (60), wherein the longitudinal control of an autonomous vehicle (60) may comprise decreasing the speed or increasing the speed at which the autonomous vehicle (60) travels, or a combination of these instructions. c3) selection of one or more adaptive ABp behaviors and simulation of their application to a specific set of input data read in sub-steps c1) and c2); c4) determining ci scenarios that relate a specific set of input data to one or more adaptive ABp behaviors, resulting from the training of the model carried out in sub-step c3); c5) storing the results of sub-step c4) in a database 2 housed in the computer memory of the aforementioned control and evaluation computer module (30); wherein step d) of recognition of traffic profile patterns, identifies and communicates adaptive behaviors, and performs continuous training of the initially trained model, step d) comprising the following sub-steps: d1) collection of second data f n+1 , by means of the set of traffic sensors (21) installed in the zone (11) upstream of the aforementioned delimited road traffic segment (10), the second data f n+1 being relative to the composition of the vehicles circulating in the zone (11), and including information on which autonomous vehicles (60) are connected; d2) sending the second data f n+1 collected in sub-step d1), by each of the traffic sensors (21) to a decision, selection and transmission computer module (40), via some wired or wireless means of communication; d3) retrieval, by the decision, selection and transmission computer module (40), of scenarios stored in database 2; d4) comparison, by the decision, selection and transmission computational module (40), of the second data f n+1 with the traffic profiles of the scenarios retrieved in d3), and retrieval of the adaptive behaviors stored in database 2; d5) by recognizing traffic profiles defined by the second data f n+1 , deciding on at least one adaptive behavior to be applied to at least one connected autonomous vehicle circulating in the upstream zone (11) of the traffic segment (10) or in the traffic segment itself (10); d6) sending at least one adaptive behavior to be applied, by wired or wireless communication means from the decision, selection and transmission module (40), to one or more transmission units (50); d7) sending a recommendation with at least one adaptive behavior to be applied, by one or more transmission units (50), via wireless communication means, to at least one connected autonomous vehicle (60) circulating in the area (11) upstream or in the traffic segment (10) identified in the second data f n+1 collected in sub-step d1); d8) collection of the first data d n+1 , by means of the environmental sensors (20) installed on or near the delimited road traffic segment (10), the first data d n+1 relating to the first real-time parameters of the delimited road traffic segment (10) and consultation, by the decision, selection and transmission computer module (40), of the aforementioned weather conditions database; d9) sending the first data collected in sub-step d8), by each of the environmental sensors (20), to the decision, selection and transmission computer module (40), via wired or wireless means of communication; d10) the decision, selection and transmission computer module (40), based on the following input data: first data received in real time d n+1 , first data received previously d n , makes a comparison between these two sets of first data d n+1 and d n , wherein, if this difference reveals a reduction of at least one of the negative externalities identified in step b), the module concludes on the effectiveness of the application of at least one adaptive behavior sent in sub-step d7) and wherein, if this difference reveals a non-reduction of at least one of the negative externalities identified in step b), the module concludes on the ineffectiveness of the application of at least one adaptive behavior sent in sub-step d7) and inserts an adjustment in at least one of the scenarios determined in step c), wherein the adjustment comprises a change in the correspondence between negative externalities and adaptive behaviors, resulting in at least one adjusted scenario; d11) storing the results of sub-step d10, i.e. at least one scenario considered effective and / or at least one adjusted scenario; wherein step d) is performed n times, wherein n is greater than or equal to one, preferably n is greater than 50, and wherein step d) is performed in a processing cycle lasting from 1 second to 60 minutes.

[0057] In an embodiment of the first aspect of the invention, in sub-step a1) of collecting first data, the aforementioned first parameters are at least one of the group consisting of pedestrian concentration rate, vehicle concentration rate, gas concentration rate, noise, safety level, congestion or a combination of these.

[0058] In another embodiment of the first aspect of the invention, in sub-step a4), the assignment of threshold values is made for different periods of time, for example, the assignment of threshold values is made by day of the week, by hour of the day of the week, by season, or by a combination of these.

[0059] In another embodiment of the first aspect of the invention, in sub-step b5), the determination of priorities is based on the proximity between the first data d n and the corresponding threshold values, period of the day, week, month, type of population living nearby, type of buildings in the vicinity of the delimited traffic segment (10).

[0060] In another embodiment of the first aspect of the invention, sub-step d7) comprises the following sub-steps: d71) obtaining confirmation that the autonomous vehicle (60) has carried out the recommendation sent in sub-step d7) in at least one of the ways belonging to the group consisting of: collection of third data g n+1 , by the traffic sensors (21) with subsequent sending, by the traffic sensors (21), of the third data g n+1 collected, to the computational decision, selection and transmission module (40); the one or more autonomous vehicles (60) send a recommendation acceptance signal to the transmission unit (50), whereupon the transmission unit (50) sends these one or more acceptance signals to the decision, selection and transmission computational module (40); storing, in a memory of the one or more autonomous vehicles (60), the acceptance of the recommendation received in sub-step d7); d72) processing, by the decision, selection and transmission computational module (40), of the third data g n+1 and / or of the one or more recommendation acceptance signals received in sub-step d71), wherein the aforementioned processing of the third data g n+1 includes a comparative evaluation with other third data g n collected at previous time instants; wherein the third data g n , g n+1 includes data from the one or more autonomous vehicles (60) that received one or more recommendations; and in the comparison step d10), the third data g n+1 and / or the one or more recommendation acceptance signals sent by the one or more autonomous vehicles (60) are additionally included in the input data.

[0061] In another embodiment of the first aspect of the invention, in sub-step d10), the effectiveness of the at least one adaptive behavior applied is concluded on the basis of at least one of: the number of people present, the length of time people are present, the decrease in exposure to polluting gases such as NOx, PM, CO 2 , the decrease in noise, the increase in safety, the decrease in congestion.

[0062] In another embodiment of the first aspect of the invention, in sub-step c) of definition of scenarios for the delimited road traffic segment (10), the definition of scenarios is carried out for different periods of time, for example, the definition of scenarios is carried out by day of the week, by hour of the day of the week, by season, or by a combination of these.

[0063] In another embodiment of the first aspect of the invention, in the sending recommendation sub-steps a2), b2), d2), d6), d9) the transmission of instructions is carried out by wired or wireless means of communication; and in the sending sub-step d7) the transmission is carried out by wireless means of communication.

[0064] In another embodiment of the first aspect of the invention, step c) is optionally carried out by the decision, selection and transmission module (40).

[0065] In a second aspect, the present invention relates to a system configured to implement the computer-implemented method as defined above in relation to the first aspect of the invention, wherein the system comprises: a set of environmental sensors (20) installed on and / or near a delimited road traffic segment (10), wherein the set of environmental sensors (20) comprises at least one of the group consisting of video cameras, photographic cameras, pollutant gas concentration analyzers, sound level meters, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, sensing coils, thermometers, precipitation sensors, air humidity sensors, wind speed sensors, and combinations thereof; a set of traffic sensors (21) installed in an upstream area (11) of the aforementioned delimited road traffic segment (10), wherein the set of traffic sensors (21) comprises at least one of the group consisting of video cameras, photographic cameras, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, detection coils and combinations of these; a control and evaluation computing module (30), wherein the computing module (30) comprises control and evaluation computing module communication means, control and evaluation computing module database storage means, and control and evaluation computing module processing means; a decision, selection and transmission computing module (40), wherein a decision, selection and transmission computing module (40) involves decision, selection and transmission computing module communication means, decision, selection and transmission computing module database storage means and decision, selection and transmission computing module processing means; at least one CAV (60) which is configured to drive partially or fully self-controlled on the delimited road traffic segment (10), in its vicinity or in the upstream area (11) of the aforementioned traffic segment (10) and comprises communication means suitable for receiving recommendations from a transmission unit (50); at least one transmission unit (50) configured to communicate with the at least one CAV (60); optionally, at least one conventional vehicle CV (70) which is configured to circulate with a driver and without being partially or totally self-controlled in the delimited road traffic segment (10), in its vicinity or in the upstream area (11) of the aforementioned traffic segment (10); wherein the control and evaluation computational module (30) and the decision, selection and transmission computational module (40) are included in the same computational unit or in different computational units.

[0066] In an embodiment of the second aspect of the invention, the aforementioned control and evaluation computer module (30) and the aforementioned decision, selection and transmission computer module (40) are selected from the group consisting of a local computer, a remote computer, a server, a tablet, a smart phone, such as a smartphone.

[0067] In a third aspect, the present invention relates to a computer program comprising logical instructions to enable the control and evaluation computer module (30), comprised in the system as defined in relation to the second aspect of the present invention, to carry out the steps of the method as defined in relation to the first aspect of the present invention.

[0068] In a fourth aspect, the present invention relates to a computer program comprising logical instructions to enable the decision, selection and transmission computer module (40), comprised in the system as defined with regard to the second aspect of the present invention, to carry out the steps of the method as defined with regard to the first aspect of the present invention.

[0069] In a fifth aspect, the present invention relates to a means of reading by a computer apparatus comprising the installation of the computer program as defined in any of the third or fourth aspects of the present invention.DEFINITIONS

[0070] As used in this application, the term "delimited traffic segment" or "delimited road traffic segment" refers to a section of roadway and adjacent areas in which vulnerabilities exist, or potentially exist, and to which the present invention is applicable.

[0071] As used in this request, the term "vulnerability" refers to a characteristic associated with the site, i.e. the delimited traffic segment , and is related to human activity, namely the socio-demographic profile of the population, the occupation of the public road and adjacent areas, and the degree of exposure to negative externalities associated with traffic, for example, and without limitation, noise, air pollution, safety, congestion. In this application, vulnerability is measured in terms of proximity or even exceeding established or desirable thresholds for characteristics associated with the location, such as noise levels, concentration levels of certain harmful gases, speed associated with the safety of those present in the traffic segment, among others.

[0072] As used in this request, the term "negative externality" refers to a negative effect that traffic has on the vulnerabilities of the delimited road segment, i.e. it refers to the worsening of the measured values associated with one or more vulnerabilities identified in a specific road traffic segment, for example, approaching or exceeding established or desirable thresholds.

[0073] As used in this request, the term "connected autonomous vehicle or CAV" refers to motor vehicles that have the means for at least partial autonomous driving and fall within one of the levels of automation greater than or equal to 3 according to the SAE (Society of Automotive Engineers), i.e. vehicles with a degree of autonomy of level SAE3, SAE4 or SAE5, with or without a driver, and which incorporate a planning module capable of interpreting information received from an I2V / V2I transmission unit (infrastructure-to-vehicle / vehicle-to-infrastructure), adapting speed, lane choice, acceleration or deceleration rate, distance to the vehicle in front. The planning module also takes into account the mechanical, dynamic and kinetic characteristics of the vehicle.

[0074] As used in this request, the term "conventional vehicle or CV " refers to vehicles with a degree of autonomy less than or equal to SAE2 level, with a driver.

[0075] As used in this application, the term "adaptive behavior" refers to a behavior that is recommended to an autonomous vehicle by changing the expected behavior that autonomous vehicle would have, in order to contribute to causing a positive impact or impact, i.e. decreasing one or more negative externalities in the traffic segment where it is, or will be, circulating. Adaptive behavior can be an increase in speed, a decrease in speed, a change of lane, among others.

[0076] As used in this request, the term "positive impact" or "impact" refers to the positive effect that adaptive behaviors have on one or more vulnerabilities, translating into a reduction in negative externalities, i.e. a departure from the established or desirable thresholds for a specific delimited traffic segment.

[0077] As used in this request, the term "traffic profile" or "TP" refers to a set of aggregated data that characterizes the type of vehicles circulating in the delimited traffic segment, by hour of the day, by day of the week, by type of vehicle, among others. Traffic data and its negative externalities are grouped using a measure of (dis)similarity between objects, requiring the definition of the number of groups (k = 1, 2, 3, ..., n).

[0078] As used in this application, the term "deep learning" refers to a training process belonging to the concept of "machine learning", in which e.g. artificial neural networks, i.e. algorithms or sets of logical instructions modeled to function like the human brain, learn from large amounts of data.

[0079] As used in this description, the expressions "about" and "approximately" refer to a range of values of plus or minus 10% the specified number.

[0080] As used throughout this patent application, the term "or" is used in the inclusive sense rather than the exclusive sense, unless the exclusive sense is clearly defined in a specific situation. In this context, a phrase such as "X uses A or B" should be interpreted as including all relevant inclusive combinations, for example "X uses A", "X uses B" and "X uses A and B".

[0081] As used throughout this patent application, the indefinite articles "one" or "a" should generally be interpreted as "one or more" and "one or more", unless the meaning of a singular embodiment is clearly defined in a specific situation.

[0082] As presented in this description, terms related to examples should be interpreted with the purpose of illustrating an example of something and not indicating a preference.

[0083] The subject-matter described above is provided as an illustration of the present invention and should not be construed to limit it. Terminology used for the purpose of describing specific embodiments in accordance with the present invention should not be construed to limit the invention. As used in the description, the definite and indefinite articles, in their singular form, are intended to be interpreted as also including the plural forms, unless the context of the description explicitly indicates otherwise. It will be understood that the terms "comprise" and "include", when used in this description, specify the presence of the related features, elements, components, steps and operations, but do not exclude the possibility that other features, elements, components, steps and operations are also covered.

[0084] All changes, provided they do not modify the essential characteristics of the following claims, are to be considered within the scope of protection of the present invention.LIST OF REFERENCE SIGNS

[0085] a1 - collecting first data a2 - sending the first data a3 - consulting the weather database a4 - assigning threshold values a5 - comparison of the first data with the threshold values a6 - determining vulnerabilities and quantifying them a7 - storing the results of a6 b1 - collecting second data b2 - sending the second data b3 - traffic profile modeling and clustering b4 - calculating negative externalities b5 - determining priorities b6 - storing the results obtained in b4 c1 - reading priorities and traffic profiles c2 - reading adaptive behaviors (AB) c3 - selecting a AB suitable for the scenario and simulating the application of the AB in the scenario c4 - determining the relationship between scenarios and appropriate ABs to be applied in the trained scenarios c5 - storing the trained model in database 2 d1 - collecting second data d2 - sending the second data d3 - scenario recovery d4 - comparing real-time traffic profiles with AB scenarios and recovery d5 - decision on AB to be applied d6 - sending the AB d7 - transmission from AB to CAV d8 - first data Dn+1 collection d9 - sending the first data collected d10 - comparing and deciding which AB to keep or change d11 - storing the new scenarios learned 1 - LIDAR system in a CAV 2 - RADAR system in a CAV 3 - Three-dimensional imaging cameras in a CAV 4 - GNSS satellite navigation system in a CAV 5 - communication module with a transmission unit 10 - delimited road traffic segment 11- area or zone upstream of the delimited road traffic segment 20 - environmental sensors 21 - traffic sensors 30 - control and evaluation computer module 40 - decision, selection and transmission computer module 50 - transmission unit 60 - connected and / or autonomous vehicle or CAV 70 - conventional vehicle or CV 80 - pedestrians LIST OF CITATIONS

[0086] Here is the list of citations:PATENT LITERATURE

[0087] US Patent No. US 11,360,236 B1, by Prathamesh Khedekar, Downers Grove, IL (USA), published on June 14, 2022; Chinese patent application No. CN111405522, by Suzhou Kunpeng Intelligent Network Tech Co Ltd, published on July 10, 2020. NON-PATENT LITERATURE

[0088] [1] Bandeira, Jorge M., Eloisa Macedo, Paulo Fernandes, Monica Rodrigues, Mario Andrade, and Margarida C. Coelho. 2021. "Potential Pollutant Emission Effects of Connected and Automated Vehicles in a Mixed Traffic Flow Context for Different Road Types." IEEE Open Journal of Intelligent Transportation Systems 2: 364-83. [2] Mohebifard, Rasool, and Ali Hajbabaie. 2020. "Effects of Automated Vehicles on Traffic Operations at Roundabouts." In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), 1-6. IEEE. [3] Rafael, Sandra, Paulo Fernandes, Diogo Lopes, Micael Rebelo, Jorge Bandeira, Eloisa Macedo, Mónica Rodrigues, Margarida C. Coelho, Carlos Borrego, and Ana I. Miranda. 2022. "How Can the Built Environment Affect the Impact of Autonomous Vehicles' Operational Behavior on Air Quality?" Journal of Environmental Management 315 (August): 115154.

Examples

Embodiment Construction

[0026]The technological solution of the present invention relates to a computer-implemented method for minimizing the negative externalities present in a given road traffic segment and also relates to a system comprising, among others, sensors and computational units that execute the method according to the invention.

[0027]The computer-implemented method according to the invention comprises four main steps:

a) determining vulnerabilities in a defined road traffic segment; b) determining traffic profiles, negative externalities and priorities in the road traffic segment; c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile -patterns, identification and communication of adaptive behaviors, and continued training of the in...

Claims

1. A computer-implemented method for minimizing negative externalities in road traffic segments characterized by comprising the following steps: a) determination of vulnerabilities in a defined road traffic segment (10); b) determination of traffic profiles, negative externalities and priorities in the aforementioned delimited road traffic segment (10); c) definition of scenarios that relate the different traffic profiles determined in step b) with adaptive behaviors for the delimited road traffic segment (10), taking into account the negative externalities and priorities determined in step b), through the initial training of a model; and d) recognition of traffic profile patterns, identification and communication of adaptive behaviors, and continued training of the initially trained model; wherein step a) of determination of vulnerabilities in a delimited road traffic segment (10) comprises the following sub-steps: a1) collection of first data dn, by means of a set of environmental sensors (20) installed on or near the aforementioned delimited road traffic segment (10), the first data dn being related to first parameters of the delimited road traffic segment (10); a2) sending the first data collected in sub-step a1), by each of the environmental sensors (20), to a control and evaluation computer module (30), via wired or wireless means of communication; a3) consultation, by the control and evaluation computer module (30), of a weather condition database; a4) assignment of threshold values by the control and evaluation computer module (30), which can be maximum or minimum threshold values xn for the first data dn collected in sub-step a1); a5) comparison, by the control and evaluation computer module (30), of each of the first data dn collected in sub-step a1), with the respective threshold values xn assigned in sub-step a4), for each of the parameters of the delimited traffic segment (10); a6) determination, by the control and evaluation computer module (30), of the presence of vm vulnerabilities, relative to the parameters wherein the first data dn and quantification of these vm vulnerabilities, according to the range of the difference between the first data dn and the threshold xn values and considering the meteorological conditions obtained in sub-step a3); a7) storage, by the control and assessment computer module (30), of the vm vulnerabilities and their quantification, i.e. the results obtained in sub-step a6) in a database 1 housed in a non-volatile computer memory of the aforementioned control and assessment computer module (30); wherein, optionally, in sub-step a6), the control and assessment computer module (30) also considers other data associated with vulnerabilities retrieved from the aforementioned database 1 and / or receives, via input by a network operator or manager, other data relating to vulnerabilities through a user interface of the control and assessment computer module (30). wherein step b) of determination of traffic profiles, negative externalities and priorities in the aforementioned segment of road traffic delimited (10) comprises the following sub-steps: b1) collection of second data fn, by means of a set of traffic sensors (21) installed in an area (11) upstream of the aforementioned delimited road traffic segment (10), the second data fn being relative to the composition of the vehicles circulating in the area (11); b2) transmission of the second data collected in sub-step b1), by each of the traffic sensors (21) to a control and evaluation computer module (30), via some wired or wireless means of communication; b3) on the basis of the second data collected in step b1), the control and evaluation computer module (30) draws up a model of the traffic profiles, wherein the modeling includes grouping the data, by at least one of vehicle types, percentage of heavy vehicles, percentage of autonomous vehicles, by date / time, by day of the week, by day of the month, by weather conditions; b4) based on the traffic profile modeling carried out in the previous sub-step, and considering the vulnerabilities and their quantification calculated in step a), the control and evaluation computer module (30) calculates the negative externalities associated with the traffic segment (10); b5) prioritization, by the control and evaluation computer module (30), of the negative externalities associated with the traffic segment (10); (b6) storage by the control and evaluation computer module (30) of the results of sub-steps (b3), (b4) and (b5) in a database 1 housed in the computer memory of the aforementioned control and evaluation computer module (30); wherein step c) of definition of scenarios relates the traffic profiles determined in step b) with adaptive behavior for the delimited road traffic segment, taking into account the negative externalities and priorities determined in step b), through the initial training of a model and comprises the following sub-steps carried out by the control and evaluation computer module (30): c1) reading input data to train a model, starting from database 1, wherein the input data comprises: - negative externalities, - traffic profiles, and / or - priorities of the negative externalities calculated in step b); c2) reading of adaptive behaviors ABp, from database 2, wherein the aforementioned adaptive behaviors ABp are adequate to minimize negative externalities and wherein the adaptive behaviors ABp are at least one of the group constituted by: - longitudinal control of an autonomous vehicle (60), - lateral control of an autonomous vehicle (60), wherein the longitudinal control of an autonomous vehicle (60) may comprise decreasing the speed or increasing the speed at which the autonomous vehicle (60) travels, or a combination of these; c3) selection of one or more adaptive ABp behaviors and simulation of their application to a specific set of input data read in sub-steps c1) and c2); c4) determining ci scenarios that relate a specific set of input data to one or more adaptive ABp behaviors, resulting from the model training carried out in sub-step c3); c5) storing the results of sub-step c4) in a database 2 housed in the computer memory of the aforementioned control and evaluation computer module (30); wherein step d) of recognition of traffic profile patterns, identifies and communicates adaptive behaviors, and performs continuous training of the initially trained model, step d) comprising the following sub-steps: d1) collection of second data fn+1, by means of the set of traffic sensors (21) installed in the zone (11) upstream of the aforementioned delimited road traffic segment (10), the second data fn+1 being relative to the composition of the vehicles circulating in the zone (11), and including information on which autonomous vehicles (60) are connected; d2) sending the second data fn+1 collected in sub-step d1), by each of the traffic sensors (21) to a decision, selection and transmission computer module (40), via some wired or wireless means of communication; d3) retrieval, by the decision, selection and transmission computer module (40), of scenarios stored in database 2; d4) comparison, by the decision, selection and transmission computational module (40), of the second data fn+1 with the traffic profiles of the scenarios retrieved in d3), and retrieval of the adaptive behaviors stored in database 2; d5) by recognizing traffic profiles defined by the second data fn+1, deciding on at least one adaptive behavior to be applied to at least one connected autonomous vehicle circulating in the upstream zone (11) of the traffic segment (10) or in the traffic segment itself (10); d6) sending at least one adaptive behavior to be applied, by wired or wireless communication means from the decision, selection and transmission module (40), to one or more transmission units (50); d7) sending a recommendation with at least one adaptive behavior to be applied, by one or more transmission units (50), via wireless communication means, to at least one connected autonomous vehicle (60) circulating in the area (11) upstream or in the traffic segment (10) identified in the second data fn+1 collected in sub-step d1); d8) collection of the first data dn+1, by means of the environmental sensors (20) installed on or near the delimited road traffic segment (10), the first data dn+1 relating to the first real-time parameters of the delimited road traffic segment (10) and consultation, by the decision, selection and transmission computer module (40), of the aforementioned weather conditions database; d9) sending the first data collected in sub-step d8), by each of the environmental sensors (20), to the decision, selection and transmission computer module (40), via wired or wireless means of communication; d10) the decision, selection and transmission computer module (40), based on the following input data: - first data received in real time dn+1, - first data received previously dn, makes a comparison between these two sets of first data dn+1 and dn, wherein, if this difference reveals a reduction in at least one of the negative externalities identified in step b), the module concludes on the effectiveness of the application of at least one adaptive behavior sent in sub-step d7) and wherein, if this difference reveals a failure to reduce at least one of the negative externalities identified in step b), the module concludes that the application of at least one adaptive behavior sent in sub-step d7) is ineffective and inserts an adjustment in at least one of the scenarios determined in step c), wherein the adjustment comprises a change in the correspondence between negative externalities and adaptive behaviors, resulting in at least one adjusted scenario; d11) storing the results of sub-step d10, i.e. at least one scenario considered effective and / or at least one adjusted scenario; wherein step d) is performed n times, where n is greater than or equal to one, preferably n is greater than 50, and wherein step d) is performed in a processing cycle lasting from 1 second to 60 minutes.

2. The computer-implemented method according to claim 1, characterized in that, in sub-step a1) of collecting first data, the aforementioned first parameters are at least one of the group consisting of pedestrian concentration rate, vehicle concentration rate, gas concentration rate, noise, safety level, congestion or a combination of them.

3. The computer-implemented method according to any of the preceding claims, characterized in that, in sub-step a4), the assignment of threshold values is made for different time periods, for example, the assignment of threshold values is made per day of the week, per hour of the day of the week, per season of the year, or a combination of these.

4. The computer-implemented method according to any of the preceding claims, characterized in that, in sub-step b5), the determination of priorities is based on at least one of proximity between the first data dn and the corresponding threshold values, period of the day, week, month, type of population residing nearby, type of buildings in the vicinity of the delimited traffic segment (10).

5. The computer-implemented method according to any of the preceding claims, characterized in that sub-step d7) comprises the following sub-steps: d71) obtaining confirmation that the autonomous vehicle (60) has carried out the recommendation sent in sub-step d7) in at least one of the ways belonging to the group consisting of: - collection of third data gn+1, by the traffic sensors (21) with subsequent sending, by the traffic sensors (21), of the third data gn+1 collected, to the computational decision, selection and transmission module (40); - the one or more autonomous vehicles (60) send a recommendation acceptance signal to the transmission unit (50), whereupon the transmission unit (50) sends these one or more acceptance signals to the decision, selection and transmission computational module (40); - storing, in a memory of the one or more autonomous vehicles (60), the acceptance of the recommendation received in sub-step d7); d72) processing, by the decision, selection and transmission computational module (40), of the third data gn+1 and / or the one or more recommendation acceptance signals received in sub-step d71), wherein the aforementioned processing of the third data gn+1 includes a comparative evaluation with other third data gn collected at previous time instants; wherein the third data gn, gn+1 include data from the one or more autonomous vehicles (60) that received one or more recommendations; and by in the comparison step d10), the third data gn+1 and / or the one or more recommendation acceptance signals sent by the one or more autonomous vehicles (60) are additionally included in the input data.

6. The computer-implemented method according to any one of the preceding claims, characterized in that, in sub-step d10) the effectiveness of the at least one applied adaptive behavior is concluded on the basis of at least one of: number of people present, dwell time of the people present, decrease in exposure to polluting gases such as NOx, PM, CO2, decrease in noise, increase in safety, decrease in congestion.

7. The computer-implemented method according to any of the preceding claims, characterized in that, in step c) of definition of scenarios for the delimited road traffic segment (10), the definition of scenarios is carried out for different periods of time, for example, the definition of scenarios is carried out by day of the week, by hour of the day of the week, by season, or by a combination of these.

8. The computer-implemented method according to claim 1, characterized in that in the sending sub-steps a2), b2), d2), d6), d9) the sending is performed by some wired or wireless communication means; and in that the sending sub-step d7) the sending is carried out by wireless means.

9. The computer-implemented method according to claim 1, characterized in that step c) is optionally performed by the decision, selection and transmission module (40).

10. A system configured to implement the computer-implemented method as defined in any one of claims 1 to 9, characterized in that it comprises: - a set of environmental sensors (20) installed on and / or near a delimited road traffic segment (10), wherein the set of environmental sensors (20) comprises at least one of the group consisting of video cameras, photographic cameras, pollutant gas concentration analyzers, sound level meters, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, sensing coils, thermometers, precipitation sensors, air humidity sensors, wind speed sensors, and combinations of them; - a set of traffic sensors (21) installed in an upstream area (11) of the aforementioned delimited road traffic segment (10), wherein the set of traffic sensors (21) comprises at least one of the group consisting of video cameras, photographic cameras, pneumatic tube counters, pulse counters, radars, lasers, magnetic plates, detection coils and combinations of these; - a control and evaluation computing module (30), wherein the computing module (30) comprises control and evaluation computing module communication means, control and evaluation computing module database storage means, and control and evaluation computing module processing means; - a decision, selection and transmission computing module (40), wherein a decision, selection and transmission computing module (40) comprises decision, selection and transmission computing module communication means, decision, selection and transmission computing module database storage means and decision, selection and transmission computing module processing means; - at least one connected and / or autonomous vehicle CAV (60) which is configured to drive partially or fully self-controlled in the delimited road traffic segment (10), in its vicinity or in the upstream area (11) of the aforementioned traffic segment (10) and comprises communication means suitable for receiving recommendations from a transmission unit (50); - at least one transmission unit (50) configured to communicate with the at least one connected and / or autonomous vehicle CAV (60); - optionally, at least one conventional vehicle CV (70) which is configured to circulate with a driver and without being partially or totally self-controlled in the delimited road traffic segment (10), in its vicinity or in the upstream area (11) of the aforementioned traffic segment (10); wherein the control and evaluation computational module (30) and the decision, selection and transmission computational module (40) are included in the same or in different computational units.

11. The system according to the preceding claim, characterized in that the aforementioned control and evaluation computational module (30) and the aforementioned decision, selection and transmission computational module (40) are selected from the group consisting of a local computer, a remote computer, a server, a tablet, a smart phone, such as a smartphone.

12. A computer program characterized in that it comprises logic instructions to enable the control and evaluation computer module (30), comprised in the system as defined in any one of claims 10 and 11, to perform the steps of the method as defined in any one of claims 1 to 9.

13. A computer program characterized in that it comprises logic instructions to enable the decision, selection and transmission computational module (40), comprised in the system as defined in any one of claims 10 and 11, to perform the steps of the method as defined in any one of claims 1 to 9.

14. A computer-readable medium, characterized in that it comprises the installation of the computer program as defined in any one of claims 12 to 13.

Images