Classifications

• • G—PHYSICS
  • G08—SIGNALLING
  • G08G—TRAFFIC CONTROL SYSTEMS
  • G08G1/00—Traffic control systems for road vehicles
  • G08G1/01—Detecting movement of traffic to be counted or controlled
  • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
  • G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
  • G08G1/0112—Measuring and analyzing of parameters relative to traffic conditions based on the source of data from the vehicle, e.g. floating car data [FCD]
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
  • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
  • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
  • G01C21/34—Route searching; Route guidance
  • G01C21/3453—Special cost functions, i.e. other than distance or default speed limit of road segments
  • G01C21/3492—Special cost functions, i.e. other than distance or default speed limit of road segments employing speed data or traffic data, e.g. real-time or historical
• • G—PHYSICS
  • G08—SIGNALLING
  • G08G—TRAFFIC CONTROL SYSTEMS
  • G08G1/00—Traffic control systems for road vehicles
  • G08G1/01—Detecting movement of traffic to be counted or controlled
  • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
  • G08G1/0125—Traffic data processing
  • G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
• • G—PHYSICS
  • G08—SIGNALLING
  • G08G—TRAFFIC CONTROL SYSTEMS
  • G08G1/00—Traffic control systems for road vehicles
  • G08G1/01—Detecting movement of traffic to be counted or controlled
  • G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed

Definitions

• the present invention relates to the prediction of the speed of a vehicle on a road network.
• Atmospheric pollution is also a major financial stake: a senatorial commission of inquiry estimates that the total cost of air pollution is between 68 and 97 billion euros per year for France, in an assessment made in July 2015, integrating both the health damage of pollution, but also its consequences on buildings, ecosystems and agriculture.
• the transport sector is still one of the most important sources of pollutants, despite the many measures put in place by the public authorities and technological advancements in the field. All modes of transport are responsible for around 50% of global nitrogen oxide (NOx) emissions and around 10% of PM2.5 particulate emissions.
• NOx global nitrogen oxide
• the current air quality monitoring tools do not make it possible to isolate and estimate with precision the share of emissions in real use due to road transport as well as their spatial location.
• the estimation of pollutant emissions is based on the use of an average method adapted to large scales, typically road segments of several kilometers so that the journey can be considered as representative of all traffic conditions, as in the COPERT methodology (standing for “COmputer Program to calculate Emissions from Road Transports”, which can be translated into a computer program for calculating emissions from road transport).
• Polluting emissions (chemical and / or sound) are linked to the speed of movement of vehicles on the road. This is why in order to have good forecasts of polluting emissions, it is important to accurately predict the speed of vehicles on the road, taking into account the topology of the road (slope, bend, road signs, etc.) and the conditions. traffic.
• the consumption of the vehicle is also linked to the speed of the vehicle. This is why, to accurately determine a vehicle's consumption, it is important to accurately predict the speed of vehicles on the road, taking into account the topology of the road (slope, bend, road signs, etc. ) and traffic conditions.
• Another area in which prediction of vehicle speed is useful is the area of determining vehicle routes for navigation. Indeed, an accurate prediction of the speed of the vehicles on the road, taking into account in particular the topology of the road and the traffic conditions, allows optimized navigation, in particular in terms of travel time.
• Another method is based on an estimate of a statistical speed corresponding to the 85% percentile using statistical models.
• This method has the same drawbacks as the previously described method: lack of precision, no consideration of the impact of the road infrastructure, nor of the different driving styles, nor of the different behaviors linked to the signs, impossibility of predicting a speed for a road segment without traffic measurement.
• FCD data from English "floating car data"
• FCD data from English "floating car data”
• This database is broken down then grouped according to macroscopic descriptors, such as the type of road. Then each road segment is identified as belonging to a group.
• One or more speed profiles are built on this segment by combining portions of real speeds from the FCD data belonging to this group.
• This method also lacks precision; in fact, the models used are only relevant on a large spatial scale, but can give rise to inconsistent behavior.
• this method which can be computationally expensive, does not allow the impact of road infrastructure to be taken into account directly and in detail.
• Another method is based on the calculation of a speed profile as a function of distance by means of mathematical functions by considering the signaling and the infrastructure on each road segment. This method is not satisfactory in terms of taking into account of different behaviors and driving styles, nor of the topology of the road. In addition, the majority of the methods based on this method cannot be used for a road segment for which there is no history.
• the document Andrieu, C. (2013). Functional modeling of speed profiles in connection with the infrastructure and methodology of construction of an aggregated profile (Doctoral dissertation, University Paul Sabatier-Toulouse III) describes such a method.
• the present invention aims to predict an accurate speed profile at a fine spatial scale by considering the different behaviors and styles of driving without microscopic data.
• the invention relates to a method for predicting at least one speed profile of a vehicle for a portion of a road network, in which a vehicle speed model is constructed by means of macroscopic data from the network. road and trip data, then this model is applied to the portion of the road network considered.
• the invention relates to a method for predicting at least one profile of the speed of a vehicle on a portion of a road network. For this process, the following steps are implemented:
• a vehicle speed model is constructed by a machine learning method by means of macroscopic data of said road network and by means of data from journeys made on said road network, said vehicle speed model associates with at least one subdivision of said road network at least one profile vehicle speed as a function of said macroscopic data of said road network and of said data of journeys made;
• At least one profile of the speed of said vehicle on said portion of said road network is predicted by applying said vehicle speed model to macroscopic data of each subdivision of said portion of said road network.
• said vehicle speed model is constructed by implementing the following steps:
• Each segment of said road network is categorized as a function of said macroscopic data of said road network
• At least one vehicle speed profile is generated by means of said data of journeys made.
• At least one vehicle speed profile is predicted on said portion of said road network by implementing the following steps:
• said at least one speed profile is assigned to each segment of said portion of said road network, taking into account data from journeys made on each segment in order to specify said at least one speed profile.
• each segment of said portion of said road network is also assigned a distribution of said at least one speed profile.
• said road network is segmented by dividing said road network into triplets of links, each triplet of links being formed by a link formed between two nodes of said road network, its origin and its destination.
• said category of said segment of the road network is chosen from:
• said data of journeys made by a classification algorithm in particular the k-means algorithm, is classified.
• At least one speed profile of the vehicle is generated by a method based on at least one neural network, in order to parameterize a function of the speed depending on the distance, preferably said function of the speed is a linear function, a parabolic function, or a combination of linear and / or parabolic functions of distance.
• said macroscopic data of said road network are the topology and traffic conditions, preferably said macroscopic data of said road network are provided by a geographic information system.
• said data of journeys made comprises data of speed, position and altitude measured during previous journeys, preferably by means of a geolocation system.
• said at least one vehicle speed profile of said portion of the road network is displayed on a road map, preferably by means of a smart phone or a computer system.
• the invention relates to a method for predicting chemical and / or sound polluting emissions on a portion of a road network, in which the following steps are implemented:
• a microscopic model of chemical and / or sound polluting emissions is applied to said at least one speed profile in order to predict said pollutant emissions, said model relating the speed of the vehicle and said pollutant emissions.
• the invention also relates to a method for predicting the consumption of a vehicle on a portion of a road network, in which the following steps are implemented: a) at least one speed profile of the vehicle is predicted on said portion of said road network by means of the method of predicting at least one speed profile according to one of the preceding characteristics; and b) A vehicle consumption model is applied to said at least one speed profile in order to predict said consumption of said vehicle, said model relating the speed of the vehicle and said consumption of said vehicle.
• the invention relates to a method for determining a route to be taken by a vehicle, for which the departure and arrival of said route are identified, by implementing the following steps:
• a route to be covered is determined to connect said departure and said arrival, by taking into account said at least one speed profile of the vehicle, preferably by minimizing the journey time.
• Figure 1 illustrates the steps of the method according to one embodiment of the invention.
• Figure 2 illustrates the construction of the speed model according to one embodiment of the invention.
• Figure 3 illustrates the prediction of a speed profile according to one embodiment of the invention.
• Figure 4 illustrates a portion of a road network.
• Figure 5 illustrates a speed profile for an example in the case of a road with a traffic light, when the traffic light is red.
• Figure 6 illustrates a speed profile for an example in the case of a road with a traffic light, when the traffic light is green.
• FIG. 7 illustrates, for an example, a comparison of measured speed profiles and predicted speed profiles by means of the method according to one embodiment of the invention, in the case of a road with a traffic light, when the traffic light is red.
• FIG. 8 illustrates, for an example, a comparison of measured speed profiles and of speed profiles predicted by means of the method according to an embodiment of the invention, in the case of a road with a traffic light, when the traffic light is green.
• FIG. 9 illustrates, for an example, a comparison of the NOx emissions measured and estimated by means of the method according to an embodiment of the invention.
• FIG. 10 illustrates, for an example, a comparison of the NOx emissions measured and estimated by means of the method according to an embodiment of the invention.
• the present invention relates to a method for predicting at least one speed profile of a vehicle on a portion of a road network.
• the method makes it possible to predict the speed of a vehicle traveling on a portion of a road network, the speed can be expressed as a function of the distance from an end of the road. Since this is a prediction, it can be performed even on a portion of a road network for which no previous route is available.
• speed profile the variation of the speed of the vehicle along a road of a road network ;. the road network being made up of all the roads for a given territory, for example for a country or for a region.
• the speed profile is dynamic (unlike an average speed).
• the road network portion is a part of this road network for which it is desired to determine at least one vehicle speed profile.
• the portion of road network can consist of a set of roads between a starting point and an end, a set of roads in a city or a neighborhood, etc.
• the vehicle is a motorized vehicle traveling within the road network, such as a motor vehicle, a two-wheeler, a heavy vehicle, a coach, a bus.
• Step 1 can be done offline, and Step 2 can be done online. These steps are detailed in the remainder of the description.
• FIG. 1 illustrates, schematically and in a non-limiting manner, the steps of the method of predicting at least one speed profile according to one embodiment of the invention.
• MAC macroscopic data
• DTR data on journeys made
• the vehicle speed model (MOD) is constructed. Then, the vehicle speed model (MOD) allows the prediction (PRED) of at least one speed profile (v) for the portion of the road network considered (POR).
• a vehicle speed model is constructed by a machine learning method using macroscopic data from the road network and using data from journeys made on the road network.
• the macroscopic data of the road network make it possible to take into account information related to the road network, such as infrastructure, slope, signage, traffic etc.
• the data of journeys made makes it possible to take into account real behavior in order to form a representative and precise vehicle speed model.
• the vehicle speed model associates with at least one subdivision of the road network (preferably with a link of the road network) at least one vehicle speed profile as a function of the macroscopic data of the road network and of the data of journeys made.
• any division of the road network can be a road network link.
• the road network link is an elementary subdivision of the road network between two consecutive nodes of the road network.
• a road network link can be a road between two consecutive intersections, between two consecutive signs, between an intersection and a sign, or part of a motorway between two consecutive exits, etc.
• the macroscopic data of the road network can be topology (i.e., slope, turns, intersections, signage, etc.) and traffic conditions.
• the macroscopic data of the road network can be provided by a geographic information system (GIS).
• GIS geographic information system
• Maps TM, Google Maps TM, OpenStreetMap TM are examples of geographic information systems. Macroscopic data is always available and from any location. Thus, they can serve as unique inputs to the vehicle speed model.
• the data of journeys made may include data measured during previous journeys, in particular speed, position and altitude.
• the data of journeys made can be measured by means of a geolocation system, such as a satellite positioning sensor, such as the GPS system (standing for Global Positioning System), the Galileo system, etc. .
• the geolocation system can be on board the vehicle or deported (for example by means of a smart phone, English "smartphone").
• the vehicle speed model can be constructed by implementing the following steps:
• FIG. 2 illustrates, schematically and in a non-limiting manner, the steps in the construction of the vehicle speed model according to this embodiment of the invention.
• SEG segment
• MAC data macroscopic
• CAT classification step
• DTR trip data
• CAT category of road segment
• PRO speed profile
• the road network is segmented using macroscopic data from the road network.
• the road network is divided into segments based on the macroscopic data of the road network.
• the purpose of this step is to obtain subdivisions of the road network based on data such as topology and traffic conditions.
• the road network can be segmented by dividing the road network into link triplets, each link triplet comprising a link formed between two nodes of the road network, its origin and its destination.
• link triplets comprising a link formed between two nodes of the road network, its origin and its destination.
• This triplet segmentation of links makes it possible to limit this dispersion, by considering (in the following steps) only the speed data of the vehicles which come from the same origin and which have the same destination.
• this segmentation makes it possible to obtain characteristics for each segment (each triplet of links), for example the maneuver angle, the number of triplets having the same central link (number of connections), etc.
• Figure 4 illustrates a road including an intersection.
• This route includes a link between nodes A and B.
• the vehicle then has a single possible origin O and two possible destinations D1 and D2.
• a first segment corresponding to the route of FIG. 4 may be the triplet of links (O, link AB, D1)
• the second segment corresponding to the route of Figure 4 can be the triplet of bonds (O, bond AB, D2).
• the central link AB is then common for two distinct segments (triplets of links).
• the road can be segmented on the basis of the road network links, or by considering half of a link to capture the effect of a signage that may be in the middle of a segment defined by the macroscopic data of a geographic information system.
• each segment of the road network obtained in step 1 .1 is categorized using the macroscopic data of the road network.
• each road segment is associated with a category that groups together road segments with the same characteristics.
• the segment is a link triplet
• two link triplets having the same central link can be found in different categories. Indeed, they can have different characteristics.
• the bond triplet (O, AB, D1) does not have a bend, unlike the bond triplet (O, AB, D2).
• the categories can be formed from the following criteria: road congested or not, presence or absence of signaling (for example of a traffic light), presence or absence of an intersection, priority road or not , importance of the curvature of a bend, functional class (which characterizes the hierarchy of the road network and the level of importance of the segment, for example motorway, small street, etc.), number of lanes, etc. These criteria are obtained directly from the macroscopic data of the road network.
• the categories of segments can be:
• each segment can be assigned one of these six categories. Indeed, it is generally not useful to subdivide the case of the road congested, because in this case the speed is very low, and neither the signage nor the curvature of the road have a significant impact on the speed of the vehicle.
• the data of the journeys made is classified. To do this, we associate with each segment of the road network the data of journeys made, in particular speed. Then, for each category, the data of the trips made which are similar are classified. This step makes it possible to limit the dispersions of the measured data, in particular of the speed, these dispersions being in particular induced by random phenomena (driving style, alternation of traffic lights, etc.).
• This classification can be made from data (descriptors) such as the average speed on the segment, the speed of the 75% percentile, minimum / maximum speed, sum of positive / negative accelerations, etc.
• the classification can be carried out by a "k-means” algorithm, because the data used is digital.
• the number of "k” classes is a parameter of the algorithm, which is determined with an iterative method aimed at maximizing a dissimilarity metric such as "silhouette".
• An important advantage of this method is the evaluation of the proximity of a data sample (in this case a speed profile resulting from the data of journeys made) to the center of a classification by also comparing it with the minimum average distance d 'another class. In general, a figure value greater than 0.5 indicates good classification, with very little confusion and dispersion between classes.
• step 1 .2 For each category of road segment (step 1 .2) and for each classification obtained in the previous step, at least one speed profile is generated by means of data from previous journeys.
• the speed profile is dynamic and that it corresponds to a variation in speed as a function of distance within the same portion of the road segment.
• a speed profile is generated on this link which approximates the data of the journeys made on this link.
• the aim of this step is to represent by a speed profile the typical behavior of the vehicles according to the characteristics of the road and the previous journeys made.
• the speed profiles are representative of real behavior.
• this step can consist in generating a function of the speed depending on the distance on the link considered.
• the speed function can be configured with the data of the journeys made.
• the function of the speed can be a polynomial function.
• the speed function can be a linear function, a parabolic function, or a combination of linear and / or parabolic functions.
• each category of segment of the road network can also be assigned a distribution of speed profiles.
• speed profiles it is possible to predict a probability of the speed profile.
• At least one vehicle speed profile can be generated by a neural network method, a support vector machine method (standing for “Support Vector Machine”), a method of decision tree forests (from the English "Random Forest”), or other methods of supervised learning.
• the neural network method makes it possible to parameterize a function of the speed which depends on the distance, this function being able to be a linear function, a parabolic function, or a combination of linear and / or parabolic functions. An example of this embodiment is detailed in the remainder of the description.
• these parameters useful for the generation of the speed profile can be the initial and final speeds of the profile on the segment considered, as well as its maximum / minimum speed and the position of the possible stopping point. Learning of these parameters can be carried out in a supervised manner (thanks to the data of the journeys made) in order to correlate them directly with macroscopic descriptors.
• the supervised learning tool used can be a neural network, which can use the following macroscopic descriptors as input: classification (from the previous step) of membership of the speed profile to be estimated, functional class of the connections of the triplet , number of channels on the triplet links, speed limit on the triplet links, average traffic speed on the triplet links, length of the triplet links, maneuver angle at the entry and exit of the central link of the triplet, number incoming / outgoing links of the central link, etc.
• classification from the previous step
• functional class of the connections of the triplet a neural network
• number of channels on the triplet links speed limit on the triplet links
• average traffic speed on the triplet links length of the triplet links
• maneuver angle at the entry and exit of the central link of the triplet number incoming / outgoing links of the central link, etc.
• the method can be split into two stages with cascade neural networks:
• the first neural network can estimate the average of the initial speed and the final speed, as well as their standard deviation, to obtain a Gaussian probability density.
• the Gaussian probability density was chosen for the simplicity of definition with few variables and for the good representativeness of the phenomenon.
• This first neural network can be common to all the classes determined in the previous step.
• the following neural networks can estimate the maximum and minimum speed as well as the position of the stopping point.
• This neural network depends on the membership class of the profile to be estimated (for example a neural network which estimates the breakpoint is used if the class foresees a profile with a breakpoint) and takes as input the estimation of the initial speed and the final speed performed by the preceding neural network.
• the parameters estimated with deterministic, probabilistic, or other polynomial methods can be used, without loss of generality.
• the polynomial functions used to generate the predicted speed profiles can be inspired by observing the actual profiles of the trip data of each class. They can mainly be reconstructed profiles with linear or parabolic functions.
• the identified parameters can be “drawn” at random according to their Gaussian distribution in order to generate several representative speed profiles. These generated speed profiles can meet the constraints of maximum and minimum length and speed of the link considered for which the prediction is made.
• FIG. 5 illustrates, schematically and in a nonlimiting manner, a function of the speed V as a function of the distance D.
• This speed function corresponds to a road link with a traffic light in the “red” state.
• the speed function is made up of two parabolic functions: a first decreasing until a stop point and a second increasing from the stop point.
• FIG. 6 illustrates, schematically and in a non-limiting manner, a function of the speed V as a function of the distance D.
• This speed function corresponds to a road link with a traffic light in the "green" state.
• the function of the speed is formed by a decreasing linear function.
• At least one profile of the speed of the vehicle is predicted on the portion of the road network considered. It is recalled that the speed profile is dynamic and that it corresponds to a variation in speed as a function of the distance within the portion of the road network considered. It may be a portion of the road network that has been traveled by previous trips made, a portion of the road network that has been partially covered by previous trips made, or a portion of the road network that has not been completed. not been traveled by previous journeys made (it can even be a portion of a non-existent road network, for which you want to predict the speed profile).
• we apply the speed model of the vehicle built in step 1 to the macroscopic data of the portion of the road network considered.
• the topological data of the portion of the road network considered is taken into account.
• at least one speed profile is assigned to each subdivision (preferably to each link) of the portion of the road network considered.
• a plurality of speed profiles is determined for each link of the portion of the road network considered.
• the plurality of speed profiles can be obtained in different ways, in particular according to the embodiments implemented.
• the plurality of speed profiles for each link can come from the fact that for each category of road segment, a plurality of speed profiles are generated (step 1 .4), each speed profile corresponding to a behavior or a driving style.
• the plurality of speed profiles for each link may arise from the fact that each link may belong to more than one link triplet, the link triplets may belong to distinct categories.
• the plurality of speed profiles may be from a plurality of random draws among the determined speed profile distribution.
• the prediction of the speed profiles can implement the following steps:
• FIG. 3 illustrates, schematically and in a non-limiting manner, the steps of this embodiment.
• the portion of the road network (POR) is segmented (SEG).
• SEG segmented
• CAT categorization
• MOD vehicle speed model
• CAT categorization
• each subdivision of the road network portion considered is assigned (ATT) at least one speed profile (v).
• the portion of the road network considered is segmented.
• the portion of the road network considered can be segmented in the same way as the segmentation implemented in step 1.1.
• the portion of the road network considered can be segmented by link triplets, comprising an origin, a central link and a destination.
• the segments of the road network portion considered are categorized.
• the segments of the road network portion considered can be categorized in the same way as the categorization provided in step 1 .2.
• the segments of the portion of the road network considered can be categorized into the following six categories:
• each segment of the road network portion considered is assigned at least one speed profile generated by means of the vehicle speed model, as a function of the categorization of the road network portion.
• the segment of the road network portion considered has a speed profile identical to the speed profile of the segment having the same category in the vehicle speed model.
• a segment of the portion of the road network considered which is a road with no or little congestion with a traffic light can have at least one speed profile as shown in figure 5 when the light is red, and at least one speed profile as shown in figure 6 when the light is green.
• the speed profile can be matched with the data of completed journeys, so as to optimize the precision of the prediction of the speed profile.
• the method may include an optional step of displaying the speed profile for the portion of the road network considered.
• This display can take the form of a note or a color code. If necessary, a note or a color can be associated with each link of the road network.
• This display can be produced on board a vehicle: on the dashboard, on an autonomous portable device, such as a geolocation device (of the GPS type), a mobile telephone (of the smart phone type). It is also possible to display the speed profile on a website.
• the predicted speed profile can be shared with the public authorities (for example road network manager) and construction companies. public. Thus, public authorities and public works companies can optimize the road infrastructure, to improve safety or polluting emissions.
• instantaneous measurements of the speed of at least one vehicle traveling on the road network can be made, in particular by means of a geolocation system (for example: GPS, smart phone) or by fewer vehicles connected (for example: with a sensor placed on the vehicle's OBD diagnostic socket).
• the instantaneous speed data measured in real time during the journey can then be used to enrich and possibly reset the prediction of the speed profiles, if necessary during step 2.3.
• the predicted speed profiles are representative of the driving conditions in real time.
• the prediction of the associated indicators is representative of driving conditions in real time.
• the present invention also relates to a method for predicting polluting, chemical (for example NOx, particles) and / or sound emissions on a portion of a road network.
• a method for predicting polluting emissions the following steps are implemented: a) at least one vehicle speed profile is predicted on the portion of the road network considered by means of the method for predicting at least one speed profile according to any one of the variants or combinations of variants described above; and
• a microscopic model of pollutant, chemical and / or sound emissions is applied to the predicted speed profile in order to predict the pollutant emissions on the portion of the road network considered, the pollutant emissions model being a model which relates the speed of the vehicle with polluting emissions.
• the method can include an optional step of displaying the polluting emissions for the portion of the road network considered.
• the polluting emissions can be displayed on a road map.
• This display can take the form of a note or a color code. If necessary, a note or a color can be associated with each link of the road network.
• This display can be produced on board a vehicle: on the dashboard, on an autonomous portable device, such as a geolocation device (of the GPS type), a mobile telephone (of the smart phone type). It is also possible to display polluting emissions on a website.
• the predicted pollutant emissions can be shared with the public authorities (eg road manager) and public works companies. Thus, public authorities and public works companies can optimize road infrastructure to improve polluting emissions.
• the present invention relates to a method for predicting the consumption of a vehicle, on a portion of a road network.
• this method of predicting the consumption of a vehicle the following steps are implemented: a) at least one vehicle speed profile is predicted on the portion of the road network considered by means of the prediction method of at least a speed profile according to any of the variants or combinations of variants described above; and
• a vehicle consumption model is applied to the predicted speed profile in order to predict the vehicle consumption on the portion of the road network considered, the vehicle consumption model being a model which relates the vehicle speed to its consumption.
• the method can include an optional step of displaying the consumption for the portion of the road network considered.
• the consumption can be displayed on a road map.
• This display can take the form of a note or a color code. If necessary, a note or a color can be associated with each link of the road network.
• This display can be produced on board a vehicle: on the dashboard, on an autonomous portable device, such as a geolocation device (of the GPS type), a mobile telephone (of the smart phone type). It is also possible to display the vehicle's consumption on a website.
• the vehicle's consumption can be shared with the public authorities (for example road network manager) and public works companies. Thus, public authorities and public works companies can optimize the road infrastructure, the location of gas stations, charging stations, etc.
• the invention relates to a method for determining a route to be covered by a user, for which the departure and arrival are identified, by implementing the following steps: a) at least one speed profile is predicted vehicle on the portion of the road network considered by means of the method of predicting at least one speed profile according to any one of the variants or combinations of variants described above,
• a route to be covered is determined to connect the start and the finish, taking into account the predicted speed profile.
• Step b) can minimize conventional criteria of navigation processes: such as travel time, distance traveled, energy consumed, etc.
• step b) can minimize the associated risk thanks to the associated probability distribution. These minimization criteria are dependent on the speed of the vehicle. Consequently, the precision obtained by the speed profile prediction method allows optimization of the determination of the path to be traveled.
• step b we can use a shorter path algorithm.
• the method can include an optional step of displaying the route to be traveled, optionally with the speed profile for each link of the route.
• the trip can be displayed on a road map.
• This display can be carried out on board a vehicle: on the dashboard, on an autonomous portable device, such as a geolocation device (GPS type), a mobile phone (smart phone type). It is also possible to display the route to be covered by the vehicle on a website. In addition, the route to be covered by the vehicle can be shared with a vehicle fleet manager.
• the emissions associated with the actual speed profiles of the data of journeys made and calculated with a microscopic model of emissions were used as a reference, thus obtaining a reference per classify (in accordance with to the classes defined in step 1.3).
• the emissions associated with the speed profiles generated by classification were compared to their reference.
• the first example was performed on a segment (link) belonging to the learning road network (Paris / Lyon), but which was not used directly in the learning of neural networks to generate the vehicle speed model. (step 1.4).
• a link with little or no congestion with a traffic light at the end of the link was chosen for the prediction of speed profiles.
• the neural networks were used. to estimate the parameters of the speed profiles (initial speed, final speed, stopping point, maximum speed) as a function of the macroscopic data of the link considered.
• FIG. 7 illustrates profiles of the speed V (km / h) as a function of the distance D (m), in the case of a link with a red traffic light.
• FIG. 8 illustrates profiles of the speed V (km / h) as a function of the distance D (m), in the case of a link with a green traffic light.
• the speed profiles illustrated correspond to the measured MES speed profiles and to the predicted speed profiles PRED by means of the method according to the invention. From a qualitative point of view, the predicted speed profiles PRED reproduce well the shape real MES profiles (acceleration level, speed, position of the stopping point) in these two situations.
• FIG. 9 illustrates for each case, the distribution of NOx emissions in mg / km (area in gray) for the MES measured speeds, and for the PRED predicted speeds by means of the method according to the invention. On each curve, the horizontal line AVG indicates the average value of the NOx emissions.
• the average obtained with the predicted speed values PRED are close to the average obtained with the measured MES speed values.
• a statistical analysis of the impact of the accuracy of the speed profile prediction on emissions was also carried out.
• the average absolute error on emissions is 10 mg / km with a percentage error of 6%.
• the second example was carried out on a road segment (link) in Marseille not belonging to the learning road network, and therefore which was not used in the learning of neural networks.
• a link with a traffic light at the end of the link was chosen for the prediction of speed profiles. This time the link does not belong to the data of the learning network in order to verify the ability of the invention to extrapolate and generalize information (data).
• the neural networks were used to estimate the parameters of the profiles (speed initial, final speed, stopping point, maximum speed) according to the macroscopic data of the link.
• FIG. 10 illustrates for each case, the distribution of the NOx emissions in mg / km (area in gray) for the measured MES speeds, and for the predicted speeds PRED by means of the method according to the invention. On each curve, the horizontal line AVG indicates the average value of the NOx emissions.
• the average obtained with the predicted speed values PRED are close to the average obtained with the measured MES speed values.
• a statistical analysis of the impact of the accuracy of the speed profile prediction on emissions was also carried out.
• the absolute average error on emissions is 37.8 mg / km with a percentage error of 7.5%.

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Engineering & Computer Science (AREA)
• Radar, Positioning & Navigation (AREA)
• Remote Sensing (AREA)
• Chemical & Material Sciences (AREA)
• Analytical Chemistry (AREA)
• Automation & Control Theory (AREA)
• Traffic Control Systems (AREA)
• Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=68281558

Family Applications (1)

Country Status (4)

Families Citing this family (6)

Family Cites Families (8)

• 2019
  • 2019-05-28 FR FR1905686A patent/FR3096822B1/fr active Active
• 2020
  • 2020-05-18 EP EP20725566.2A patent/EP4026110A1/de active Pending
  • 2020-05-18 WO PCT/EP2020/063828 patent/WO2020239503A1/fr unknown
  • 2020-05-18 US US17/610,694 patent/US20220215749A1/en active Pending

Also Published As

Similar Documents

Legal Events