WO2009116105A2 - Procédé d'attribution de trafic pour des réseaux de transport multimodaux - Google Patents

Procédé d'attribution de trafic pour des réseaux de transport multimodaux Download PDF

Info

Publication number
WO2009116105A2
WO2009116105A2 PCT/IT2008/000194 IT2008000194W WO2009116105A2 WO 2009116105 A2 WO2009116105 A2 WO 2009116105A2 IT 2008000194 W IT2008000194 W IT 2008000194W WO 2009116105 A2 WO2009116105 A2 WO 2009116105A2
Authority
WO
WIPO (PCT)
Prior art keywords
transportation
traffic
disaggregate
route
aggregate
Prior art date
Application number
PCT/IT2008/000194
Other languages
English (en)
Other versions
WO2009116105A3 (fr
Inventor
Gianfranco Antonini
Sergio Sirgi
Original Assignee
Gianfranco Antonini
Sergio Sirgi
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Gianfranco Antonini, Sergio Sirgi filed Critical Gianfranco Antonini
Priority to PCT/IT2008/000194 priority Critical patent/WO2009116105A2/fr
Publication of WO2009116105A2 publication Critical patent/WO2009116105A2/fr
Publication of WO2009116105A3 publication Critical patent/WO2009116105A3/fr

Links

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"

Definitions

  • the present invention relates in general to a method for traffic assignment and, more particularly, to a probabilistic traffic assignment method for multimodal transportation networks .
  • known methods are substantially based on theorems and algorithms proper to the theory of graphs.
  • the traffic flow assignment does usually combine the algorithms for identifying the shortest route with other particular algorithms, such as the Ford-Fulkerson algorithm.
  • sections of the transportation network become saturated by traffic, and therefore excluded during the following iterations from the utilisable part of the system.
  • the maximum capacity of each section of the network is progressively achieved, until, by using any of the available routes (up to all the routes, if necessary), the given demands for travel are fully met.
  • the present invention stems from the observation made by the Applicant that in known traffic assignment methods the choice of "the most convenient route" is not defined prior to starting the trip, but during the simulated trip, depending on the simulated traffic conditions encountered by the "virtual user" on the way towards its destination. Therefore, in most of such assignment methods, the user's route results from a progressive and unpredictable adjustment of traffic streams to arbitrarily simulated accidental events rather than from the user's rational choice. Moreover, most of such methods are biased by arbitrary criteria concerning, for instance, priority to be given in the assignment to certain stream categories, demand stocks and so forth. Therefore, it is not appropriate to state that these methods account for the user' s "most convenient" choice.
  • the sequence of links which connect any trip origin to a destination, includes usually a very high number of possible routes. However, many of such routes are either unlikely or never used in practice. But unlikely or quite unreasonable routes are involved by known assignment simulation methods when, on the one hand, the simulation saturates all the likely routes, while, on the other hand, the given demand for travel must be met by the simulation, but the saturated likely routes can no more cope with the demand.
  • the saturation occurs when the volume of the traffic stream achieves the level of the route capacity.
  • traffic congestion is a frequent occurrence, the known assignment methods are not able to simulate processes of traffic congestion, which instead arise when the loads of the traffic flows exceed the route saturation levels.
  • the Applicant has carried out an in- depth study in order to develop a new traffic assignment methodology which is more realistic and reliable than the known ones, which are, as previously said, inadequate to simulate the real conditions that determine the behaviour of users of transportation networks .
  • the objective of the present invention is, then, to provide a methodology of the aforesaid type.
  • This objective is achieved by the present invention in that it relates to a traffic assignment method, and a software program configured to implement the method, as defined in the appended claims.
  • this objective is achieved by the present invention in that it relies on a probabilistic assessment of the factors that may influence the decisions as to the choice of the most convenient routes by the users of a given transportation network, unlike known traffic assignment methods which adopt a deterministic approach.
  • the simple principle, on which the present invention is based, is that any user of a transportation network makes trips after choosing the routes that are subjectively considered as the most convenient ones.
  • An important feature of present invention is that it accounts for preliminary assessments, as supposedly made by users prior to undertaking their travel, concerning the traffic conditions relevant to a selected set of routes, which are considered as the only reasonable alternative routes for each possible travel.
  • preliminary assessments are not made by separate categories of users and allow for the users' expectations relative to aggregate traffic streams. Nevertheless, preliminary assessments of the traffic conditions are often wrong and traffic congestions are not avoided, while it's also realistic not to exclude that part of the users don't reject the risk of "acceptable levels" of congestion during their travel. Therefore, the model provides a consistent logical way to simulate the overcoming of traffic congestions, as these emerge during the simulation exercises .
  • a major feature of the present invention consists of avoiding the adoption of questionable "optimisation criteria" , such as minimisation of travel cost, of route length or trip duration, etc.
  • optimisation criteria such as minimisation of travel cost, of route length or trip duration, etc.
  • the mathematical structure of the present method exploits the amount of probabilistic uncertainty that is always associated with any distribution of probabilities, such as the unknown distribution of trip probabilities to be determined.
  • Figure 2 shows a graph representing traffic flow as a function of average flow speed relating to road links
  • Figure 3 shows a graph of traffic flow as a function of the average flow density relating to road links ;
  • Figure 4 shows a graph of exceeding traffic flow densities in a congestion state in a road link.
  • the present invention is implemented by means of a software program, loadable in a memory of an electronic processor, and comprising software code portions for implementing, when the software program is run on the processor, the traffic assignment method described hereinafter.
  • the proper system referred to by the present invention consists of multi-mode transportation links between different cities or regions. Nevertheless, making particular simplifications and assumptions, the present invention may be fit also for applications to intra-urban transportation networks .
  • the present invention allows for possible cases of traffic congestion in some sections, i.e. road links, of the transportation network, and provides logical tools to overcome the relevant problems that may arise during assignment simulations.
  • a graph is a geometrical figure consisting of a set of scattered points connected with each other by linear segments named "links" .
  • the points of the graph are distinguished according to the representation of two different functions.
  • the points of the graph that represent origins and/or destinations of traffic streams are marked as poles.
  • the points of the graph that indicate only link junctions are marked as nodes .
  • Every pole since it is always at a link junction or at a link-end, is also a node; whereas a simple node, which is merely a link junction, is never meant to be a pole.
  • an isolated node at one end of a single link can only be a pole.
  • Any transportation infrastructure connecting pairs of nodes is represented by a link of the graph.
  • Each pair of nodes may be connected by more than one type of infrastructure and, therefore, by more than one link.
  • each link consists of oriented lines, which represent the traffic canals that convey the traffic flows.
  • the line orientation indicates the flow direction allowed by the relevant canal.
  • route is meant any sequence of compatibly oriented and inter-connected links between two poles .
  • travelling is here used to mean a given set of routes between any pair of poles.
  • Nodes, poles, routes and travels are the main components of a transportation network.
  • N(N-I) is the maximum number of possible travels between different poles of the same network.
  • the analysis assumes in general that the travel from Pole A to Pole B differs from the travel from B to A. Moreover, for the sake of simplicity, the travels internal to each pole (i.e., for example, from A to A itself) are not allowed for in this text. Such an omission doesn't affect at all the logics of the theory expounded here. These "internal travels" could easily be introduced into the model whenever the case should occur. Nodes and poles of . a transportation network are here addressed as components different from the links of the network .
  • the present invention aims at formulating and solving the following problems:
  • the degree of users concentration in a particular route may bring the traffic flow (which is expressed in terms of transportation units moving per time unit toward their destination) to drop to a level lower (or much lower) than the maximum allowed by the physical capacity of the route .
  • the maximum traffic flow allowed by the physical capacity of any given route does not necessarily correspond to the optimum flow expected by the users.
  • the maximum flow in fact, is normally not associated with the optimum trip speed/duration planned by the users .
  • the time taken by the trips tends to increase with increasing traffic flows.
  • the fear of meeting, along the theoretically best route, with a traffic volume close to its maximum may induce users to choose "inferior" alternative routes, which can nevertheless allow shorter trip times.
  • the method according to the present invention systematically distinguishes demand for route from actual flow expected on the same route.
  • the method aims at assessing the respective probability, in consideration of the non-deterministic approach adopted.
  • the method according to the present invention relies on preliminary probabilistic assessments of conditions that may influence decisions of various users as to choice of the most convenient routes to go in a given transportation system.
  • preliminary probabilistic assessments as supposedly made by users prior to undertaking trips, concern the expected trip and traffic conditions relevant to the selected range of routes that is considered as the set of reasonable alternative routes for each possible travel.
  • the data regard the disaggregate flows of the different categories of vehicles and transported loads, surveyed independently of each other.
  • Figure 1 shows a flow chart describing the traffic assignment method 1 according to the present invention.
  • the traffic assignment method 1 comprises a probabilistic assignment 11, a congestion analysis 12 and a results analysis 13.
  • the traffic assignment method 1 may further comprise a calibration 14.
  • the calibration 14, as shown in figure 1 may be a preliminary calibration performed before implementing the traffic assignment method 1 or it may be an a posteriori calibration based on the results of the implementation of the traffic assignment method 1.
  • the traffic assignment method 1 is based on the mathematical theory that will described in paragraph 2 (and respective subparagraphs) .
  • the traffic assignment method 1 simulates the distribution of the entire transport demand on selected alternative routes of the network.
  • the simulation calculates the volumes of traffic that are expected by the user on the identified routes pertaining to each travel, accounting also for the conditioning effects (feed-back) caused by the traffic itself.
  • the "aggregate" demand of all transport categories (passengers, goods and relative classes), which is needed to evaluate the real effect of the load on links, is measured with classical equivalence methods, vehicle load, etc., applied to the different transportation modes considered.
  • the reference of the transport demand to the "analysis period”, as the time interval considered by the particular application (i.e. day or week or month or year) is reduced to an average equivalent unit (pcu, passenger car unit) per hour, through the analytical classical procedures of yearly, seasonal, weekly, daily and hourly distribution of traffic that characterises the study network and period.
  • the transportation network As to the transportation network, this is analysed in all its components, starting from Origin-Destination (0-D) poles and links up to the junction or interchange nodes, through the determination, by engineering or empirical means, of the unit costs of the unit type- vector in free flow condition, i.e. with no other load affecting the route links.
  • the unit transport cost (relevant to the equivalent unit) on each link is a oneway cost, since the motion direction on a same link may significantly affect the cost. It is also advisable to point out the nature of such unit costs, which, as in any assignment application, summarize vehicle operation and time costs by a generalized transport cost.
  • any route unit cost is the summation of the one-way unit costs relative to all the components of the same route: links, nodes, crossing poles as well as origin and destination poles, so representing the generalized unit cost of a free flow on the route.
  • Equation (*) The definition of the unit costs adopted by the traffic assignment method 1 is given by equation (*) here below, which describes the feed-back effect caused by the expected stream volume on the free flow unit cost, i.e.
  • S is the specific volume of traffic (stream) expected on the route,- it is the problem's principal unknown.
  • Quantity (1 + k LnS) is here referred to as loading factor.
  • the probabilistic assignment 11 comprise two consecutive phases: an aggregate assignment and a disaggregate assignment.
  • the first step of the aggregate assignment is the aggregate setting, necessary to determine the "network constants" proper to the system under consideration. These constants, which include coefficient k, secure the consistent use of the various units of the measurement system that has been adopted to quantify the parameters of the mathematical model, and accompany the entire implementation of the traffic assignment method 1. This input makes also possible to re-balance the scale of costs, when inconsistencies are detected in the light of the given reference streams, as it may happen because of some approximate cost estimates.
  • the second step of the aggregate assignment consists of an iterative calculation carried out by use of the network constants and under the constraints imposed by the probabilistic theory, to determine all the "travel constants” pertaining to the study network.
  • the “travel constants” differ from the "network constants” , and are associated with the physical and functional aspects that characterise the set of selected alternative routes by which every "travel” is identified.
  • the "travel constants" are the necessary key to start the next step.
  • the third step of the aggregate assignment implements, on the basis of the aggregate costs of the single routes, the preliminary aggregate assignment, whose result is the set of values that quantify the trip demands assigned to the network routes, according to the probabilistic approach proper to the mathematical theory.
  • This preliminary aggregate assignment accounts for all the transport categories considered together.
  • the stream volumes calculated by this preliminary aggregate assignment enter definitions (*) for a first determination of the generalized costs to use in the subsequent simulation stages. These costs, however, do not represent real monetary transport costs, but only a simulated scale of these. The actual monetary transport unit costs, relevant to the study network, will be calculated during the results analysis 13.
  • the application may be repeated as many times as the number of available reference travels, to choose eventually, among the different sets of results obtained, those sets which are most similar to each other and, among them, the most credible one, keeping all of the relevant calculated parameters for the subsequent simulation stages.
  • the second part of the probabilistic assignment 11 implements the assignment of disaggregate route demands, i.e. the disaggregate assignment.
  • the initial step of the disaggregate assignment in an analogy with the procedure followed for the aggregate assignment, is the disaggregate setting, which determines the "category constants" that pertain to the study network by use of the available data relevant to the same "reference travel” , for which streams and unit costs are known.
  • This setting is carried out through specific probabilistic equations defined by the mathematical theory. Also at this stage, it's possible to re-balance the scale of the category costs in relation to the category streams surveyed, should such an operation be necessary upon detection of inconsistencies .
  • the next steps of the disaggregate assignment consists of an iterative process concerning the assignment of the disaggregate streams, operating a specific iterative procedure for each transport category.
  • This procedure develops through a series of subsequent computerised assignment runs, each of which yields a set of "tentative category streams" , to be re- aggregated all together by summation and introduced in the loading factors (one for each route) , in order to quantify the category unit costs to be used in the disaggregate assignment run that follows. And so on, up to the convergence on the final values searched for the category streams .
  • the iterations converge on a final situation that represents the probabilistic distribution of the demand for category streams in the network and, by summation of these per route, also the final aggregate demand assignment.
  • the latter does not coincide with the preliminary aggregate assignment, which, by definition, is only a preliminary assessment.
  • the final aggregate assignment together with the relevant disaggregate assignment, forms the conclusive result of the probabilistic determination of the route demands .
  • the functional characteristics of the network links are preponderant in determining the assignment of the route demands, when the study system is a large regional/interregional transportation network.
  • the role of the crossing nodes and poles is certainly less important, though the simulation associates appropriate crossing costs with these network components in the probabilistic process.
  • the method in its present version focuses the congestion analysis 12 only on the road transport mode.
  • the principle, statistically corroborated, which governs the simulation of states of congestion is that road users show a certain degree of tolerance toward congestion regimes, provided these do not last too long and that the traffic stream comes back to normality after a reasonable interval.
  • the simulation considers that the users can get timely informed about the risk of traffic blockages in links of their route, so as to be in condition to opt for alternative routes .
  • the simulation compares any demanded stream on road, as previously calculated, with the capacity of each link of the road-route regarded, on the assumption that a certain amount of stream exceeding the link capacity can still be accommodated, despite the intervening congestion, up to a predetermined volume of traffic.
  • the criterion adopted is that the excess is admissible and accepted by the link if it does not bring the average speed of the traffic flow below a conventional level.
  • the computerised simulation does automatically transfer any amount of not accommodate excess of road-route demand to the most convenient alternative road-routes of the same travel, getting a response that follows the same criterion mentioned above. In accommodating the residual demand remainders, the procedure may involve the whole set of the road- routes of the same travel.
  • the main innovation of this congestion analysis consists of having turned traditional empirically drawn "speed-flow diagrams" into mathematical equations, defined in the mathematical theory, which describe precisely the relationships that exist between traffic flow speed, traffic intensity, traffic density, traffic congestion and road capacity.
  • these equations have made it possible to point out the quantitative relation between level of congestion and respective traffic flow speed. This also implies a more accurate calculation of the mean trip duration in each road link and, therefore, a more reliable assessment of the actual travel costs as resulting at the simulation conclusion.
  • the results of the congestion analysis 12 lead to a necessary revision of the probabilistic assignment 11 of the route demand, in order to obtain the expected actual traffic assignment.
  • the revision is made and saved in the last part of the method application.
  • the simulation revises the probabilistic demand assignment, according to the outcomes of the congestion analysis 12 on road links, so as to outline a final picture of the expected actual assignment of the traffic streams. It should go without saying that the probabilistic demand assignment coincides with the expected actual stream distribution if no case of congestion is detected by the simulation.
  • the category stream values are translated from pcu/hour units into effective number of category vehicles per hour, by means of the equivalence parameters and average loads initially adopted in quantifying the input for the traffic assignment method 1.
  • This operation regards every transport mode considered, and concludes with the summation of overlapping streams on links, thus determining the relevant flows in vehicle units (passenger cars, light and heavy trucks, buses, trains, etc.) per hour of operation.
  • results analysis 13 returns and lists the original values of category transport units (passengers, tons) transported on each route,- which are then summed-up all together to display the total volumes transported by transport mode and category.
  • the results analysis 13 proceeds on to assess the network total costs, number of vehicles per km, and average unit costs per vehicle and per transported unit.
  • This final table constitutes a fundamental information basis when alternative transformation projects proposed for the system have to be mutually compared, in order to enable the planners to select the most convenient options.
  • the traffic assignment method 1 may further comprise a calibration 14, which, as shown in figure 1, may be a preliminary calibration performed before implementing the traffic assignment method 1 or an a posteriori calibration based on the results of the implementation of the traffic assignment method 1.
  • the calibration 14 is based on the inverse use of the traffic assignment method 1.
  • the opportunity and convenience for a preliminary calibration becomes evident when all the data relevant to the traffic in the study network are made available through the findings of direct field investigations, i.e. when all the real traffic streams are directly measured on the network routes.
  • an a posteriori calibration based on all traffic streams resulting from the implementation of the traffic assignment method 1, may be very useful for subsequent implementations of the traffic assignment method 1.
  • the inverse use of the traffic assignment method 1 can be exploited for a consistent assessment of the whole set of transport costs, provided that, along with the data concerning the traffic streams, mean actual transport costs per category are also provided as for two routes of any reference travel, plus one category mean actual transport cost, whatever category, per every other travel of the network. Once such data are made available, the inverse use of the traffic assignment method 1 allows for the determination of all the remaining unknown costs, according to the logic consistency of the traffic assignment method 1.
  • the basic relationship existing between costs and streams comes into evidence from the objective conditions in which the system operates, so as a reliable calibration of the transport costs can be obtained through the mathematical theory equations and used for the input settings of possible subsequent simulation exercises.
  • the mathematical procedure that performs the calibration 14 by the inverse use of the traffic assignment method 1 is the solution of systems of linear equations derived, by logical symmetry, from the probabilistic equations applied in the direct use of the traffic assignment method 1, i.e. for the route demand assignment.
  • the computerised process for such an inverse use requires a simplified procedure with respect to the procedure required by the direct use of the traffic assignment method 1.
  • the calibration procedure doesn't imply any congestion analysis 12, since it starts from an actual situation, real or simulated, which obviously includes the possible cases of traffic congestion.
  • the calibration 14 doesn't change the probabilistic nature of the traffic assignment method 1 and is applicable to either aggregate or disaggregate transport.
  • the main purpose of the calibration 14 is that of providing method operators with a consistent series of route cost parameters, in view of credible assignment simulations concerning the study system under various design conditions. Exercises of this kind are very useful for transport analyses, either for preparing feasibility studies or for project optimisation purposes, on complex networks and under different hypotheses of demand expansion. 2.
  • the total traffic flow that, in a given time unit, moves from one pole to another consists of the summation of all the streams that engage the set of alternative routes connecting the two poles .
  • the term “stream” indicates the amount of traffic engaging one single route in the time unit.
  • f ⁇ ⁇ is the stream in any route i of the r ⁇ routes of travel T;
  • F is the overall traffic demand engaging the whole system's network in the time unit considered.
  • the general problem to solve is the determination of the distribution of all the streams in the whole set of routes of the system' s network, which means the determination of the expected value for every fj 7 .
  • the first stage accounts for the data incompleteness.
  • the available information is supposed to be the more or less accurate information shared by most network users, so that each "datum" must be taken as a coarse mean estimate of many different subjective estimates.
  • Such a probable demand for route is substantially a subjective theoretical assessment. As said above, this demand for route does not necessarily correspond to the actual stream to be expected in each route . This also depends on the route capacity, which is not always in condition to accommodate all the relevant traffic flow demanded.
  • each given travel demand is compared with the respective set of both estimated demand for routes and the capacities of these, to find a logical criterion for assessing the probable traffic flow distribution to be actually expected.
  • Cj r the free-route mean transportation unit cost for going route i of travel T.
  • Any user knows this cost may only exceptionally be the actual one. Any network user does normally associate the expected transportation cost (at least in terms of time) also with the expected intensity of the traffic on the route .
  • T is also a function of the overall demand for the same route expected, or guessed, by users.
  • the model formalizes the resulting cost function as follows:
  • Jc is a positive constant proper to the study system
  • Ln is the operational symbol for "natural logarithm”
  • d ⁇ ⁇ is the expected demand for route i of travel T.
  • Demand d ⁇ ⁇ represents the first set of the problem' s unknowns .
  • K ⁇ ⁇ is the expected mean unit cost
  • the cost to be associated with demand d ⁇ ⁇ is expressed by:
  • Function K ⁇ ⁇ is a working hypothesis, which uses logarithms upon the assumption that no demand for route is less than one, i.e., no route of the alternative routes of T 1 is engaged by less than one transportation unit per time unit. This, from another standpoint, is also the basic assumption and criterion for identifying the set of alternative routes of any travel .
  • the summation of route demands coincides with the summation of the travel streams.
  • quantities B ⁇ and U are fictitious data. Such quantities, in the mathematical process that follows, are either eliminated or incorporated into other compound quantities that can easily be determined.
  • the first problem is to determine the probable distribution of route demand d ⁇ ⁇ amongst all the routes of the system.
  • Relation [8] defines variable p ⁇ ⁇ as the probability that any unit of route demand in the system relates to route i of travel T.
  • Condition [9] facilitates the solution of the problem, considering that the number of the unknowns is much greater that the number of equations involving the same unknowns .
  • F being one of the problem's data.
  • probability distributions there is one of a particular importance for this model.
  • Uncertainty E is a positive or nil quantity. If R is the number of probabilities of the distribution considered, then the absolute maximum uncertainty is expressed by:
  • Uncertainty E is nil when one event only is possible and is therefore certain ⁇ E is also nil when no event is possible and, therefore, all probabilities are nil) .
  • the determination of the probabilities can be obtained by calculation of the maximum amount of uncertainty that is possible under a set of given constraints.
  • constraints are formed by the equations that bind the values of the distribution probabilities to verify certain given conditions . It must be observed that, by hypothesis, all the available information is expressed by the constraint- equations, which means that the uncertainty is objectively maximum as for the rest.
  • the determination of the constrained maximum of any mathematical function can be made according to various methods, of which the Lagrange Multipliers Method is the most usual.
  • ⁇ 3(C/ Lnef/- CJ Lnef/ ) + CJ Ln (fj/fj) - for any m and n of any T, e being the base of natural logarithms .
  • Equations from [18] to [22] above immediately explain the reason why it is necessary to know at least two streams f m ⁇ and £ a ⁇ , or two stream probabilities pj and p n ⁇ , relative to any travel T 7 considering the general definition by which
  • the solution to the problem is usually requested separately for either vehicle or load category.
  • the two main load categories to consider consist of passengers, i.e. human travelers, and goods .
  • any network user prior to selecting the route to use, is supposed to estimate the impact of aggregate transport flows on the particular mean aggregate unit cost associated with the traffic flow that the same user is contributing to create .
  • any network user In addressing disaggregate streams, any network user is still supposed to estimate the impact of aggregate transport flows, but the estimated impact is now on the mean unit cost of the particular category stream that concerns the user.
  • the aggregate flow is then taken as a datum that contributes to the determination of the transportation category mean unit cost assessed by the category user. Access to the disaggregate model therefore requires an initial application of the aggregate model.
  • Disaggregate model intends to formulate the probable route demand expected by category users, such as, for example, passenger or commodity transporters.
  • di T is, initially, the route aggregate demand determined through previous application of the aggregate model and, therefore, a given input necessary to start the application of the disaggregate model.
  • di T is the summation of the disaggregate demands (category streams) relevant to route i of travel T, resulting, each time, from the previous iteration.
  • the disaggregate model assumes that category travel demands G ⁇ are exogenous given data.
  • [30.J ⁇ is a category constant for the entire network,-
  • D ⁇ ⁇ must therefore be considered as the category actual mean unit costs, as conditioned by the actual traffic stream intensity.
  • the second stage of the problem solution consists of the determination of the traffic load that may actually be expected on each route of the system.
  • This aspect of the problem is particularly important in the analysis of the road network, where the traffic is not kept under a centralized planning control. This is one of the main causes of frequent traffic congestions in sections of the road network, where the flocking users, often unaware of the actual traffic conditions they are going to meet with, do not only cause the saturation of the road capacity but also an over-concentration of vehicles that lowers the flow below or much below the road capacity.
  • F is the flow intensity, as given by the number of transportation units crossing a conventional road section in a given time unit;
  • D is the mean vehicular flow density in transportation units per road-length unit;
  • V is the mean flow speed in road-length units per time unit .
  • (length) varies with the flow speed from an average conventional minimum b (body) , when the speed is nil, to a maximum length that depends on the maximum speed the transportation unit is allowed to attain in the road considered.
  • a conventional transportation unit 2 can be expressed by the following function of its speed:
  • coefficient a depends basically on the road characteristics .
  • n pcu passenger car unit
  • This model uses both the definitions above, according to the analysis or the simulation circumstances involved.
  • Definition [33] is particularly used to determine the values for F 1 D and V by use of equations [32] .
  • Acronym pcu is instead currently used to express the value of traffic flow intensity, in terms of pcu per time unit, and the value of flow density, in terms of pcu per road-length unit.
  • h is the adopted road-length unit
  • the value of flow density can in particular be expressed in pcu/h by:
  • coefficient a has the physical dimension of the inverse of an acceleration.
  • the road-length unit h relates to a one-lane road. If the road lanes considered for each flow direction are more than one, h must be multiplied by the number of lanes .
  • the overall traffic conditions in the road network may cause the flow demand for some oriented links to exceed the capacity of these links .
  • the analytical function which expresses the vehicular flow density in any congested road link x as depending on the excess of demand, must be subject to a number of constraints .
  • a x d x - ⁇ x > 0 represents the excess in the flow demand. It should go without saying that A x ⁇ 0, when d x ⁇ ⁇ x , represents instead the balance to the link saturation.
  • Congestion density D x in the link is a function of A x subject to the following constraints:
  • compression regards the flow density, though compressed density may be taken as a measurement of the relevant flow potential. Normally, in fact, as soon as the cause of the congestion ceases, the various streams that have been reduced/compressed by congestion can regain their respective initial values.
  • Demand d x for link x brought about by the bundle of different routes that have congested link x in common, may be considered, in terms of probability, as composed by a set of different sub-demands d x r translated into compressed streams f x r as expressed by:
  • r indicates any route of the bundle.
  • the initial values of the streams, which have been compressed by the congestion in x, can be recovered, if the road links subsequent to congested link x provide the physical conditions that allow the re-expansion of the compressed streams.
  • any compressed stream whose initial value is f r , must regain its proper mean velocity V r to re-expand completely in traversing the link subsequent to congested link x.
  • all compressed streams have the same mean speed V x , as given by equation [46] .
  • any stream incurring congestion in any link x of its route is not suppressed, if it can re- expand in a subsequent link whose length is greater than [51] .
  • Practical examples indicate that compressed flows can recover their original intensities in some thousand meters, however low the congestion speed may be.
  • the first case occurs when link x has attained a flow stoppage state, and no additional transportation unit can physically enter the clogged link.
  • the simulation addresses this case by calculation of the portion of rejected demand for link x on the basis of a conventional minimum average speed tolerated in the congested or quasi- blocked link.
  • equation [35] it is possible to assess the intensity of the flow associated with the conventional minimum speed
  • equation [48] it is possible to determine the portion of the demanded flow that is rejected by the clogged link.
  • the second case is the one in which the impossibility of re-expansion for any stream compressed by congestion results from the insufficient length of the link subsequent to the congested one; or it results from the fact that the subsequent link, though sufficiently long, enables the vehicles to quickly achieve an average speed that is much higher than that proper to the original sub-flow.
  • flow intensity declines with the average vehicular speed allowed by the road. It doesn't seem reasonable to assume that this is a frequent case, considering the statistical evidence provided by real traffic situations. The case, however, is possible, and deserves some theoretical attention.
  • the route to consider is formed by a sequence of several links, which sooner or later allow the re-formation of the streams compressed by congestion.
  • the analysis must be extended to the whole set of links that follow the congested one to establish whether and where the solution is found.
  • Figure 1 shows a graph of the traffic flow F as a function of the average flow speed V relating to road links
  • Figure 2 shows a graph of the traffic flow
  • Figure 3 shows a graph of exceeding traffic densities in a congestion state in a road link.
  • synoptic list of the above- presented basic mathematical relations used for determining the expected flow assignment and for assessing the running times .
  • the synoptic list includes:
  • parameter h is 1000 m in the metric system, and varies with the measurement system adopted.
  • a preliminary overall calibration exercise may be needed, with a view to making all the transportation unit costs mutually consistent and compatible with the logic structure of the model .
  • the preliminary calibration exercise is to be based on a direct traffic survey aimed at measuring the intensity of the streams in the various routes of the study system.
  • the mean transportation unit costs are supposed to be the known parameters, while the traffic streams are considered as the unknowns to be determined.
  • rh(N 2 -N) is the number of category streams to be determined, h being the number of transportation categories considered.
  • the use of the disaggregate model requires that two category streams, concerning any pair of routes of any reference travel, i.e. 2h category streams, be given for the determination of h category constants ⁇ a as well as of the remaining (N 2 -N) rh-2h unknown category streams. Which results in the overall number of h[ (N 2 -N) r-1] unknowns .
  • the inverse use of the aggregate model is rather different from the inverse use of the disaggregate model .
  • the basic purpose of the calibration is not that of assessing the real transportation mean unit costs, but rather that of establishing a logically consistent set of data that work as an appropriate input in a correct application of both the aggregate and disaggregate models.
  • the calibration of the category conditioned mean unit costs D y s requires that at least one (anyone) of the category conditioned costs, relevant to any one route of each travel, be known; together with one additional category cost to be associated, in a pre-fixed "reference travel” , with the other unit cost that is known by hypothesis.
  • D ⁇ ⁇ and D k ⁇ of known category mean unit costs that pertain to reference travel T it is immediately possible to determine the system' s category constant ⁇ a by virtue of equation [31] , i.e.
  • ⁇ a Ln ( g k a ⁇ /g ⁇ a ⁇ ) / ( D 1 a ⁇ - D k a ⁇ )
  • Jc is the system's constant proper to the aggregate model (see [17] and [18] )
  • d/ is the aggregate traffic stream that engages route y of travel S.
  • any aggregate stream d y s can immediately be determined if the whole set of category streams in the network is given. In fact, in that case, any aggregate stream can soon be calculated as a summation of its component category streams, i.e. as
  • the problem to solve, for using [64] is the determination of k, which is associated with the calibration of the aggregate model.
  • the present invention has been successfully tested through realistic and satisfactory applications.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Navigation (AREA)
  • Traffic Control Systems (AREA)

Abstract

L'invention porte sur un procédé d'attribution de flux de trafic dans un réseau de transport qui comprend des liaisons de transport aptes à permettre un transport de flux de trafic. Chaque liaison de transport est comprise entre deux nœuds et chaque nœud est soit une extrémité de liaison de transport simple, soit interconnecte au moins deux liaisons de transport. Le réseau de transport est associé à des demandes de transport entre des couples de nœuds donnés et chaque couple de nœuds donné est interconnecté par des itinéraires respectifs, chaque itinéraire étant une séquence respective de liaisons de transport et étant associé à des coûts de transport d'itinéraire respectifs. Le procédé comprend les opérations consistant à calculer pour chaque itinéraire une probabilité de demande de transport respective basée sur les coûts de transport d'itinéraire et sur les demandes de transports; et attribuer à chaque liaison de transport un flux de trafic respectif sur la base des probabilités de demande de transport.
PCT/IT2008/000194 2008-03-21 2008-03-21 Procédé d'attribution de trafic pour des réseaux de transport multimodaux WO2009116105A2 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/IT2008/000194 WO2009116105A2 (fr) 2008-03-21 2008-03-21 Procédé d'attribution de trafic pour des réseaux de transport multimodaux

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/IT2008/000194 WO2009116105A2 (fr) 2008-03-21 2008-03-21 Procédé d'attribution de trafic pour des réseaux de transport multimodaux

Publications (2)

Publication Number Publication Date
WO2009116105A2 true WO2009116105A2 (fr) 2009-09-24
WO2009116105A3 WO2009116105A3 (fr) 2010-01-28

Family

ID=40451282

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IT2008/000194 WO2009116105A2 (fr) 2008-03-21 2008-03-21 Procédé d'attribution de trafic pour des réseaux de transport multimodaux

Country Status (1)

Country Link
WO (1) WO2009116105A2 (fr)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102063744A (zh) * 2010-11-11 2011-05-18 天津高速公路集团有限公司 基于车牌识别的高速公路二义性路径通行费拆分方法
WO2013045150A1 (fr) * 2011-09-26 2013-04-04 Robert Bosch Gmbh Procédé de transmission de données d'itinéraire pour la télématique de circulation routière
WO2013120765A1 (fr) * 2012-02-17 2013-08-22 Bayerische Motoren Werke Aktiengesellschaft Procédé de construction d'un modèle pour une banque de données de temps de parcours
WO2014091674A1 (fr) * 2012-12-12 2014-06-19 Toyota Jidosha Kabushiki Kaisha Appareil et procédé d'aide à la création de plans de transport
EP3186797A4 (fr) * 2014-08-27 2018-07-18 Sparkcity.com Ltd. Système et procédé de stationnement individuel et régional
CN108388542A (zh) * 2018-01-18 2018-08-10 中国科学院遥感与数字地球研究所 一种轨道交通拥挤度计算方法及系统
US10345109B2 (en) 2013-09-15 2019-07-09 Tomtom Navigation B.V. Generating routes to optimise traffic flow
CN111458769A (zh) * 2020-05-26 2020-07-28 南京大学 用于输电线路环境气象数据预测的方法及系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6131064A (en) * 1996-02-06 2000-10-10 Mannesmann Aktiengesellschaft Vehicle-autonomous detection of traffic backup
WO2005088493A1 (fr) * 2004-03-10 2005-09-22 Ingo Morgenstern Procede et systeme d'optimisation de travaux de transport
EP1657693A2 (fr) * 2004-11-16 2006-05-17 Microsoft Corporation Prédiction du trafic avec l'aide du modelage et de l'analyse des interdépendances probabilistiques et de données contextuelles
WO2008009271A1 (fr) * 2006-07-20 2008-01-24 Deutsche Telekom Ag Procédé et dispositif pour générer des alertes précoces signalant des ruptures de trafic au niveau de goulots d'étranglement

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6131064A (en) * 1996-02-06 2000-10-10 Mannesmann Aktiengesellschaft Vehicle-autonomous detection of traffic backup
WO2005088493A1 (fr) * 2004-03-10 2005-09-22 Ingo Morgenstern Procede et systeme d'optimisation de travaux de transport
EP1657693A2 (fr) * 2004-11-16 2006-05-17 Microsoft Corporation Prédiction du trafic avec l'aide du modelage et de l'analyse des interdépendances probabilistiques et de données contextuelles
WO2008009271A1 (fr) * 2006-07-20 2008-01-24 Deutsche Telekom Ag Procédé et dispositif pour générer des alertes précoces signalant des ruptures de trafic au niveau de goulots d'étranglement

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102063744B (zh) * 2010-11-11 2012-11-21 天津高速公路集团有限公司 基于车牌识别的高速公路二义性路径通行费拆分方法
CN102063744A (zh) * 2010-11-11 2011-05-18 天津高速公路集团有限公司 基于车牌识别的高速公路二义性路径通行费拆分方法
CN103890823B (zh) * 2011-09-26 2016-08-17 罗伯特·博世有限公司 传输用于交通远程信息处理的路线数据的方法及处理单元
WO2013045150A1 (fr) * 2011-09-26 2013-04-04 Robert Bosch Gmbh Procédé de transmission de données d'itinéraire pour la télématique de circulation routière
CN103890823A (zh) * 2011-09-26 2014-06-25 罗伯特·博世有限公司 用于传输用于交通远程信息处理的路线数据的方法
US9053631B2 (en) 2011-09-26 2015-06-09 Robert Bosch Gmbh Method for transmitting route data for traffic telematics
WO2013120765A1 (fr) * 2012-02-17 2013-08-22 Bayerische Motoren Werke Aktiengesellschaft Procédé de construction d'un modèle pour une banque de données de temps de parcours
CN104169986B (zh) * 2012-02-17 2016-10-26 宝马股份公司 用于为旅行时间数据库建立模型的方法
CN104169986A (zh) * 2012-02-17 2014-11-26 宝马股份公司 用于为旅行时间数据库建立模型的方法
US9122983B2 (en) 2012-02-17 2015-09-01 Bayerische Motoren Werke Aktiengsellschaft Method for model construction for a travel-time database
CN104838433A (zh) * 2012-12-12 2015-08-12 丰田自动车株式会社 交通运输计划创建支持设备和交通运输计划创建支持方法
WO2014091674A1 (fr) * 2012-12-12 2014-06-19 Toyota Jidosha Kabushiki Kaisha Appareil et procédé d'aide à la création de plans de transport
US10345109B2 (en) 2013-09-15 2019-07-09 Tomtom Navigation B.V. Generating routes to optimise traffic flow
EP3186797A4 (fr) * 2014-08-27 2018-07-18 Sparkcity.com Ltd. Système et procédé de stationnement individuel et régional
CN108388542A (zh) * 2018-01-18 2018-08-10 中国科学院遥感与数字地球研究所 一种轨道交通拥挤度计算方法及系统
CN111458769A (zh) * 2020-05-26 2020-07-28 南京大学 用于输电线路环境气象数据预测的方法及系统

Also Published As

Publication number Publication date
WO2009116105A3 (fr) 2010-01-28

Similar Documents

Publication Publication Date Title
WO2009116105A2 (fr) Procédé d'attribution de trafic pour des réseaux de transport multimodaux
Li et al. Vehicle routing problems with time windows and stochastic travel and service times: Models and algorithm
Tirachini et al. Accounting for travel time variability in the optimal pricing of cars and buses
Ma et al. Modeling bus travel time reliability with supply and demand data from automatic vehicle location and smart card systems
Zhao The planning and analysis implications of automated data collection systems: rail transit OD matrix inference and path choice modeling examples
Park et al. A stochastic emergency response location model considering secondary incidents on freeways
Liu et al. Observability quantification of public transportation systems with heterogeneous data sources: An information-space projection approach based on discretized space-time network flow models
Aron et al. Estimating travel time distribution for reliability analysis
Tavassoli et al. Application of smart card data in validating a large-scale multi-modal transit assignment model
Si et al. Urban transit assignment model based on augmented network with in-vehicle congestion and transfer congestion
Godachevich et al. Does the measured performance of bus operators depend on the index chosen to assess reliability in contracts? An analysis of bus headway variability
Lu et al. A bilevel model with a solution algorithm for locating weigh-in-motion stations
Su et al. Simulation-Based Method for the Calculation of Passenger Flow Distribution in an Urban Rail Transit Network Under Interruption
Li et al. Reliability evaluation for stochastic and time-dependent networks with multiple parking facilities
Boyles et al. Incorporating departure time choice into high-occupancy/toll (HOT) algorithm evaluation
Kim et al. Integrated decision model of mode, line, and frequency for a new transit line to improve the performance of the transportation network
Mehrabani et al. Development, calibration, and validation of a large-scale traffic simulation model: Belgium road network
Di Mino et al. Pavement Management System model using a LCCA-microsimulation integrated approach.
Hu Developing the analysis methodology and platform for behaviorally induced system optimal traffic management
Yu et al. Departure time and route choices with accurate information under binary stochastic bottleneck capacity in the morning commute
Shen et al. Modeling stochastic overload delay in a reliability-based transit assignment model
Graham A comparative evaluation of FDSA, GA, and SA non-linear programming algorithms and development of system-optimal methodology for dynamic pricing on I-95 Express
Bar-Gera et al. Evaluating the assumption of independent turning probabilities
Abbas-Turki et al. Public transportation systems: modeling and analysis based on a new Petri net approach
Srotyr et al. Prediction system of occupancy of parking spaces

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 08763784

Country of ref document: EP

Kind code of ref document: A2

NENP Non-entry into the national phase in:

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 08763784

Country of ref document: EP

Kind code of ref document: A2