I research complex social systems: urban mobility. Congested, urban multimodal networks used by millions of agents to reach their destinations and leaving huge sets of mobility traces ready to be applied for modelling, optimization, understanding and control. Full CV.
Currently, I run the ERC Starting Grant COeXISTENCE where I bridge between ML and transportation - we simulate how intelligent machines compete with humans for limited urban resources (space) and what can we expect. See details here.
I am an Associate Professor at the Faculty of Mathematics and Computer Science, Jagiellonian University in Krakow (Poland) with the Group of Machine Learning Research GMUM. I am involved in SUM project, Horizon Europe. From 2021 to 2024 I led the NCN Opus Grant on Shared Mobility in the pandemic times
Before, I worked (2019-2021) with prof. Oded Cats at TU Delft in his ERC Starting Grant Critical MaaS. I modelled two-sided mobility platforms, specifcally focusing on ride-pooling (ExMAS) and agent-based simulator for Uber-like systems (MaaSSim). I did PhD at Cracow University of Technology with an excellent group of prof. Andrzej Szarata and working closely with Guido Gentile from La Sapienza on non-equilibrium dynamic traffic assignment. In the interdisciplinary field of urban mobility I did research which can be classified as:
- model estimation, optimization, system control, network design;
- agent-based simulation, game-theory, network science, stochastic simulation, epidemic modelling;
- machine learning, spatial analysis, big data analysis, pattern recognition, unsupervised learning;
- behavioural modelling, economic discrete choice models, policy, sustainability.
Teaching materials
I run the seminar on Complex Social Systems (transport is one of them) at Jagiellonian University - materials and papers are on my github
List of main publications and preprints
-
Hyper pooling private trips into high occupancy transit like attractive shared rides
npj Sustainable Mobility and Transport
2024
The size of the solution space associated with the trip-matching problem has made the search for high-order ride-pooling prohibitive. We introduce hyper-pooled rides along with a method to identify them within urban demand patterns. Travellers of hyper-pooled rides walk to common pick-up points, travel with a shared vehicle along a sequence of stops and are dropped off at stops from which they walk to their destinations. While closely resembling classical mass transit, hyper-pooled rides are purely demand-driven, with itineraries (stop locations, sequences, timings) optimised for all co-travellers. For 2000 trips in Amsterdam the algorithm generated 40 hyper-pooled rides transporting 225 travellers. They would require 52.5 vehicle hours to travel solo, whereas in the hyper-pooled multi-stop rides, it is reduced sixfold to 9 vehicle hours only. This efficiency gain is made possible by achieving an average occupancy of 5.8 (and a maximum of 14) while remaining attractive for all co-travellers.
-
Simulating two-sided mobility platforms with MaaSSim
PLoS ONE
2022
Two-sided mobility platforms, such as Uber and Lyft, widely emerged in the urban mobility landscape. Distributed supply of individual drivers, matched with travellers via intermediate platform yields a new class of phenomena not present in urban mobility before. Such disruptive changes to transportation systems call for a simulation framework where researchers from various and across disciplines may introduce models aimed at representing the complex dynamics of platform-driven urban mobility. In this work, we present MaaSSim, a lightweight agent-based simulator reproducing the transport system used by two kinds of agents: (i) travellers, requesting to travel from their origin to destination at a given time, and (ii) drivers supplying their travel needs by offering them rides. An intermediate agent, the platform, matches demand with supply. Agents are individual decision-makers. Specifically, travellers may decide which mode they use or reject an incoming offer; drivers may opt-out from the system or reject incoming requests. All of the above behaviours are modelled through user-defined modules, allowing to represent agents’ taste variations (heterogeneity), their previous experiences (learning) and available information (system control). MaaSSim is a flexible open-source python library capable of realistically reproducing complex interactions between agents of a two-sided mobility platform. MaaSSim is available from a public repository, along with a set of tutorials and reproducible use-case scenarios, as demonstrated with a series of illustrative examples and a comprehensive case study.
-
Modelling virus spreading in ride-pooling networks
Kucharski, Rafał,
Cats, Oded,
and Sienkiewicz, Julian
Scientific Reports
2021
Urban mobility needs alternative sustainable travel modes to keep our pandemic cities in motion. Ride-pooling, where a single vehicle is shared by more than one traveller, is not only appealing for mobility platforms and their travellers, but also for promoting the sustainability of urban mobility systems. Yet, the potential of ride-pooling rides to serve as a safe and effective alternative given the personal and public health risks considerations associated with the COVID-19 pandemic is hitherto unknown. To answer this, we combine epidemiological and behavioural shareability models to examine spreading among ride-pooling travellers, with an application for Amsterdam. Findings are at first sight devastating, with only few initially infected travellers needed to spread the virus to hundreds of ride-pooling users. Without intervention, ride-pooling system may substantially contribute to virus spreading. Notwithstanding, we identify an effective control measure allowing to halt the spreading before the outbreaks (at 50 instead of 800 infections) without sacrificing the efficiency achieved by pooling. Fixed matches among co-travellers disconnect the otherwise dense contact network, encapsulating the virus in small communities and preventing the outbreaks.
-
If you are late, everyone is late: late passenger arrival and ride-pooling systems’ performance
Kucharski, Rafał,
Fielbaum, Andres,
Alonso-Mora, Javier,
and
Cats, Oded
Transportmetrica A: Transport Science
2021
Sharing rides in on-demand systems allow passengers to reduce their fares and service providers to increase revenue, though at the cost of adding uncertainty to the system. Notably, the uncertainty of ride-pooling systems stems not only from travel times but also from unique features of sharing, such as the dependency on other passengers’ arrival time at their pick up points. In this work, we theoretically and experimentally analyse how late arrivals at pick up locations impact shared rides’ performance. We find that the total delay is equally distributed among sharing passengers. However, delay composition gradually shifts from on-board delay only for the first passenger to waiting delay at the origin for the last passenger. Sadly, trips with more passengers are more adversely impacted. Strategic behaviour analysis reveals Nash equilibria that might emerge. We analyse the system-wide effects and find that when lateness increases passengers refrain from sharing and eventually opt-out.
-
Exact matching of attractive shared rides (ExMAS) for system-wide strategic evaluations
Transportation Research Part B: Methodological
2020
The premise of ride-sharing is that service providers can offer a discount, so that travellers are compensated for prolonged travel times and induced discomfort, while still increasing their revenues. While recently proposed real-time solutions support online operations, algorithms to perform strategic system-wide evaluations are crucially needed. We propose an exact, replicable and demand-, rather than supply-driven algorithm for matching trips into shared rides. We leverage on delimiting our search for attractive shared rides only, which, coupled with a directed shareability multi-graph representation and efficient graph searches with predetermined node sequence, narrows the (otherwise exploding) search-space effectively enough to derive an exact solution. The proposed utility-based formulation paves the way for model integration in travel demand models, allowing for a cross-scenario sensitivity analysis, including pricing strategies and regulation policies. We apply the proposed algorithm in a series of experiments for the case of Amsterdam, where we perform a system-wide analysis of the ride-sharing performance in terms of both algorithm computations of shareability under alternative demand, network and service settings as well as behavioural parameters. In the case of Amsterdam, 3000 travellers offered a 30% discount form 1900 rides achieving an average occupancy of 1.67 and yielding a 30% vehicle-hours reduction at the cost of halving service provider revenues and a 17% increase in passenger-hours. Benchmarking against time-window constrained approaches reveals that our algorithm reduces the search-space more effectively, while yielding solutions that are substantially more attractive for travellers.
-
Lewis-Mogridge Points: A Nonarbitrary Method to Include Induced Traffic in Cost-Benefit Analyses
Kucharski, Rafał,
Drabicki, Arkadiusz,
Paszkowski, Jan,
and Szarata, Andrzej
Journal of Advanced Transportation
2020
We propose a new method to estimate benefits of road network improvements, which allows to include the induced demand without arbitrary assumptions. Instead of estimating induced demand (which is nontrivial and hardly possible in practice), we search for demand induction where initial benefits are mitigated to zero. Such approach allows to formulate a dual measure of benefit, covering both the potential benefits and the likelihood of consuming them by the induced traffic. We first estimate benefits of road network improvement assuming that traffic demand is fixed. Consequently, we find demand model configurations at which the benefits of the new investment become null, i.e., all the initial benefits are consumed by the traffic demand growth. We call such states of induced demand the Lewis–Mogridge points of the analysed improvement. We select the most probable of such points and use it to calculate the proposed novel indicator μ, for which the initial benefits (obtained under a fixed-demand assumption) are multiplied with a demand increase rate needed to consume them. We believe that such measure allows to include the critical phenomena of induced traffic and, at the same time, to overcome problems associated with reliable estimation of induced demand. As we illustrate with the case of two alternative road improvement schemes in Kraków, Poland, the proposed method allows to estimate maximal threshold of induced traffic and to select scenario more resilient to induced traffic.
-
Simulation of rerouting phenomena in Dynamic Traffic Assignment with the Information Comply Model
Transportation Research Part B: Methodological
2019
We present the Information Comply Model (ICM) which extends the framework for macroscopic within-day DTA proposed by Gentile (2016) to represent the rerouting of drivers wrt a single traffic event. Rerouting is reproduced as a two-stage process: first, drivers become aware about the event and its consequences on traffic; second, drivers may decide to change path. At each arc, unaware drivers have a probability of being informed by multiple ATIS sources (radio, VMS, mobile apps), which depends not only on devise penetration rates, but also on users space and time coordinates wrt the position and interval of the event. At each node, aware drivers, who are somehow reluctant to change, may finally modify their path based on a random rerouting utility, which is composed of expected gains and avoided losses. ICM is thus capable of representing the evolution of rerouting phenomena in time and space when the information about a traffic event and its consequences on congestion spread among drivers and onto the network. This way, ICM extends the concept of dynamic user equilibrium to a case of imperfect information related to availability and awareness rather than to individual perception, as well as to a case of bounded rationality with prudent drivers. Besides the model architecture and specification, this paper provides a workable methodology which can be applied both off-line for transport planning and in real-time for traffic management on large size networks.
-
Multichannel queueing behaviour in urban bicycle traffic
Kucharski, Rafał,
Drabicki, Arkadiusz,
Żyłka, Klaudia,
and Szarata, Andrzej
European Journal of Transport and Infrastructure Research
2019
The objective of this paper is to propose a method to analyse and describe cyclists’ behaviour at signalized intersections with specific focus on the multichannel (multi-lane) queue phenomenon. As we observed, cyclists form queues without a fixed-lane and FIFO discipline, for which the classical, car-oriented analytical approach becomes insufficient. Cyclists’ multichannel queueing behaviour is common and characterized by substantial degree of variability, especially in case of shorter queues which emerge regularly at cycle crossings. Although cyclist behaviour has been widely studied by transportation research community, their queueing behaviour picture is still incomplete. Namely, there is no method addressed to analyse the full scope of these phenomena and to quantify their impact on the cyclist queue performance. To bridge this gap, we introduce the technique to observe multichannel queues and report relevant observations, which we then complement with a methodological framework to analyse obtained results and provide a complete multichannel queue description. We video-record cyclists as they enqueue to one of multiple channels, form the queue and smoothly merge into a single lane again as the queue discharges. We apply the method to analyse results from a pilot study of 160 cyclists forming 50 queues in the city of Krakow, Poland. The proposed method allows us to analyse and quantify the observed queue performance and its characteristics: the number of channels, their emergence process, channel and queue lengths, discharge process with FIFO violations, starting and discharging times. Findings from pilot study reveal that both queue length and discharge times strongly depend on queue formation process. The contribution of this paper is the method to describe multichannel cyclist queueing behaviour, enriching current picture of bicycle flow and cyclists’ behaviour. Since the method has been developed on relatively short queues (up to 10 cyclists), findings included in this paper primarily refer to such queue sizes. Nonetheless, the method is formulated in a generic way, applicable also for longer bicycle queues. Possible practical implications are new estimates for queue lengths and discharge times - useful for bicycle infrastructure design and traffic engineering purposes.
-
Method to Decompose Regional Travel Demand Model-Case Study of Kraków Region
Kucharski, Rafał,
Kulpa, Tomasz,
Mielczarek, Justyna,
and Drabicki, Arkadiusz
In Directions of Development of Transport Networks and Traffic Engineering: 15th Scientific and Technical Conference" Transport Systems. Theory and Practice 2018"
2019
In this paper we propose a new approach to model regional travel demand with a four-step model. Based on a comprehensive travel survey results for the Małopolska region in Poland (over 3mln inhabitants) we analysed the demand model with specific focus on regional trips. As a result, we introduce a modified model structure where long-distance trips are exposed. We breakdown the demand to four demand strata of various destination types to improve regional demand model quality. Such decomposition allows better representation of regional travel demand as compared to the classical four-step approach.
-
Metoda aktualizacji modelu podróży z wykorzystaniem macierzy przemieszczeń telefonów komórkowych
Kucharski, Rafał,
Mielczarek, Justyna,
Drabicki, Arkadiusz,
and Szarata, Andrzej
Transport Miejski i Regionalny
2018
W artykule prezentujemy metodę aktualizacji modelu podróży z użyciem macierzy przemieszczeń telefonów komórkowych. Proponujemy rozwiązanie dla sytuacji, gdy model popytu przestaje być aktualny, a jednocześnie dostępne są duże zbiory danych o faktycznych przemieszczeniach. Jako że budowa modelu popytu jest pracochłonna i kosztowna, a duże zbiory danych coraz bardziej dostępne, wskazane jest opracowanie metody aktualizacji. Jest to jednak problematyczne. Po pierwsze szczegółowość podziału obszaru na rejony jest zazwyczaj większa niż ta dostępna dla macierzy komórkowych, po drugie model popytu zawiera pełny opis mobilności (ruchliwość, motywacje, modele wyboru celu, środka podróży i trasy), a macierze komórkowe zawierają jedynie wielkości podróży. W niniejszym artykule proponujemy metodę wykorzystującą zalety obydwu źródeł danych. W szczególności korzystamy z formuł generacji ruchu, zmiennych objaśniających i struktury motywacji ujawnionych w badaniach, na podstawie których sformułowano model podróży. Dopełniamy te dane dwustopniowym modelem wyboru celu podróży opracowanym na podstawie struktury przemieszczeń telefonów komórkowych. Uzyskane wielkości podróży kalibrujemy do wielkości ujawnionych w macierzach, a następnie potoki pasażerów i pojazdów kalibrujemy do wyników uzyskanych w pomiarach. Propozycja ta nie eliminuje wszystkich problemów i niejasności, ale niewątpliwe pozwala na pełniejsze wykorzystanie tego cennego źródła danych. Metodę ilustrujemy wynikami aktualizacji małopolskiego modelu ruchu z użyciem macierzy przemieszczeń między powiatami całej Polski klientów jednego z operatorów. W efekcie model podróży oparty na zdezaktualizowanych wynikach badań został zaktualizowany do aktualnej struktury przemieszczeń i wielkości podróży, ale i sam opis mobilności stał się dokładniejszy. Dzięki wydzieleniu na etapie generacji ruchu liczby podróży: a) do Krakowa, b) do Tarnowa lub Nowego Sącza, c) międzypowiatowych d) wewnątrzpowiatowych e) zewnątrzwojewódzkich udało się uzyskać obraz podróży regionalnych zgodny z rzeczywistym.
-
Trip volume seasonal variations at regional level – case study of Małopolska GSM OD matrices
Kucharski, R.,
Szarata, A.,
Mielczarek, J.,
and Drabicki, A.
Archives of Transport System Telematics
2018
In this paper we analyze the big-data set of GSM origin-destination matrices collected between 360 districts (Powiat) in Poland to indirectly observe the trip volume fluctuations: weekly and seasonal. We utilized the dataset obtained from BTS position for all clients registered to a single provider. The data was collected over several days, which allows for a valuable temporal analysis of trip volumes. We analyze internal and external (inbound and outbound) trips in Małopolska region, intra-zonal trips (within district), inter-zonal (between districts of Małopolska). We discuss the general variability of observed trip volumes. We present the weekly fluctuations (working day, Friday, Saturday, Sunday) and the seasonal ones (winter, holiday, etc.). We verify the results obtained from GSM data with the traffic counts and their seasonal variations provided by national road administration (GDDKiA). Main contribution of the paper is presenting the observed fluctuations from the GSM data and comparing them with the classically collected data, as we demonstrate the results are comparable.
-
Real-time traffic forecasting with recent DTA methods
2017
In this paper we revisit the real-time traffic forecasting problem. We review recently proposed Dynamic Traffic Assignment (DTA) methods and verify how they can improve the practice of traffic forecasting. In particular, we analyze: 1) the Gradient projection DTA model of Gentile (2016), 2) Day-to-day model by Watling and Cantarella (2016), 3) the Marginal Computation (MaC) method by Corthout et al. (2014), 4) dynamic origin-destination (O-D) demand estimation methods (Kostic and Gentile, 2015) and 5) the event rerouting model (Kucharski and Gentile, 2014). We discuss how these methods can be applied to improve short-term forecasting and, most importantly, if they are efficient and mature enough for practical, real-time implementations. We formulate the real-time DTA forecasting problem which searches for the solution using all of the above DTA methods. The main contribution of this paper can be seen as a review and synthesis of recently proposed DTA methods, summarized with conceptual real-time forecasting framework.
-
Estimating Macroscopic Volume Delay Functions with the Traffic Density Derived from Measured Speeds and Flows
Kucharski, Rafał,
and Drabicki, Arkadiusz
Journal of Advanced Transportation
2017
This paper proposes a new method to estimate the macroscopic volume delay function (VDF) from the point speed-flow measures. Contrary to typical VDF estimation methods it allows estimating speeds also for hypercritical traffic conditions, when both speeds and flow drop due to congestion (high density of traffic flow). We employ the well-known hydrodynamic relation of fundamental diagram to derive the so-called quasi-density from measured time-mean speeds and flows. This allows formulating the VDF estimation problem with a speed being monotonically decreasing function of quasi-density with a shape resembling the typical VDF like BPR. This way we can use the actually observed speeds and propose the macroscopic VDF realistically reproducing actual speeds also for hypercritical conditions. The proposed method is illustrated with half-year measurements from the induction loop system in city of Warsaw, which measured traffic flows and instantaneous speeds of over 5 million vehicles. Although the proposed method does not overcome the fundamental limitations of static macroscopic traffic models, which cannot represent dynamic traffic phenomena like queue, spillback, wave propagation, capacity drop, and so forth, we managed to improve the VDF goodness-of-fit from of 27% to 72% most importantly also for hypercritical conditions. Thanks to this traffic congestion in macroscopic traffic models can be reproduced more realistically in line with empirical observations.
-
Modeling information spread processes in dynamic traffic networks
In 16th International Conference on Transport Systems Telematics, TST 2016
We propose the probabilistic information spread model to represent the spatiotemporal process of becoming aware while traversing the traffic network. In the contemporary traffic networks drivers are exposed to multiple traffic information sources simultaneously. Traffic managers look for a realistic estimate on when, where and how many drivers become informed about the actual traffic state (e.g. about the event). To this end we propose the probabilistic Information Spread Model (ISM) representing the process of spreading information to the drivers via multiple information sources (radio, VMS, on-line information, mobile applications, etc.). We express the probability of receiving information from a given information source using specifically defined spreading profile (formalized through the probability density function) and market penetration of respective source, with a novel information spreading model for on-line sources (websites, mobile apps, social networks etc.). Moreover, by assuming the information sources are mutually independent, the simplified formula for the joint probability can be used so that the model becomes practically applicable in real-time applications. Model is designed to work within the macroscopic dynamic traffic assignment (DTA) as a part of the network flow propagation model. Thanks to that, the informed drivers can be traced as they propagate through the network towards their destinations. We illustrate the model with the simulations on Dusseldorf network showing how information is spread in several ATIS scenarios (VMS, radio news, online sources, and simultaneous sources).
-
Modelowanie oporu skrzyżowań w modelach makroskopowych
Kucharski, Rafał,
Drabicki, Arkadiusz,
and Szarata, Andrzej
Transport Miejski i Regionalny
2016
Artykuł prezentuje metodę obliczania oporu relacji skrętnych na skrzyżowaniach dla makroskopowych miejskich modeli ruchu. Pokazano, jak na podstawie dostępnej bazy danych o warszawskiej sieci drogowej można zbudować model sieciowy wraz z określeniem przepustowości i czasów traconych na relacjach skrętnych. Wykorzystano w nim podstawowe formuły inżynierii ruchu (Gaca i in., 2008) pozwalające obliczyć przepustowość (poprzez oszacowanie natężenia nasycenia, udziału efektywnego sygnału zielonego, potoku nadrzędnego) oraz czasy przejazdu (swobodny oraz tracony) dla skrzyżowań sygnalizowanych, niesygnalizowanych i rond. Proponowana metoda jest uogólnieniem dostosowanym do dostępnej bazy danych i potrzeb makroskopowego modelu ruchu dużego obszaru (np. Aglomeracja Warszawska). Nie uwzględnia wszystkich czynników wpływających na przepustowość, jednak pozwala zbudować model sieciowy, w którym, tak jak w rzeczywistości, o czasie przejazdu i powstawaniu kolejek decyduje ograniczona przepustowość skrzyżowań, a nie odcinków. Metoda może być zastosowana w budowie modelu sieciowego dla dużego miasta przy użyciu dostępnej bazy danych bez znacznego zwiększenia czasów obliczeń. Wyniki metody ilustrują przykłady dla wybranych skrzyżowań w Warszawie.
-
Model wyboru środka transportu w dojazdach do i z pracy w Warszawie
Kucharski, Rafał,
Kulpa, Tomasz,
and Szarata, Andrzej
Transport Miejski i Regionalny
2016
Artykuł przedstawia model wyboru środka transportu przez podróżnych. Zaproponowano model wyboru dyskretnego: dwumianowy model logitowy, który określa prawdopodobieństwo wyboru w danej sytuacji jednego z dwóch rozważanych środków transportu (nazywanych dalej opcjami): komunikacji zbiorowej (KZ) i indywidualnej (KI). Sytuację, w której dokonywany jest wybór, opisuje odpowiednio zdefiniowana dla każdej z opcji użyteczność obejmująca m.in. dostęp do samochodu, czas przejazdu, liczbę przesiadek, częstotliwość kursowania. W artykule przetestowano kilka postaci modelu i oceniono ich dopasowanie do wyników prawie 7 tysięcy podróży do i z pracy, zbadanych w Warszawskim Badaniu Ruchu 2015 (WBR 2015). Parametry modelu szacowano przy użyciu pakietu do kalibracji modeli dyskretnych BIOGEME (Bierlaire, 2003), który szukał postaci gwarantującej największą zgodność modelu z faktycznymi wyborami podróżnych. Zaproponowano formuły o najwyższej zgodności uzyskanej przy użyciu zmiennych dostępnych w modelu i łatwo prognozowalnych. Efektem jest model, który objaśnia kiedy i dlaczego w dojazdach do pracy w Warszawie wybierana jest komunikacja zbiorowa, a kiedy indywidualna.
-
Multichannel Cyclist Queuing Behaviour at Signalised Cycle Crossings
Kucharski, Rafał,
Drabicki, Arkadiusz,
Kulpa, Tomasz,
and Szarata, Andrzej
In hEART 2016 - 5th Symposium of the European Association for Research in Transportation
2016
We observed cyclist queuing behaviour at signalised cycle crossings. We examined how they spontaneously form queues of various number of channels and physical length. We analyzed how the queue dissipates, showing that clearance times is strongly related with the queue formation process. We mapped cyclists’ trajectories on the space-time diagram to illustrate observed phenomena. Key findings from 50 queues observed in Kraków are: 1) physical length of five-cyclists queue vary from 7m for one channel to 2.5m for three channels 2) its clearance time drops from 16s for one channel to 11s for three channels 3) number of formed channels grows with number of queuing cyclists 4) cyclists minimizing their queuing time by forming multiple channels improve the overall crossing efficiency. This research contributes towards a developing stream of research on bicycle traffic flow. Increasing popularity of cycling, especially in modern-day urban areas, reinforce the need for in-depth understanding of bicycle traffic flow and cyclist behaviour, which is necessary to formulate an adequate approach towards the design of cycling infrastructure and traffic control systems.
-
Modelowanie wyboru środka transportu–porównanie regresji logistycznej i logitowego modelu wyboru dyskretnego
Kucharski, Rafał,
Szarata, A,
Bauer, M,
and Kulpa, T
X Poznańska Konferencja Naukowo-Techniczna, Poznań-Rosnówko
2015
Niniejszy artykuł jest teoretyczną analizą porównawczą dwóch modeli matematycznych używanych w modelowaniu wyboru środka transportu przez podróżnego. Opisujemy i porównujemy tu dwa modele: regresję logistyczną i model logitowy. Obydwa powszechnie stosowane w praktyce i często niesłusznie uznawane ze identyczne. Dzięki formalnemu zdefiniowaniu modeli możliwa jest ich analiza, porównanie założeń i przedstawienie interpretacji. W artykule tym pokazujemy różnice w założeniach modeli i ich konsekwencje. Obydwa modele określają prawdopodobieństwo wyboru danego środka transportu jako funkcję jego atrakcyjności. Wywodzą się jednak z różnych założeń. Model logitowy jest modelem wyboru dyskretnego, regresja logistyczna jest z kolei dopasowaniem funkcji jednej zmiennej do zaobserwowanego zjawiska. Pokażemy jakie są konsekwencje tych założeń w: doborze zmiennych objaśniających wybór (ograniczony w regresji i nieograniczony w modelu logitowym); liczbie alternatyw (dwie w regresji, dowolna ilość w modelu logitowym); postaci pochodnej, czyli wrażliwości na zmianę atrakcyjności (silnie nieliniowa i złożona w regresji) i procesie kalibracji (tradycyjny w regresji z wykorzystaniem zagregowanych obserwacji i symulacyjny z wykorzystaniem wszystkich obserwacji w modelu logitowym).
-
Observing rerouting phenomena in dynamic traffic networks
-
Direct observation of rerouting phenomena in traffic networks
Archives of Transport
2014
In this paper we propose how available dataset can be used to estimate rerouting phenomena in traffic networks. We show how to look at set of paths observed during unexpected events to understand the rerouting phenomena. We use the information comply model [1] and propose its estimation method. We propose the likelihood formula and show how the theoretical and observed rerouting probabilities can be obtained. We conclude with illustrative example showing how a single observed path can be processes and what information it provides. Contrary to parallel paper [2] where rerouting phenomena is estimated using real traffic flow measures from Warsaw, here we use only synthetic data. The paper is organized as follows. First we elaborate on rerouting phenomena and define the traffic network, then we summarize the literature behind rerouting phenomena. We follow with a synthetic definition of dynamic traffic assignment needed to introduce ICM model in subsequent section. Based on that introduction we define the observations and propose estimation method based on them followed by illustrative example. Paper is summarized with conclusions and pointing of future directions.
-
Modelowanie zjawiska zmiany trasy przejazdu w dynamicznym rozkładzie ruchu w sieci drogowej
In Zeszyty Naukowo-Techniczne Stowarzyszenia Inżynierów i Techników Komunikacji w Krakowie. Seria: Materiały Konferencyjne
2014
W artykule omówiono problem modelowania stanu sieci drogowej w następstwie sytuacji nietypowych, w szczególności w następstwie zdarzeń nieoczekiwanych (wypadków). Pokazano trudności jakie pojawiają się, gdy w dynamicznym modelu ruchu próbuje się uwzględnić reakcje użytkowników na zdarzenia nieoczekiwane. Zdefiniowano pojęcie ‘rerouting’, czyli zmianę trasy przejazdu w reakcji na informacje o zdarzeniach nieoczekiwanych i pokazano jak można je uwzględnić w dynamicznym modelu ruchu. Przedstawiono i omówiono istotę dwóch rozwiązań problemu: model przesuwającego się horyzontu, oraz model przyswajania informacji. Artykuł podsumowany jest prezentacją wyników modelu przyswajania informacji.
-
Indirect observation of rerouting phenomena in traffic networks - Case study of warsaw bridges
Archives of Transport
2014
In this paper we propose estimation procedure in which traffic flows resulting from rerouting model are matched with traffic flows observed during unexpected events. We show practical value of observing a entire cut-set of the transportation network and propose theoretical closed-form formulation of estimation problem for the rerouting model. We apply proposed framework on field-data from Warsaw bridges to observe rerouting phenomena. Most importantly we observed that: a) around 20% of affected traffic flow reroutes, b) rerouting flows are increasing in time, c) drivers show strategic capabilities, d) and maximize their utility while rerouting. All of the which were hypothesized in Information Comply Model (Kucharski et. al., 2014) and are now supported with field observations.
-
Makroskopowy model przepływu ruchu w sieci drugiego rzędu–alternatywny opis stanu sieci
Kucharski, Rafał
In Wydajność systemów transportowych, Materiały IX Konferencji naukowo-technicznej
2013
Macroscopic volume-delay functions are widely criticized due to their inability to reproduce fundamental traffic phenomena, i.e. queue formation and dispersion, capacity drop, within-day temporal changes, spillbacks, etc. Microsimulation is widely recognized as the only alternative. In this paper another approach is shown: second order macroscopic traffic flow models. General Link Transmission Model described here can reflect above mentioned phenomena with much shorter computation time and requiring much smaller input database than microsimulation. In author’s opinion results of GLTM are more valuable for traffic engineer than results of microsimulation, which are typically stochastic. GLTM can be computed for real-size network in acceptable time.
-
Gromadzenie danych do budowy modeli ruchu-przegląd możliwości
Kucharski, R
Transport Miejski i Regionalny
2013
W polskich warunkach możliwa jest poprawa jakości modeli ruchu , a jednym z potencjalnych sposobów poprawy jest zgromadzenie i wykorzystanie w modelowaniu baz danych. Baza danych wejściowych to kluczowa sprawa dla jakości modelu ruchu. Jeśli model ma spełniać swoje zadanie, to musi być zbudowany w oparciu o bogatą i dokładną bazę danych. Równie istotną sprawą jest jej regularna aktualizacja , a jej zmiany powinny pociągać za sobą aktualizacje modelu. W artykule autor opisuje, jakie dane są dostępne i w jaki sposób można je wykorzystać w modelowaniu ruchu. Na kilku przykładach pokazano w praktyczny sposób jak można wzbogacić bazę danych używanych do modelowania ruchu. Opisano cztery przykłady dobrych praktyk, w szczególności: protokół przesyłu danych o sieci i rozkładzie jazdy transportu zbiorowego z serwisu jakdojade.pl do programu do modelowania ruchu, możliwość użycia danych przestrzennych GIS do tworzenia modelu sieci transportowej, możliwość użycia danych przestrzennych z podkładów GIS do obliczania produkcji i atrakcji rejonów komunikacyjnych w jednym z programów do modelowania ruchu oraz wykorzystanie wyników automatycznej rejestracji tablic rejestracyjnych (ARTR) w modelowaniu. Na wielu konferencjach dotyczących modelowania ruchu powtarzana jest teza, że najpilniejszą sprawą w polskim modelowaniu jest problem pozyskania danych. Poniższe akapity pokazują jak można ten stan rzeczy małymi krokami zmieniać. Przykłady te powinny służyć jako dobra praktyka w budowie modeli ruchu. Należy, na poziomie definiowania zamówienia na budowę modelu ruchu, zapewnić jak największą ilość i jakość danych wejściowych. Dzięki temu modele będą mogły pełnić istotną rolę w kształtowaniu systemu transportowego miasta, tworząc wspólną platformę zarządzania transportem.
