doktorska disertacija
Robert Rijavec (Avtor), Igor Grabec (Mentor), Peter Lipar (Član komisije za zagovor), Tomaž Maher (Član komisije za zagovor), Drago Sever (Član komisije za zagovor), Marijan Žura (Komentor)

Povzetek

Zanesljivost podatkov in ustrezna kratkoročna napoved prometa imata zelo pomembno vlogo pri upravljanju avtocestnega prometa, še posebej v povezavi s sistemom nadzora in vodenja prometa ter sistemom informiranja udeležencev o izrednih dogodkih v prometu. Temeljijo na modelih, ki se učijo z uporabo značilnih vzorcev prometnih stanj, podanih z različnimi karakteristikami prometnega toka v času in prostoru, ki smo jih imenovali profili prometnega toka. Njihov pomen je povezan z značilnimi dogodki v preteklosti, kar bo aktualno, vse dokler ne bo znano prometno povpraševanje, ki bo določeno z najavo prihoda posameznega vozila. Tako smo analizirali različne poznane metode gručenja in predlagali najprimernejšo, s katero želimo določiti med seboj izključujoče, homogene skupine značilnih dogodkov - profile značilk karakteristik prometnega toka s poudarkom na dnevnem profilu pretoka vseh vozil ter težkih vozil, neodvisno od mesta meritve in družbenega okolja. Določitev skupin profilov prometnega toka in njihovih značilk je eno izmed dejanj, s katerimi lahko določimo poljuben matematični prometni model obstoječega prometnega stanja v znanih cestno-vremenskih razmerah. A prometnega modela ni možno definirati brez poznavanja razporeditve in razpršenosti karakteristik prometnega toka po prometnih pasovih smernega vozišča avtoceste, ki smo ga poimenovali prečni profil prometnega toka. Iz obeh profilov se lahko naučimo, kakšne so skrajne vrednosti karakteristik prometnega toka v času in prostoru. Na podlagi teh in glede na dejavnike okolja lahko določimo, preverimo in spreminjamo merila vodenja prometa. Tako je izvedena empirična raziskava, s katero smo določili model vpliva cestno-vremenskih razmer na prečni profil prometnega toka. Z uporabo mikrosimulacijskega prometnega modela smo prikazali pomen natančne karakterizacije prometnih tokov za potrebe sistema za nadzor in vodenje prometa na avtocesti. Zamude, ki smo jih izračunali s prometnim modelom, vgrajenim v računalniški program PTV Vissim, smo primerjali z referenčnim scenarijem brez vodenja prometa, a pri različnih cestno-vremenskih in enakih prometnih pogojih. Zamude so na določenem območju avtocestne mreže lahko mera uspešnosti vodenja prometa in hkrati kazalnik intenzivnosti zastojev, zato je pomembno poznavanje rezultata, da enako vrednost mere uspešnosti dosežemo pri različni vrednosti gostote prometnega toka ob različnih cestno-vremenskih razmerah.

Ključne besede

Grajeno okolje;gradbeništvo;disertacije;karakterizacija prometa;avtoceste;SNVP;profili prometnega toka;časovne vrste;porazdelitev prometa po pasovih;prometni koledar;vremenski pojavi;

Podatki

Jezik: Slovenski jezik
Leto izida:
Tipologija: 2.08 - Doktorska disertacija
Organizacija: UL - Univerza v Ljubljani
Založnik: R. Rijavec]
UDK: 656.11:625.711.1(043)
COBISS: 9026401 Povezava se bo odprla v novem oknu
Št. ogledov: 737
Št. prenosov: 216
Ocena: 0 (0 glasov)
Metapodatki: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Ostali podatki

Sekundarni jezik: Angleški jezik
Sekundarni naslov: Motorway traffic characterisation for the Slovenian traffic control system
Sekundarni povzetek: The reliability of traffic data and useful short-term traffic forecasts play a very important role in the management of motorway traffic, especially in connection with real-time traffic control systems and the provision of traffic and travel information to road users regarding unexpected incidents. What follows are learning based models which use typical traffic patterns that have different traffic flow characteristics in relation to time and space, and which we refer to as "profiles of traffic flow characteristics" or in short "traffic profiles". Their significance is related to distinct historical events, which are current until the known traffic demand is determined, which will be determined by the arrival of a particular vehicle. Therefore, we analysed various known clustering methods and proposed the one we considered most suitable to define specific groups of typical and representative time-based profiles of traffic flow characteristics with an emphasis on the daily traffic flow profile of all vehicles and heavy vehicles, regardless of the location of the measurement station and the social environment. Finding representative profiles of traffic flow characteristics is one of the steps by which we can determine any mathematical traffic-state estimation model for predicted road and environmental conditions. A traffic model cannot be defined without knowing the lane flow distribution, the speed distribution and variability of other traffic flow characteristics across the traffic lanes of the motorway in both directions, which we refer to as "lane distribution of motorway traffi"c o"r cross-sectional traffic profile". From both profiles we can learn the peak values of traffic flow characteristics in terms of time and space. Based on these and their relationship to environmental factors, we can determine, review and update traffic control algorithms and criteria. Accordingly, an empirical study was carried out to determine the effect of weather conditions on the lane distribution of motorway traffic. Using a micro-simulation traffic model, we showed the importance of precise traffic characterisation for the needs of the traffic control system on motorways. Delays, calculated with the PTV Vissim traffic simulation model, were compared with the results obtained using reference scenarios without traffic control, but under different road and weather conditions and in the same traffic conditions. If delays in a certain area of a motorway network are a measure of the effectiveness of traffic control, and at the same time an indicator of the degree of congestion, then it is important to know the result that the same value of the measure of effectiveness is achieved at different traffic flow density values under different weather conditions.
Sekundarne ključne besede: Built Environment;civil engineering;doctoral thesis;traffic characterisation;motorway traffic control;traffic profiles;time series;lane flow distribution;traffic calendar;weather conditions;Avtoceste;Disertacije;Promet;Karakterizacija;
Vrsta dela (COBISS): Doktorsko delo/naloga
Študijski program: 0
Konec prepovedi (OpenAIRE): 1970-01-01
Komentar na gradivo: Univ. v Ljubljani, Fak. za gradbeništvo in geodezijo
Strani: XXXII, 163 str.
ID: 11372672