doktorska disertacija
Anamarija L. Mrgole (Author), Drago Sever (Mentor), Rok Strašek (Co-mentor)

Abstract

Prometni tok je gibanje motornih vozil na cestni mreži in je kot tak spremenljiv, dinamičen in kompleksen sistem, ki je nelinearen in nepredvidljiv. Pojav zgostitve prometnega toka v cestnem prometu ocenjujemo, ko se prometna obremenitev na določenem odseku prometnice in v določenem časovnem obdobju približa ali prekorači kapaciteto cestne infrastrukture, zato je pod določenimi pogoji mogoče opaziti, v zgostitvi prometnega toka, kaotične parametre. Pregled literature o prometnem toku in povezavi z kaosom nakazuje, da ima tema veliko teoretično in praktično vrednost. Raziskane metode za identifikacijo kaosa v prometnem toku so pokazale omejitve do sedaj uporabljanih metod, hkrati pa nakazale usmeritve za možnost izboljšave identifikacije kaotičnih parametrov v prometnem toku. Predlagana nova metoda kratkoročnega napovedovanja zgostitev v prometnem toku uporablja frekvenčno Wigner-Villerjevo porazdelitev, ki omogoča prikaz kaotičnega atraktorja brez uporabe rekonstrukcije faznega prostora, s tem je mogoče skrajšati čas izračuna napovedi zgostitve prometnega toka in zmanjšati število uporabljenih časovnih vrst. Rezultati eksperimentalnega dela naloge prikazujejo verodostojnost modela identifikacije kaosa in napovedovanja zgostitev v prometnem toku, s primerjavo dobljenih rezultatov kaotične nevronske mreže in usmerjene nevronske mreže, se pokaže da z izpuščenim korakom rekonstrukcije faznega prostora dobimo natančnejšo kratkoročno napoved. Na osnovi dobljenih rezultatov je mogoče sklepati, da je nov model kratkoročnega napovedovanja zgostitev prometnega toka zanesljiv, natančen, verodostojen in veljaven.

Keywords

nelinearni dinamični sistem;Wigner-Ville distribucija;prometni tok;napovedovanje prometa;zgostitev prometnega toka;analiza časovnih vrst;aidentifikacija kaotičnega parametra;nevronske mreže;doktorske disertacije;

Data

Language: Slovenian
Year of publishing:
Typology: 2.08 - Doctoral Dissertation
Organization: UM FGPA - Faculty of Civil Engineering, Transportation Engineering and Architecture
Publisher: [A. Ljubič Mrgole]
UDC: [519.216:519.246.8]:656.1.05(043.3)
COBISS: 18862870 Link will open in a new window
Views: 1732
Downloads: 154
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Short-term prediction of traffic flow congestion using time series analysis of nonlinear dynamic system
Secondary abstract: Traffic flow is variable, dynamic and complex system, which is non-linear and unpredictable. We estimate the emergence of traffic flow congestion in road traffic when the traffic load on a specific section of the road in a specific time period is close to exceeding the capacity of road infrastructure. Under certain conditions, it can be seen in concentrating chaotic traffic flow parameters. A literature review on the traffic flow and chaotic features implies that this kind of method has great theoretical and practical value. Researched methods for identification of chaos in traffic flow have shown certain restrictions techniques, but suggested guidelines for improving the identification of chaotic parameters in traffic flow. The proposed new method of forecasting short-term congestion in traffic flow is using Wigner-Viller frequency distribution. The method enables the display of a chaotic attractor without the use of reconstruction phase space. It is possible to shorten the time of the forecast calculation denser traffic flow and reduce the time series. The results of the experimental work shows the credibility of the model identification and prediction of congestion chaos in traffic flow. Comparison of the results obtained chaotic neural network and targeted neural networks shows that with the missed step we have a more accurate short-term forecast. Based on the results it can be concluded that the new model of forecasting short-term traffic flow congestion is reliable, accurate, credible and valid.
Secondary keywords: non-linear dynamic system;Wigner-Ville distribution;traffic flow;alumintraffic forecasting;congestion of traffic flow;identification of chaotic parameter;neural network;Cestni promet;Disertacije;Napovedovanje;
URN: URN:SI:UM:
Type (COBISS): Doctoral dissertation
Thesis comment: Univ. v Mariboru, Fak. za gradbeništvo, prometno inženirstvo in arhitekturo
Pages: 110 str.
ID: 8751919