BOMBOL, Kristi M. - Prvi autor
Vukanovic, Smiljan M. - Drugi autori
Cucuz, Nikola - Drugi autori
Milosevic. Stanisa - Mentor

Monografska publ./Monography

Macedonian/makedonski

[K. Bombol]

1995

Bitola

165 listova : ilustr. ; 30 cm

Na makedonskom jeziku. - Umnozeno za odbranu. - Univerzitet "Sv. Kliment Ohridski", Tehnicki fakultet, Soobrakaen otsek, Bitola, odbranjeno 18.01.1996. - Bibliografija: listovi 119-127. - Registar. - Abstract.

The issue presented in this dissertation is due to my staz as a British Council scholarship holder for miscellaneous research student at the University in Leeds (UK), i.e. at the Institute for Transport Studies.The origin of the idea to do such a research work came out from the fact that many of traffic accidents happen on two-lane roads. Most European contries face the fact thet traffic accidents occur on two-lane roads where the central lines effectiveness is jet to be established. The problem complexity consequence results in urgent need for undertaking rational measures which assume a modern approach, i.e. road users' behaviour knowledge.The thesis is based on empirical experiences obtained from the experimental investigations done on a driving simulator "Rover" as a very updated and sophisticated tool for evaluation purposes. Pioneers' steps have been made to a modern approach establishing the evaluation procedures of central line markings (CLM) impact on driver behaviour and driving safety. The problem identification is being established from the road users' behaviour complexity and variability. Therefore the CLM safety effectiveness is to be established in the context of driver behaviour and driving safety.The research issue is defined as a drivers adaptation behaviour on a 2-lane road with/without CLM. The research objectives are being defined as to obtain scientific based information on driver behaviour on a 2-lane rural road with/without CLM, and to evaluate the CLM effectiveness on driver behaviour and driving safety.The methodology applied is based on experimental investigation on a driving simulator which allows great research opportunities in a domain of evaluation procedures. The conceptualisation of experimental work is based on the idea to get a random sample of 26 drivers volunteers (18 males and 8 females) of different nationalities, with different driving experience and age. A 2-lane rural network about 7 km long was designed. The total number of sections (straight, curves, intersections) was 29. Different types of road central lining were applied (double solid, dashed, hazardous, dash/solid). The car was activated by electrical power engine and was equipped with brakes, clutch, acceleration pedal as a real vehicle. The road scenario was projected on a wide screen in front of the driver by computer generated images.The drivers had 4 experiments each, i.e. twice on a road with/without CLM. The observation of driver behaviour implied driving on straight sections, in curves and approaching junctions. There were 3 dependant variables of interest: speed, lateral displacement and time in the opposite lane. The independent variables were defined as: road width, CLM type, section type (straight/curves), drivers' sex.In data analysis statistical methods in traffic engineering have been applied. The statistical procedures applied are: 85th percentile speed, means before/after CLM implementation, multivariate model of analysis, ANOVA (Analysis of Variance), hypothesis testing and PROBIT analysis. The main assumption made was that speeds, lateral displacements and the time in the opposite lane are normal distributed values which allow the application of above mentioned statistical procedures. Main effects, interactions among various factor varialbes combinations, factor level comparing contrasts and the probability of going out of lane were being investigated.It was concluded that there are changes in driver behaviour under conditions no/yes line markings in terms of speed changes, overtaking manoeuvres, i.e. lateral displacements and the probability of going out of lane. No exact data are being obtained for driving in curves. All the results apply to driving on straights. It suggested that further research on mathematical driver behavioural models and empirical investigation for driving in curves are to be carried out. The methodology of process evaluation procedure was presented.The directions and recommendations for further work in terms of driver behaviour response on other horizontal lining changes, drivers' making decisions in judgement and gap acceptance, using the driving simulator for research in curves and evaluation system modelling for drivers, vehicles and road network using advanced transport technologies are given.