The relationship between pedestrian crash rates and injuries sustained and the location, time of day, traffic volume and the hypothetical travel speed at the time of the crash on the Gauteng Freeways

Snijder, Janneke (2019-12)

Thesis (MEng)--Stellenbosch University, 2019.

Thesis

ENGLISH ABSTRACT: Informal settlements are situated adjacent to the freeway, resulting in pedestrians that cross and walk along the South African freeways regularly. This phenomenon is due to the historical and current demographics, where certain racial groups were forced to move out to the city perimeter. The current economic situation results in informal settlements that grow rapidly. The low-income areas are not provided with proper transport facilities, forcing the residents of the settlements to walk to their destinations. Research indicates that pedestrians tend to walk the shortest path to their destination, resulting in pedestrians that cross the freeway to get to their destinations. The aim of this study is to determine the impact of the land use surrounding pedestrian crashes, the day and time on which the pedestrian crash occurred, the traffic volume and vehicle speed, the vehicle type involved in the crash as well as the number of lanes to cross have on the severity as well as the crash rate of the pedestrian crashes on the Gauteng FMS network. The injuries sustained in the pedestrian crashes was modelled using a binomial logistic model. The crash rate of the pedestrian crashes was modelled using a multinomial logistic model. It was determined that the independent variables for the two models are different, and a clear relationship between the two models was found. The traffic volume on the road, the average travel speed, the day of the crash, the vehicle type involved in the crash as well as the number of lanes to cross had an impact on the probability to sustain fatal injuries in a pedestrian crash. The probability of sustaining fatal injuries increase during the weekend, during the off and night peaks, when the travel speed increase, when the number of lanes to cross increase as well as for the case when a heavy vehicle is involved in the crash. The probability of sustaining fatal injuries decreases as the number of vehicles on the road decreased. The crash rate, on the other hand, was dependent on the traffic volume, vehicle speed, land use, day of the crash, as well as the peak hour in which the crash occurred. It was, however, also determined that the land use attribute might cause this model to predict biased results, and this model was therefore not regarded as an accurate model. The injury model was regarded as a good model, and it was concluded that this model could be applied in other locations in South Africa.

AFRIKAANSE OPSOMMING: Informele nedersettings is geleë aangrensend aan die snelweg, wat lei tot voetgangers wat op ‘n gereelde basis die Suid-Afrikaanse snelweë oorsteek en langs die snelweë loop. Dit is as gevolg van die historiese en huidige demografie waar sekere rassegroepe gedwing is om na die stad grense te skuif. Die huidige ekonomiese situasie lei tot informele nedersettings wat vinnig groei. Die lae-inkomste gebiede is nie voorsien met 'n goeie vervoer fasiliteite nie, en dwing die inwoners van die nedersettings om na hul bestemmings te loop. Navorsing dui daarop dat voetgangers geneig is om die kortste pad na hul bestemming te loop, wat lei tot voetgangers wat die snelweg oorsteek om hul bestemmings te bereik. Die doel van hierdie studie is om te bepaal wat die impak van die grondgebruik rondom voetgangerongelukke, die dag en tyd waarop die voetgangerongeluk plaasgevind het, die verkeersvolume en voertuig se spoed, die tipe voertuig wat betrokke is in die ongeluk asook die aantal lane om oor te steek op die erns van die voetgangerongeluk asook op die ongelukskoers van die voetgangerongelukke op die Gauteng FMS netwerk het. ‘n Binomiale logistiese model is ontwikkel vir die beserings wat opgedoen is in die voetgangerongelukke en 'n multinomiale logistiese model is ontwikkel vir die ongelukskoers model. Daar is vasgestel dat die onafhanklike veranderlikes vir die twee modelle verskil en dit het tot die gevolgtrekking gelei dat daar nie 'n duidelike verband tussen die twee modelle is nie. Die verkeersvolume op die pad, die gemiddelde reisspoed, die dag van die ongeluk, die tipe voertuig wat betrokke is in die ongeluk asook die aantal lane om oor te steek het 'n impak op die waarskynlikheid om noodlottige beserings in 'n voetgangerongeluk op te doen. Die waarskynlikheid vir noodlottige beserings verhoog tydens die naweek, tydens die af en nag pieke, wanneer die voertuig se spoed toeneem, wanneer die aantal lane om oor te steek verhoog sowel as vir die geval wanneer 'n swaar voertuig betrokke is in die ongeluk. Die waarskynlikheid vir noodlottige beserings verminder as die aantal voertuie op die pad afgeneem. Die ongelukskoers, aan die ander kant, was afhanklik van die verkeersvolume, spoed van die voertuig, grondgebruik, dag van die ongeluk, sowel as die piek uur waarin die ongeluk plaasgevind het. Dit is egter ook bepaal dat die grondgebruik kan veroorsaak dat hierdie model onakkurate resultate voorspel en hierdie model is dus nie as 'n akkurate model beskou nie. Die beseringsmodel was beskou as 'n goeie model en dit is tot die gevolgtrekking gekom dat hierdie model in ander plekke in Suid-Afrika toegepas kan word.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/107080
This item appears in the following collections: