Enhancing public transportation time from Stavanger to Forus

Abstract

This thesis investigates how heuristic methods and simulation modeling can enhance the efficiency of rush hour travel times for public transportation from Stavanger to Forus. With Forus being a big business area, it is important to have a route that is time effective and connects it with the city centre of Stavanger.

The master’s thesis applies a simulation model to imitate the real world actions of route 3. Distributions matching the movements of passengers and bus drivers were used to simulate the time spent waiting and stopping at each stop. Equations were used to approximate driving time in between stops, and time spent in the different stages. Highlighting how individuals time perception differs depending on the way of travel. Heuristic methods were used for minimization, due to its ability to closely mimic nature. Of two potential heuristic methods, simulated annealing was the preferred heuristic method for reducing the number of stops along the route and minimizing weighted distance to a stop. Genetic algorithm was discarded due to difficulties of escaping local optima.

The regression was fitted with two independent variables, number of stops and average waiting time, explaining the dependent variable, total travel time for the whole selected part of route 3. Average walking time and bus travel time were affected by number of stops, while average waiting time was given by the frequency of buses. The regression shows that reducing the number of stops only marginally decreased total travel time. On the other hand, average waiting time affected the total travel time more influentially. Thus, change in frequency is the public transport agencies’ greatest tool in controlling the total travel time. The total number of stops can be utilised to reduce total travel time, but at the cost of coverage and will only gain a slight reduction in time.

This thesis investigates how heuristic methods and simulation modeling can enhance the efficiency of rush hour travel times for public transportation from Stavanger to Forus. With Forus being a big business area, it is important to have a route that is time effective and connects it with the city centre of Stavanger.

The master’s thesis applies a simulation model to imitate the real world actions of route 3. Distributions matching the movements of passengers and bus drivers were used to simulate the time spent waiting and stopping at each stop. Equations were used to approximate driving time in between stops, and time spent in the different stages. Highlighting how individuals time perception differs depending on the way of travel. Heuristic methods were used for minimization, due to its ability to closely mimic nature. Of two potential heuristic methods, simulated annealing was the preferred heuristic method for reducing the number of stops along the route and minimizing weighted distance to a stop. Genetic algorithm was discarded due to difficulties of escaping local optima.

The regression was fitted with two independent variables, number of stops and average waiting time, explaining the dependent variable, total travel time for the whole selected part of route 3. Average walking time and bus travel time were affected by number of stops, while average waiting time was given by the frequency of buses. The regression shows that reducing the number of stops only marginally decreased total travel time. On the other hand, average waiting time affected the total travel time more influentially. Thus, change in frequency is the public transport agencies’ greatest tool in controlling the total travel time. The total number of stops can be utilised to reduce total travel time, but at the cost of coverage and will only gain a slight reduction in time.