Optimizing Road Network Weight Calculation for Emergency Vehicles
Ladda ner
Publicerad
Författare
Typ
Examensarbete för masterexamen
Master's Thesis
Master's Thesis
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Correctly calculating routes and the future positions of emergency vehicles is impor tant for traffic safety and can reduce response time during distress calls. Carmenta
Automotive is a company that supports the EU-funded NordicWay3 project, which
aims to create a safer traffic environment. Their objective in the project is to antic ipate the path and location of emergency vehicles with the goal of sending warning
messages to road users.
This thesis is a continuation of a previous thesis in which an algorithm for pre dicting the routes and positions of emergency vehicles was implemented. However,
no data or metric was used to validate the developed algorithm’s performance. In
this thesis the goal was to develop validation metrics for this algorithm and try to
optimize the cost of selecting each road, called its weight, in the road network. To
achieve this an SQL database was created containing the mission data of emergency
vehicles, which was used for validation purposes. Two validation metrics were also
developed, the first uses the mean distance between the actual and predicted posi tions in a shorter time frame, the second focusing on more extended missions and
selecting correct routes. Furthermore, parameters were optimized for three different
weight equations, which are used to calculate the weight of each road using infor mation about its speed, size and distance. The parameters were optimized using
Competitive Particle Swarm Optimization with respect to one validation metric at
a time. Through this process a maximum improvement of 4% and 7% compared to
the previous thesis weight equation with respect to the first and second validation
metrics was found respectively. Finally, the conclusion reached was that the goal of
the thesis was accomplished and in order to further improve the algorithm’s perfor mance, a weight calculation considering more aspects of the road than just its size,
length, and speed is required.
Beskrivning
Ämne/nyckelord
algorithm, PSO, CPSO, A-star, stochastic, optimization, route, predic tion.