Master thesis presentation Viktor Wiklund & Anton Rosenberg, MPCAS

Abstract: Correctly calculating routes and the future positions of emergency vehicles is important for traffic safety and can reduce response times. Carmenta Automo- tive is a company aiding the NordicWay3 project, funded by the EU, and aims to create a safer traffic environment. Their objective is to anticipate the path and location of emergency vehicles to send warning messages to road users. This the- sis is a continuation of a previous thesis in which an algorithm for predicting the routes and positions was implemented. However, no data or metric was used to validate the developed algorithms’ performance. In this thesis, a SQL database was created to contain actual mission data of emergency vehicles, which was used for validation purposes. Two validation metrics were also developed, one focusing on more extended missions and selecting correct routes, the other on the mean distance between the actual and predicted positions in a shorter time frame. Furthermore, parameters were optimized with three weight equations to determine the probability of choosing a particular road. The parameter was optimized using Competetive Swarm Optimization considering both validation metrics, and a maximum improvement of 7% was found using this setup. Fi- nally, it was concluded that a cost calculation that takes more aspects of the road than its size, length, and speed limit is required to improve the algorithm’s performance further.

Supervisor: Anna Brunzell

Examiner: Giovanni Volpe

Opponents: Jesper Jäghagen & Carl Hjalmarsson