Abstract
We describe mathematical models and practical algorithms for a problem concerned with monitoring the air pollution in a large city. We have worked on this problem within a project for assessing the air quality in the city of Rome by placing a certain number of sensors on some of the city buses. We cast the problem as a facility location model. By reducing the large number of data variables and constraints, we were able to solve to optimality the resulting MILP model within minutes. Furthermore, we designed a genetic algorithm whose solutions were on average very close to the optimal ones. In our computational experiments we studied the placement of sensors on 187 candidate bus routes. We considered the coverage provided by 10 up to 60 sensors.
Highlights
A current project is concerned with the quality of air in the city of Rome
We decided to run a set of preliminary tests in which we tried to get an estimate of the running time for the Mixed Integer Linear Programming (MILP) and the genetic algorithm
In this paper we have investigated a particular problem of sensor placement originated by a real application
Summary
A current project is concerned with the quality of air in the city of Rome. Part of this project is devoted to measuring the air pollutants in the city. In order to reduce the costs and at the same time assure a good significance to the measurements, it has been decided to mount the sensors on the public buses. These sensors measure the pollutants and immediately send the data to a central station, while they are moving within the city according to the bus route. Given that there are only a relatively small number of available sensors, one needs to choose which buses to place the sensors on in order to get the best possible covering of the city area
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
