COVERING A MAXIMUM NUMBER OF POINTS BY A FIXED NUMBER OF EQUAL DISKS VIA SIMULATED ANNEALING

Abstract

The presented paper considers the problem of covering a maximum number of n given points in the plane by m equal disks of radius r. A point is covered if it is inside one or more than one disk. The disks need to be placed in the plane in such a way that a maximum number of points are covered. To solve the problem, an objective function, called energy, is introduced in such a way that the greater the covering is, the lower the energy is. Thus, a configuration of disks with minimum energy is a configuration with maximum covering. To find a configuration of disks that minimizes the energy, a stochastic algorithm based on the Monte Carlo simulated annealing technique is proposed. The algorithm overcomes potential local minima, which, as shown in the paper, are quite likely to occur. The computational complexity of the algorithm is O(mn). The algorithm is tested on several cases demonstrating its efficiency in finding global minima of the energy, i.e. configurations with maximum covering.