Optimization of Weather Radar Network using PSO

To study weather phenomena over a landscape bounded by a regional boundary, a decision support system is needed for optimally placing the radars of a Weather Radar Network (WRN). Optimization is to be achieved in terms of higher area coverage by radar beam when obstructed through ground elevations in mountainous region and in low cost of radar installations. Traditional exhaustive search algorithm can't be useful here as it would computational infeasible to find out the optimum positions from a huge number of possible radar site locations. In this work applies Particle Swarm Optimization (PSO) to address the complex multi-dimensional optimization problem of WRN. The approach considers several factors in optimization such as radar beam blockage due to complex terrain, limitations of radar range from low level ground based coverage, radar locations confined to a regional boundary, cost of radar installations in terms of number of radars and radar's proximity to nearest road network. The PSO based approach has unique strength of addressing multiple dimensions of the optimization problem together and it provides a simple way for reaching an optimal solution in relatively short span of time.