Design of global positioning system (GPS) networks using different artificial intelligence techniques

Abstract

The selection of optimal GPS baselines can be realized by solving the geodetic second-order design (SOD) problem. Basically, there are two techniques to be used for selecting optimal baselines in GPS network, namely, traditional techniques and artificial techniques. Traditional techniques include the method of trial and error and the analytical method, while artificial methods include both local and global optimization techniques. The global optimization techniques, such as Genetic Algorithms (GAs), Simulated Annealing (SA) method, Particle Swarm Optimization (PSO) Algorithm, and Butterfly Optimization Algorithm (BOA), have been used recently in geodesy. In the current study, BOA has been used for the selection of the optimal GPS baselines to be measured in the field that will meet the postulated criterion matrix, at a reasonable cost. It has been tested on a GPS network. The BOA is already designed, and it determined the number of baselines that would be observed because of obtaining high accuracy. The results showed that the BOA method was more efficient than the traditional ones by 19.2%. It was better than the artificial methods in terms of length as it enhanced SA method by 21.7% and PSO method by 4.6%. Consequently, the use of the BOA is proven to be more effective and applicable.