SAEnet: an application for adjustment of horizontal network using particle swarm optimization algorithm

Abstract

The horizontal network of control stations plays an important role in geometric measurement activities, such as determining positions of point, distance, area, or even volume. One of the methods of determining horizontal network values is the least squares method. The method of least squares is one of the most frequently used methods to obtain unique estimates for a set of redundant measurements. The equations that are used in the horizontal network surveys are the nonlinear equations so it must be linearized in order to be computed using least squares method. The problem that arises when linearizing the nonlinear observation equation is the approximate coordinates of the network stations are needed for its computations. In order to handle this problem, the Particle Swarm Optimization (PSO) algorithm is applied as an algorithm for determining the approximate coordinate values of the network stations. The algorithm of PSO is a heuristic function that optimizes a problem by iteratively trying to improve a candidate solution toward the best solutions. The advantages of PSO are that PSO is designed for solving a problem more quickly when classic methods are too slow and finding an approximate solution when classic methods fail to find any exact solution. This paper presents the design of an application for horizontal network adjustment using PSO algorithm, namely SAEnet. The Algorithm was written in Matlab which is very effective in matrix manipulation and has been widely used in various fields of numerical computations. The computation of horizontal network adjustment using SAEnet application consists of determination of initial approximation values using PSO and the adjustment computation using the least squares method. SAEnet is also equipped with save menu to Microsoft Excel format and load menu from Microsoft Excel format. The test shows that the input and output of the program can run as expected. The results are then compared to other program, like STAR*NET, using a variety of data sets. The results of both computations show the maximum difference between both programs is 0.000040 while the average difference is 0.000020. The test shows that there is no significant difference between the results of the SAEnet and STAR*NET computation. Hence, SAEnet can be used for horizontal network adjustment computation.