Application of Enhanced Particle Swarm Optimization for Determining the Optimal Geometry of Horizontal Curves

Geometry of horizontal curves is an important issue in the road alignment and it requires applying methods for determining the curve recognition. The curve fitting is an approach tointegrate the in-filed collected points with acceptably accuracy and usually conducts by empirical selection of curves using the trial-and error-technique. In this paper, the EnhancedParticle Swarm Optimization (EPSO) is applied for determining the optimal geometry ofhorizontal curves. The Global Positioning System (GPS) data collection of horizontal curves in a rural road consider for studying on effect and validation of proposed approach in 27curves with various characteristics in the considered road as case study. The EPSO with an objective function as minimization of total least squares is coded in an m-file of MATLAB software. A comparison is conducted between the results of Auto Cad Civil 3D and EPSOin fitting the curves and EPSO showed very efficient to the Civil 3D. The results of this research can to present an automated framework for determining and implementing the horizontal curves. This approach shows that reliable for road geometry recognition of curves.

Keywords:

Horizontal curves , Road recognition , Global positioning system (GPS) , Total least squares , Enhanced particle swarm optimization (EPSO)