GA optimization technique’s in interpolation for dynamic GIS

R Naveen Kumar Goud
11-1-315/1, Flat no : 304, SVS Residency,
Opp : Pragathi High School, Mylargadda, Sithaphalmandi, Secunderabad, Andhra Pradesh -- 500061.
E-Mail: naveengoud_r@rediffmail.com


Abstract
Generating cross-sectional data of arbitrary time slice is a basic function to support temporal operations such as time-series analysis and integration of dynamic models in Global GIS environment. To generate time slice data, it is necessary to interpolate/integrate existing global data with different temporal coverage and spatial resolution. Although interpolate/ integrate methods have been developed for continuous variable data, there are only very primitive interpolation methods such as nearest neighbor interpolation considered about class variable data. Here we proposed a spatio-temporal interpolation scheme for pixel-based class variable data under the framework of optimization of likelihood. To optimize the likelihood of spatio-temporal data, a Genetic-Algorithm Hill-Climbing model (GA/HC), which combined genetic-algorithm and Hill-climbing method together to increase the efficiency and quality of optimization, was developed. In GA/HC, a direct coding method and new reproduction and crossover operators are proposed for 3D spatio-temporal data. The evaluation function of GA/HC is defined in the respect to the effect of neighboring spatial-temporal relations on the class change. Through several testing experiments, it showed that GNHC can be a good spatio-temporal interpolation.