Comparison of traditional method and genetic algorithm optimization in the land reallocation stage of land consolidation

The land reallocation phase as a process has a direct influence on the success of land consolidation (LC) studies. In Turkey, land reallocation is performed according to ‘farmer preferences’ as determined by interviews with landowners. This traditional method is called the ‘interview-based method’. In land reallocation studies, parcel locations and property rights are re-arranged according to both farmer preferences and project productivity. A variety of solutions is inevitable in this re-arrangement, as the person responsible for the project and the series of criteria involved in the process differ between projects. How acceptable the solution is for landowners, and the expected benefit from an LC project, depends on the determination of an optimal solution. Genetic algorithms (GA) are widely considered to be an effective method for finding the optimum solution. In this study, land reallocation was performed automatically using GA and the results compared with the traditional land reallocation method. The main purpose of this study is to advance a land reallocation model which minimizes human interference while still meeting traditional land reallocation criteria. Both farmer preferences and block occupancy status were taken into consideration in this model by using GA. The polygonal areas (blocks) within which plots were defined were planned according to topography and the road and irrigation systems. The study was tested with real data, and it was observed that the success rate of block placement using GA was 93%, according to the farmer's preferences, and that full occupancy rate in the blocks was 73%. The results derived from both models were compared in terms of number of parcels, average parcel size, number of parcels per landholding, and number of shared parcels; these criteria are commonly accepted as prominent comparison criteria in the relevant literature. The comparison shows that more successful land reallocation results were derived from the GA model than from the interview-based model. A survey of landowners was also conducted to verify the success of the GA method. The results showed that an interactive land reallocation which includes an optimization process through which the landowners are able to change their preferences can be carried out effectively.