Maximum Irrigation Benefit Using Multiobjective Differential Evolution Algorithm (MDEA)
6 Pages Posted: 22 Aug 2010
Date Written: August 21, 2010
This study presents the application of strategies of multiobjective differential evolution algorithm (MDEA) to the maximization of irrigation benefit in the lower orange catchment of South Africa. The two strategies presented are MDEA1 and MDEA3 with binomial and exponential crossover methods respectively. The study compares the non-dominated solutions generated by the two algorithms to find the better algorithm for the irrigation model presented. From the analysis of the results, the results generated by MDEA1 with binomial crossover method are found superior to the results generated by MDEA3 with exponential crossover method. The average total irrigation water of 104 Mm3 was generated with the corresponding averages of 32,208ha of planting areas and ZAR 1257 million total benefit using MDEA1 while the averages of total irrigation water, total area and total benefit of 128.1 Mm3, 28,021 ha and ZAR 808 million respectively were generated by MDEA3. This study concludes that MDEA with binomial crossover method is better in terms of quantity and quality of non-dominated solutions generated. It is further shown that the maximum irrigation water of 3503 m3 per hectare of land cultivated and ZAR 11.25 per m3 of irrigation water used were generated using MDEA1 while MDEA3 generated the maximum irrigation water of 4570 m3 per hectare of land cultivated and ZAR 5.92 per m3 of irrigation water use. This shows that MDEA1 is better in achieving higher profit for farmers using lower volume of irrigation water.
Keywords: Irrigation Planning, MDEA, Multi-Objective, Optimization
JEL Classification: C9
Suggested Citation: Suggested Citation