A clustering model based on an evolutionary algorithm for better energy use in crop production

Energy consumption and its negative environmental impacts are of interesting topics in the recent centuries. Agricultural systems are both energy users and suppliers in the form of bio energy and play a key role in world economics as well as food security. A high amount of energy from different sources is used in this sector while researchers who investigated energy flow in crops production especially in developing countries, have reported a high degree of inefficiency. In order to differentiate between efficient and inefficient farms, a clustering model based on imperialist competitive algorithm (ICA) has been developed and the surveyed watermelon farms have been clustered based on three features, i.e. greenhouse gas (GHG) emission, input energy and farm size. The results showed that of the three developed clusters, the best cluster performed 20 and 46 % better than the two other clusters in energy and 22 and 52 % in CO2 emissions. The average of total energy input and GHG emissions for the best cluster were calculated as 43,423 MJ per ha and 8,120 CO2eq. The results of this study demonstrate the successful application of ICA for better use of energy in cropping systems which can lead to a better environmental and energy performance