Application of adaptive neuro-fuzzy inference system to model CO2 emissions of chickpea production in Kangavar county, Iran

The main objectives of this study were determination of carbon dioxide (CO2) emissions and finding relation between CO2 emissions and harvested crop usingadaptive neuro-fuzzy inference system (ANFIS) in the production process of chickpea under dry farming system in the Kangavar county of Kermanshah province that is located in the west of Iran. 25 farmers were considered to collect data. Results indicated that total CO2 emissions in chickpea farms was about225 kg CO2 eq. ha-1. Diesel fuel was determined as the main CO2 emitter input with 173.49 kg CO2 eq. that was more than 77% of total emissions. Modeling results revealed that ANFIS architecture with two-level was computed as the best model with cceptable statistical indices. According to ANFIS (3) results, R2 was calculated as 0.954 and RMSE and MAE were determined as 2.82 kg CO2 eq. and 0.51, respectively. Finally, it can be concluded that artificial intelligence methods especially ANFIS can be useful for modeling with high accuracy in the agricultural activity.

Application of adaptive neuro-fuzzy inference system to model CO2 emissions of chickpea production in Kangavar county, Iran Keywords:

Adaptive neuro-fuzzy inference system , Agriculture , Chickpea , CO2 , Prediction