Evaluating the Efficiency of Energy Consumption in Dairy Cattle Breeding Unit Using Statistical and ANFIS Methods

Abstract Evaluating the input and output process of energy in agricultural systems is one of the ways to determine the level of sustainability in these systems. For this reason, the purpose of this study was to investigate and predict the amount of input and output energies and also related indicators in cattle breeding. The information needed to conduct this study was collected by referring to Pars Company and extracting the recorded information regarding the amount of consumed inputs and the value of produced products. Also, part of the information was extracted from the financial reports presented by the company and the reports published in the annual meetings of the shareholders. In addition, the information is related to the financial period from 2011 to 2020. Based on the available data, the value of input and output energies, energy indices were calculated. Also, adaptive network-based fuzzy inference system (ANFIS) was used to predict performance and output energy. According to the results of data analysis, the average energy required for raising cattle during a one-year period was 52,604.52 MJ∙cow −1 . Also, produced milk, produced manure, and born calf were considered as the output energies. The energy equivalent of produced milk during a year was 56,821.39 MJ∙cow −1 . After examining different models of ANFIS, the best ANFIS model with the input membership function of trimf was selected and the R 2 and MAPE values of 0.98 and 0.42 were obtained, respectively.