EXTENDED ABSTRACT
INTRODUCTION
Today, in the economies of all countries, both developed and developing, increasing productivity and optimal energy consumption has become a national priority Productivity is a standard for evaluating the performance of systems and determining the degree of success or failure in achieving the goals of the system with regard to resource consumption. Considering the growth in energy consumption, the optimization of energy consumption in order to protect the environment, provide security of supply protection of national resources and wealth, and address intergenerational issues has become one of the important priorities of the country. Increasing productivity at the national level will increase people's living standards, reduce inflation, and create national competitiveness in global markets. Increasing national productivity is the result of increasing productivity in organizations, institutions, and economic enterprises. Therefore, considering to the importance of this issue in the national economy, this research the purpose of this research isaimed to investigate the factors affecting energy consumption and efficiency. For this purpose, the fuzzy regression model, which has great explanatory power in this field, was used to investigate this issue. Econometric models need complete and definite information to be accurate considering that energy consumption and efficiency are constantly fluctuating, being able to model this uncertainty is a necessity.
 
 
METHODOLOGY
In this article, to investigate the factors affecting energy consumption and efficiency in Iran, the autoregression model of soft fuzzy logistic transfer was used due to its flexibility in modeling for the time period of 1990-2019, and MATLAB software was used to estimate the membership functions.
 
FINDINGS
The results of the linear and nonlinear fuzzy models indicated that the energy consumption estimation of the fuzzy nonlinear model has significantly increased the explanatory power of the model the coefficient of determination in the fuzzy nonlinear model is 0.64, while it is 0.41 in the linear model. Because non-linear models model the behavior of economic variables more accurately than linear models, the coefficient of determination of the non-linear model is higher compared to the linear model. Based on the estimation results of the nonlinear model, the transmission speed for energy consumption is 2.5, which indicates a high adjustment speed in energy consumption. In addition, the threshold value for energy consumption is 8.1. Also, the linear model of energy consumption shows that technology and gas and oil prices reduce energy consumption. On the other hand, the added value of the industrial sector and economic growth increase energy consumption. The results of the fuzzy nonlinear model are completely different from the linear model, because the economic variables have a nonlinear behavior, so it is necessary to model these nonlinear behaviors using nonlinear models to make the proper analysis of the effect of different variables possible.
 
CONCLUSION
According to the study results, three numbers can be obtained for how the investigated variables affect the dynamics of energy consumption and efficiency. For this purpose, we examined the three cases of high threshold, low threshold, and middle threshold (when the emission of carbon dioxide gas and the price of gas are equal to the threshold value).
Regarding how the rate of carbon dioxide gas emission affects energy efficiency, as we approach the low threshold, the effect of carbon dioxide gas emission increases significantly however, this effect decreases at the high threshold. One of the reasons for this is the increase of energy consumption at a high threshold, which necessitates the optimal use of energy and higher use of technology in order to reduce environmental damage. The impact of fixed investments on energy efficiency is positive and it stabilizes as it approaches the low and middle thresholds, but this effect increases at the high threshold. Fixed investment can have a significant impact on increasing energy efficiency in the country. Sectors such as industry, household public and private buildings, transportation, construction, and agriculture are among the most important sectors in which energy efficiency can be increased by increasing large investments. Also, added value and technology have a greater impact on energy efficiency at the high threshold level and the impact decreases at the middle and low threshold levels. Added value and technology improve the partial and total efficacy of production factors, which plays a significant role in increasing productivity.
The price of gas and oil has caused an increase in energy efficiency, which has the greatest effect at the upper threshold, and this effect decreases as weapproach the lower threshold. This points to the effectiveness of increase in energy prices on the productivity index energy-capital substitution elasticity and labor supply. With a coordinated increase in the price of energy carriers, if it is not possible to substitute energy with other production factorsproduction will decrease, and the increase in prices will not have a significant effect on productivity and consumption indices. If the substitution between the factors is possible, production will not change with reduction in consumption and the productivity index will improve.
The results of the fuzzy nonlinear model of the estimation of the effect of independent variables on energy consumption at each of the three mentioned thresholds indicate that the variables have almost the same effect, which is related to the characteristics of energy consumption. Regarding the percentage change of carbon dioxide emission rate on energy consumption, there is no significant reduction at the low and middle threshold however, this effect decreases at the high threshold. The impact of fixed investments on energy consumption is positive and is almost stable at the low and middle thresholds, but this influence increases at the high threshold. Also, the impact of technology on energy consumption is almost the same at all three thresholds  although it decreases at the high threshold. The price of gas and oil reduces energy consumption. Also, Netich shows that it has the greatest effect at the upper threshold and this effect increases as it approaches the lower threshold. Regarding the effect of economic growth on energy consumption, the results indicate a direct relationship between these two variables, with the greatest effect on energy consumption observed at the high threshold.