Assessment of Emission Reduction Potential of Power Supply Enterprises Based on Fuzzy Comprehensive Evaluation Method

. In order to realize the comprehensive analysis and evaluation of the carbon emission reduction potential of power supply enterprises, four emission reduction measures were selected for evaluation, namely, building energy efficiency improvement and clean energy utilization, sulfur hexafluoride emission control, reducing the comprehensive network loss rate, and clean transportation energy utilization. In view of the correlation of various impact factors, the analytic hierarchy process was adopted to form a three-level evaluation index system, According to the index system and based on the interval number fuzzy comprehensive evaluation method, the evaluation model of carbon emission reduction potential of power supply enterprises is constructed. The results show that the potential level of carbon emission reduction of power supply enterprises is the highest "very good", that is, the capacity of carbon emission reduction is significant.


Introduction
The power industry occupies the first place in China's carbon emissions, accounting for more than forty percent of the country's total 1 . Power supply enterprises are an important link in the power industry to achieve low carbon development on the power generation side and the electricity consumption side.
In terms of carbon emission reduction in the power sector, literature 2 used a grey correlation analysis model to comprehensively evaluate the carbon reduction effect of two indicators, namely the construction mileage and transmission quantity of hybrid power grids, and verified that the hybrid power grid technology has the effect of promoting carbon reduction. In the paper 3 , the supply-side emission reduction in Shenzhen's power sector was evaluated from the perspectives of emission reduction volume and cost, and the focus of carbon emission reduction in Shenzhen's power sector was on power supply restructuring. The paper 4 used the LEAP model to analyse the emission reduction potential of the demand side of electricity, and concluded that the energy saving and emission reduction of the tertiary industry and residential electricity consumption had great emission reduction potential. Literature 5 used scenario analysis to study carbon emission reduction policy scenarios in the power industry, and concluded that the future emission reduction potential of power generation-related measures such as clean energy generation, high-power units and carbon capture technology is huge, and the emission reduction effect of demand-side end power policies is not obvious, and the focus of emission reduction in the power industry in the future is mainly on the power generation side.
Emission reduction in the power sector is divided into three perspectives from the emission reduction object: the power generation side, the demand side and the power supply side 6 . There are many factors that affect the carbon emissions of power supply enterprises, and each factor is interrelated and influences each other, at present, there is relatively little research on emission reduction of power supply enterprises.
In order to overcome the difficulties caused by the correlation of multiple factors on carbon emission analysis, this paper uses the hierarchical analysis method to establish a three-level evaluation index system, taking into account the influence of technological progress, local policies and planning on each evaluation index, and realizes the assessment of the emission reduction potential of power supply enterprises based on the interval number fuzzy comprehensive evaluation model.

Fuzzy integrated evaluation method
The fuzzy comprehensive evaluation method is developed from fuzzy mathematics, in which some quantitative data with fuzzy boundaries or unclear descriptions of natural language are represented as a quantified set, and the affiliation levels of the evaluation objects are synthesized according to the degree of influence of various factors, and finally a comprehensive evaluation is made by the maximum affiliation principle or weighted average. Based on fuzzy mathematical theory 7 , this paper treats the problem of evaluating the emission reduction potential of power supply enterprises as a complex system, and uses the fuzzy comprehensive evaluation method to transform the affiliation relationships of various indicators affecting carbon emission reduction and make a comprehensive evaluation according to the weights of the indicators to determine the level of emission reduction potential of power supply enterprises. The specific process 8 is shown in Figure 1.

Establishment of the evaluation indicator system
Comprehensive evaluation involves many factors and extensive content, which objectively requires the indicator system to be as comprehensive and reasonable as possible; according to the SMART principle, the comprehensive evaluation involves many factors and extensive content, which objectively requires the indicator system to be as comprehensive and reasonable as possible; according to the SMART principle, which means selecting specific, measurable, attainable, relevant and trackable indicators to establish the rating indicator system. The rating indicator system was established by selecting specific, measurable, attainable, relevant and trackable indicators. The set of factors (evaluation indicators) affecting the evaluation object is called the set of factors, and the first level is , , , … , , the second level is , , , … , , where m is the number of evaluation indicators.
When evaluating each factor or object, the set of various evaluation results that may be made is called the set of comments, which is noted as , , , … , , where n is the number of evaluation results. When evaluating the emission reduction potential of power supply enterprises, the set of comments is ={excellent, good, fair, poor, very poor}.

Constructing the judgement matrix
Let R be the domain of real numbers, and the closed interval , ̅ is said to be the interval number, denoted by . The interval number is the lower bound of the interval number, is the upper bound of the interval number, the interval number, , ∈ R, and ≤ . When = ,, degenerates to a real number., , so the interval number is a generalization of the real number.
According to the division of indicators, experts used the 0.1-0.9 nine-scale method to compare indicators two by two, thus establishing the interval number complementary judgment matrix Q. The 0.1-0.9 ninescale method is shown in Table 1 , --complementary judgment matrix； is the i-th row and j-th column element of the complementary judgement matrix expressed as an interval number, indicating the importance of the i-th factor compared to the j-th factor; is the number of points at the left end of the interval number; is the number of points at the right end of the number of intervals;

Determination of weights
The row sum of the complementary judgment matrix is calculated and normalized to obtain the interval number weight vector, as shown in equation (3): is the i-th component of the weight vector w expressed as an interval number.
The formula for is shown in equation (4): where is the number of points at the left end of expressed as an interval number; is the number of points at the right end of expressed as an interval number; Build the likelihood matrix. Using the likelihood formula to compare the weight vectors two by two, build the likelihood matrix , as shown in equation (5): (5) where is the likelihood matrix; is the i-th row and j-th column element of the probability matrix.
The formula for is shown in equation (6): where is the length of ， =| |; The same goes for ； The weight vector is obtained by normalise the row sum of the degradability matrix , as shown in equation (7): where is the i-th component of the weight vector expressed as a real number.
The formula for is shown in equation (8):

Constructing a fuzzy integrated evaluation matrix
Determine the affiliation of each indicator to the set of rubrics to obtain the evaluation matrix, as shown in equation (9): (9) where is the evaluation matrix; is the i-th row and j-th column element of the evaluation matrix.
The evaluation matrix rows sum to 1, satisfying equation (10): First, to do the first-level fuzzy comprehensive evaluation, multiply the weight matrix of the second-level evaluation index with the single-factor fuzzy evaluation matrix, as shown in equation (11): where ° is the take large and take small operator in the synthesis operation.
Thus, the total fuzzy synthesis evaluation matrix B ̅ can be obtained, as shown in equation (12): , , … , (12) Secondly, to do the second level fuzzy comprehensive evaluation, the total fuzzy comprehensive evaluation matrix is synthesized with the weights of the first level factor set as shown in equation (13): Normalise ̅ as shown in equation (14): Finally, based on the principle of maximum membership, the evaluation results of the emission reduction potential of power supply enterprises can be obtained.
3 Application of fuzzy comprehensive evaluation method in the assessment of emission reduction potential of power supply enterprises

Establishing an index system for evaluating the emission reduction potential of power supply enterprises
The carbon emissions of power supply enterprises are distinctive in that the sources of carbon emissions can be divided into the energy side and the non-energy side 9 . In response to the above carbon emission sources, power supply enterprises have formulated four emission reduction measures, namely building energy efficiency improvement and clean energy utilization, sulfur hexafluoride emission control, reducing comprehensive network loss rate, and clean transportation energy use. Therefore, the above four emission reduction measures are selected as primary indicators. Through analysis and field research, the factors that affect the emission reduction effect of the four emission reduction measures are set as the third level indicator.
Combining the carbon emission characteristics 10 of power supply enterprises and the selected evaluation indicators, the evaluation index system of emission reduction potential of power supply enterprises was established as shown in Figure 2 The evaluation object of the comprehensive evaluation is the emission reduction potential of power supply enterprises, and the factor set is divided into two layers. (1) The first factor set is:

Constructing the judgement matrix
According to the evaluation index system, using the 0.1-0.9 nine scale method in Table 1

Determination of weights
After the interval number complementary judgment matrix is obtained, the weight vectors expressed in the form of interval numbers are calculated according to equations (2). The weight vectors , , and for the emission reduction potential of the four emission reduction measures of the power supply enterprises are shown below: