Research on Decision-Making for a Photovoltaic Power Generation Business Model under Integrated Energy Services

Abstract

The 14th Five-Year Plan for renewable energy development proposes that renewable energy should achieve high-quality leapfrog development during the 14th Five-Year Plan period. The rapid development of integrated energy services has created more market opportunities for photovoltaics, and the photovoltaic (PV) industry has entered a new stage. In this study, from the perspective of stakeholders of a PV micro-grid under integrated energy services, a decision model for a PV micro-grid business model was established with the help of fuzzy mathematical theory. The MATLAB platform was used to carry out the decision analysis and simulation of the PV micro-grid business model. The simulation results showed that the self-consumption business model had the best overall performance among the current three different business models. It was also the business model with the highest comprehensive correlation with the government, power grids, investors, and users. Studying the best business model for PV micro-grids under integrated energy services has practical significance as it provides guidance and promotes faster and better development of PV micro-grids under integrated energy services.

1. Introduction

The global energy structure is undergoing a period of change, with the cost of various clean energy technologies, for example, wind power, declining gradually while the corresponding demand rises. In recent years, reducing carbon emissions and promoting sustainable development on a regional level have become the global consensus. In order to appropriately respond to climate issues, each country has enacted a series of policies and systems that reflect their national conditions [1]. Integrated energy services are one of the optimal ways to utilize better clean energy technologies and distributed energy technologies are used to improve the clean energy supply, energy allocation efficiency and energy efficiency in China [2]. This approach has attracted more attention from investors and entrepreneurs, including technology inventors. Integrated energy services are of great significance in promoting the restructuring of China’s supply, improving the energy system’s efficiency, and upgrading energy consumption. These services are a crucial means of achieving energy transformation in the future [3].

The energy sector in China will be transformed and upgraded with the help of integrated energy services. Integrated energy service providers have a key role in the process of advancing the PV industry around the world and they undertake an important mission. With regard to the scale of the PV industry, China’s new installed capacity for PV power generation, the cumulative installed capacity, polysilicon production and module production have been ranked first in the world for many years [4]. With integrated energy services, the PV industry will surely be able to develop even further, especially in the area of centralized photovoltaics. The PV industry can build a large diverse, multi-energy complementary base, including water and wind storage, which will expand the market for the PV industry [5].

The current industry is overgrown with the full use of integrated energy services in the PV industry. If PV power is to achieve its sustainability goals, it must first solve the problem of limiting the amount of electricity it generates for the purpose of safely managing the grid. At the same time, China must also take appropriate measures to improve and adjust its photovoltaic energy-generating structure to quickly achieve its energy market optimization and upgrading goals. Currently, the PV industry and integrated energy services are in a win–win situation as integrated energy services are driving the development of the PV industry [6]. Thus, the PV industry gains a larger market and a wider range of applications, while integrated energy services are fully utilized. Chinese PV companies prefer to develop the PV industry chain. Analysis of the experience of the PV industry in other countries shows that applying the micro-grid energy storage model to PV power generation brings economic and social benefits [7,8]. In addition, the model also promotes the achievement of carbon peaking and carbon neutrality goals [9].

At present, there are three incentive modes for household-distributed PV that are utilized across the world: the “benchmarking feed-in tariff” (FIT) policy, “NetMetering” policy, and “self-consumption” policy. In this study, PV power generation under integrated energy services was taken as the research object. The PV industry uses the same three modes as its business models, including firstly, the self-consumption mode of “self-use first and the rest of the generation to the grid”, which mainly revolves around customers and PV companies; second, the feed-in tariff model based on government subsidy policy; and third, a net metering business mode that enjoys direct access to the grid’s retail tariff. Based on the analysis, different business models have different subjects of interest. An effective value chain consists of these subjects of interest. Choosing a scientifically-based business model leads to greater revenue and value chain benefits [10].

The construction of a photovoltaic power generation project is a very complex process, and these processes include the selection, operation, installation and design of the facility. Because we wanted to focus on the uncertainty of the whole project construction process, we chose a fuzzy mathematics method for our research. The fuzzy comprehensive evaluation method is based on fuzzy mathematics, which transforms fuzzy mathematical relations into mathematical principles. For uncertain mathematical relations, such as uncertain boundary conditions and the influence of variable, uncertain factors, this method is able to comprehensively evaluate the influence of uncertain factors, and these comprehensive factors will be considered. The basic idea is to decompose the complex weight coefficient, and finally comprehensively evaluate each factor and sort them into several main levels. In this study, the decision model was constructed and simulated from the stakeholder perspective with the help of fuzzy mathematical theory. The purpose was to provide theoretical support for the selection of a PV power generation business model in China. This research provides guidance and has practical significance for further enhancing the industrialization and scale development of PV power generation under integrated energy services in China. It provides a reference for promoting the faster and better development of the PV industry under integrated energy services while helping to achieve the goal of carbon peaking by 2030 and carbon neutrality by 2060.

2. Literature Review

There are various results from the research on business models used in PV power generation in the setting of integrated energy services.

In a study on business models in the PV industry in the U.S., Yasser and Sami [11] point out that the problems in the commercialization of green energy need to be solved before the PV industry can develop a good green energy business model. By identifying business models such as those used for products and technologies, they constructed a corresponding value network diagram. They also analyzed the profitability of each industry model to solve the problems regarding the commercialization of green energy. Zhai [12] suggests that the PV industry can play a positive role by choosing a business model based on taxation. This is especially true for new energy companies that use PV power as a product because they can use the corresponding tax investment fund to build the PV micro-grid. The PV industry can obtain more revenue from the sale of PV electricity, while controlling the tax dividends, as is currently the case in the United States. Chi et al. [13] point out that establishing a business model can promote the development of distributed solar PV. They believe that business model innovation is the best way to stimulate the Chinese PV industry. The business model of distributed solar PV in China is still in the exploration stage. For a more market-oriented business model, an in-depth analysis is necessary. Using a comparative analysis of the U.S. and Chinese distributed markets, Bao [14] suggests ways for China to learn from the successful business models of solar PV towns.

In a study on the stakeholders of PV power generation, according to Sauter and Watson [15], as companies change their roles in their business models, the relationships between them and their users will unavoidably be affected. Self-generation is used by some conventional suppliers. In a company-controlled model, users are not the primary actors, but rather appear to be passive. Users can take the lead and decide how the community micro-network is built if the model is the community model. According to Drury and Miller [16], the business model chosen for the Southern California region of the United States is a third-party PV generation model. In this model, PV companies operate and own the PV systems, and generate profits by selling PV power or leasing PV equipment. PV companies providing services on a third-party basis can reduce upfront costs as well as control technical risks by utilizing monitoring systems. The study also showed that younger groups with lower education and income levels prefer this business model. Zhang [17] believes that distributed energy projects are developing rapidly due to the State’s support. Even so, the business model is not yet mature, and as subsidies continue to shrink, its development will become more fragile. Currently, China has not formed a large-scale distributed energy industry. A mature business model requires innovative thinking on many fronts at the present time [18]. Integrated energy services will become a focal point in the transformation and upgrading of China’s energy enterprises in the future [19].

In terms of business model design, Horváth and Szabó [20] used BMC to analyze three different business models for PV globally. They proved how and to what extent these business models can overcome the barriers to distributed energy deployment. Timilsina et al. [21] found that PV technology has pioneering characteristics through the study of business models for PV. By leveraging this characteristic, one can construct a business model that uses technology as its driving force, and then innovate that business model. Gabriel and Kirkwood [22] used BMC to examine the factors that might influence the choice of business models for renewable energy companies. In addition, they explored regional differences and their impact on the various business models adopted. Strupeit and Palm [23] used the concept of business models as an analytical tool to investigate how different business models facilitate the deployment of customer on-site PV systems in Germany, Japan, and the United States. Zhang [24] took the payback period in the self-consumption model and uses the NPV model to conduct his research. For the PV market to operate properly, it is necessary to ensure that the feed-in tariff payback period is reasonable.

In summary, the number of scholars examining the business model of PV generation has increased, and a large amount of research has taken place. However, most of the studies are at the stage of qualitative analyses. There is little empirical evidence on PV power generation business models, which provided the impetus for our study. This paper explores the optimal business model for PV power generation under integrated energy services from the perspective of PV power generation stakeholders, and also establishes a PV power generation business model using a decision model with the help of fuzzy mathematical theory.

3. Theoretical Analysis

3.1. Fuzzy Property Analysis

The development and operation of the photovoltaic industry is a game based on stakeholders, which reflects a transaction-based balance between the stakeholders as a whole. According to Mitchell, stakeholders must have three attributes: influence, legitimacy and urgency; thus, it was concluded that the most important stakeholders in the development of the photovoltaic industry include the government, investors, power grid companies and users. The government represents the interests of the country and is actually the biggest beneficiary; photovoltaic investors provide sufficient capital and corresponding technology and equipment support for photovoltaic projects, which is the main financial backing for the development of photovoltaic industry; the grid company is directly involved in the photovoltaic application, guiding the general direction of the new energy and low-carbon economy, and is the main backup for the development of photovoltaic industry; users are the most direct beneficiaries, and at the same time, this may have a negative effect on the development of the photovoltaic industry. The relationship between various stakeholders is very complex.

In order to maximize their own interests, stakeholders interact and transmit information and resources with the help of business models, which also means that the business models show fuzzy characteristics; these are mainly reflected in the following aspects.

First, the complexity of its business operation. From the above, it can be seen that as a large-scale system, there are multiple stakeholders in PV power generation. These stakeholders have complex interconnections, including internal influences due to cost and revenue factors. This also includes the influence of the external environment such as policy, technology, the market, etc. The formation of linkages and related systems will grow and expand, making the complexity even more so, all of which will have an impact on the economics of commercial operations [25].

Second, the non-linear relationships of the business operation elements are crisscrossed. When key elements change, the remaining elements are likely to change as well. For PV generation, the state subsidizes the difference above the FGD benchmark tariff and needs to pay more money. After changing its feed-in tariff, the tax and generation link may also change. Regardless of the type of customer, they have to sign power purchase agreements with grid companies, which increases the transaction costs. Moreover, many small and medium-sized customers cannot develop invoices for grid companies, and the operation needs to solve the problems of the industry, commerce and taxation. This clearly shows its non-linear characteristics.

Third, the internal and external randomness of its business operation. Randomness refers to the inability to predict whether an event will take place in the future. In the process of photovoltaic power generation, various conflicts will inevitably arise among the stakeholders in order to make profits, which will have an impact on the orderliness of commercial operations.

Fourth, the initial value sensitivity of its commercial operation. If observed from a theoretical perspective, chaotic systems are usually very sensitive to initial values. During the commercial operation of PV power generation, its decision makers and management need to adjust several elements in the market environment. At this point, the interested parties may either gain further benefits or be caught in the ambiguity.

3.2. Fuzzy Decision Method

From the discussion above, the business model of PV power generation has obvious ambiguity, especially when there are many uncertainties in the market, it is difficult to make decisions regarding the selection of a scientific business model. At this point, it is particularly important to clarify the needs of various stakeholders of PV power generation in a complex environment, and then make a scientific and reasonable choice. In order to do this, this paper constructed a fuzzy decision-making method based on fuzzy mathematical theory and grey system theory in fuzzy decision-making. This method involves many stakeholders, and it was found that with the help of this method, the optimal solution could be obtained when choosing a business model, and the efficiency of the decision could be significantly improved.

When constructing a tool for decision-making with the help of fuzzy mathematical theory, it is necessary to clarify which elemental indicators exist for its stakeholders, and then assign them according to the actual situation, and complete the fuzzy decision-making process to select the optimal business model with the help of two aspects of the elements, which are firstly, the weight values, and secondly, the comprehensive evaluation [26]. Therefore, the fuzzy decision-making process of PV power generation was organized and summarized to obtain this decision-making method, the details of which are shown in Figure 1.

In the first step, the construction of fuzzy decision indicators for PV power generation is completed. After analysis, when choosing the business model of PV power generation, because the interests of its stakeholders are involved and in order to complete the fuzzy decision-making work, the construction of relevant indicators must be completed first.

The second step is to clarify the weight of the PV power generation fuzzy decision indicators and complete the construction of the fuzzy decision matrix. Entropy technology is used to clarify the index weights of PV power generation, adjust and correct them, and at the same time, calculate their weights and complete the construction of the corresponding matrix.

In the third step, according to the principle of maximizing the degree of affiliation, the interests and demands of the subjects of interest of PV power generation are clarified and the business model of PV power generation is comprehensively evaluated. In this way, the advantages and disadvantages of various business models of photovoltaic power generation can be clarified, and there will be a clear basis for selection.

4. The Design of the Study

4.1. Indicators Construction

Analyses of the fuzzy decision model for PV power generation demonstrate the importance of fuzzy decision indicators for PV power generation. The indicators in this system describe the performance of a certain aspect of the fuzzy object, which can be evaluated and measured.

In the commercial operation of PV power generation, four major factors, namely, economic profit, investment cost, power supply level and social benefit, affect the interests of the stakeholders of photovoltaic power generation. They have different needs, and it is necessary to build the corresponding decision indicators based on their needs. Therefore, the indicators established involve four aspects: firstly, economic profit; secondly, investment cost; thirdly power supply level; and finally, social benefit. They belong to the first-level indicators, which can be expressed as Z1, Z2, Z3, Z4, respectively, and the set is Z = {Z1, Z2, Z3, Z4}. Under the primary indicators, secondary indicators are set, as shown in

Table 1. Fuzzy decision indicator system for photovoltaic power generation business model.

Fuzzy Decision Primary Indicators	Fuzzy Decision Secondary Indicators
Economic Profit	Grid operating income U11 Investment operating income U12 Government Taxes U13 Operating margin U14 Subsidy income U15 Additional profit U16 Operating profit U17
Investment Costs	Equipment Investment U21 Human resource investment U22 Financial Investments U23 Loss Cost U24
Power Supply Level	Electricity price level U31 Power supply quality U32 Power supply range U33 Power supply period U34 Power supply stability U35
Social Benefits	Safety benefits U41 Environmental benefits U42 Employment Benefits U43 Industry Benefits U44

.

4.2. Determination of Indicator Weights and Matrices

After specifying the primary and secondary indicators of the fuzzy decision model for PV power generation, it is necessary to specify the weights of each indicator, i.e., the importance ranking. This is an essential step for quantitative decision-making. After fully analyzing the needs of the PV power generation stakeholders and the nature of the four major indicators, the entropy technique and hierarchical analysis are used to complete the modelling operation and clarify the weights of the four first-level indicators. Then, each level corresponds to the core objectives of the final analysis. On this basis, the specific weights of each indicator are clarified [27]. The indicators are then compared in pairs and their importance is assessed following the completion of the above steps. The importance of the indicators in fuzzy decision making is analyzed using the Saaty scale. Based on the trust coefficient, the influence value of each indicator is calculated, and a decision judgment matrix is created. A description of the Saaty scale is shown in

Table 2. Saaty’s scale chart.

Rating Scale	Explanation
1	Two elements contribute equally to the objective
3	Experience and judgment slightly favour one element over another
5	Experience and judgment strongly favour one element over another
7	An element is favoured very strongly over another, its dominance is demonstrated in practice
9	The evidence favouring one element over another is one of the highest possible order or affirmation
2, 4, 6, 8	Judgment of having adjacent intermediate values

.

The values regarding fuzzy decisions are arranged by a matrix, where it is assumed that a certain value is denoted by $a_{ij}$ and its corresponding matrix takes the form of:

$$A={\left(a_{ij}\right)}_{m\times n}\left[\begin{array}{cccc}\raisebox{1ex}{$a_1$}\!\left/ \!\raisebox{-1ex}{$a_1$}\right.& \raisebox{1ex}{$a_1$}\!\left/ \!\raisebox{-1ex}{$a_2$}\right.& \dots \dots & \raisebox{1ex}{$a_1$}\!\left/ \!\raisebox{-1ex}{$a_n$}\right.\\ \raisebox{1ex}{$a_2$}\!\left/ \!\raisebox{-1ex}{$a_1$}\right.& \raisebox{1ex}{$a_2$}\!\left/ \!\raisebox{-1ex}{$a_2$}\right.& \dots \dots & \raisebox{1ex}{$a_2$}\!\left/ \!\raisebox{-1ex}{$a_2$}\right.\\ \dots \dots & \dots \dots & \dots \dots & \dots \dots \\ \raisebox{1ex}{$a_n$}\!\left/ \!\raisebox{-1ex}{$a_1$}\right.& \raisebox{1ex}{$a_n$}\!\left/ \!\raisebox{-1ex}{$a_2$}\right.& \dots \dots & \raisebox{1ex}{$a_n$}\!\left/ \!\raisebox{-1ex}{$a_n$}\right.\end{array}\right]$$

(1)

In this formula, given positive numbers, if a_ij = 1/a_ij, a_ij = 1, then the judgment matrix can be obtained. After constructing the judgment matrix, the eigenvalues of the decision index weights are calculated to obtain the eigenvectors, and the indicators are represented by n here. The corresponding relative weights can be represented by $w$. The determination process is as follows: in the first step, the geometric mean of the elements of the judgment matrix is calculated, and the following formula is used in the calculation process.

$$\overline{w_i}=\sqrt[n]{{\displaystyle \prod }_{j=1}^n{a}_{ij}\left(j=1,2,\dots \dots n\right)}$$

(2)

Normalization calculations are performed:

$$w_i=\frac{\overline{w_i}}{{{\displaystyle \sum }}_{i=1}^n\overline{w_i}}\left(i=1,2,\dots \dots .n\right)$$

(3)

The corresponding judgment matrix is usually obtained by comparing the factors, but the judgment matrix is likely to show inconsistencies, especially in terms of influence. When there is a small gap between two types of factors, the chances of inconsistency are greater. At this time, the consistency test is carried out with the judgment matrix as the test object. After obtaining the consistency indicator, the stochastic consistency ratio can be obtained by combining the scalar values.

Table 3. Corrected indicators.

Corrected Indicators
Dimensions	1	2	3	4	5	6	7	8	9
Scalar values	0	0	0.58	0.9	1.12	1.24	1.32	1.41	1.45

shows the specifics of the corrected indicators.

In the above table, CR represents the consistency ratio, and the larger the value, the worse the consistency of the judgment matrix. The smaller the value, the better the consistency and when the value of CR is 0, it means the judgment matrix is completely consistent.

4.3. Modelling

By using fuzzy mathematical theory and grey system theory, we found that the indicators are not only rich in levels but also very diverse in types. Therefore, after clarifying the indicators and their respective weights, it is no longer necessary to make fuzzy decisions based on single-factor methods.

Firstly, after clarifying the advantages and disadvantages of the business model, the construction of the fuzzy decision-making comprehensive evaluation model is completed, which is represented by V. Using this set, the analysis and description of each decision indicator is started; it should be noted that the number of factors corresponding to each decision indicator is not necessarily the same. Next, the core factors of each business model are processed using fuzzy decision making and the corresponding matrix model is obtained. The specific matrix model is as follows: assume that there exists a fuzzy factor i in V, which is in the Zth layer, and when a decision is made on it, its affiliation factor with element j is represented by r. At this time,

$$R=\left[\begin{array}{cccc}{r}_{11}& r_{12}& \dots & r_{1m}\\ r_{21}& r_{22}& \dots & r_{2m}\\ r_{31}& r_{32}& \dots & r_{3m}\\ \dots & \dots & \dots & \dots \\ r_{m1}& r_{m2}& \dots & r_{km}\end{array}\right]$$

(4)

The integrated fuzzy decision matrix model for photovoltaic power generation is as follows:

$$B=A\ast R=\left(W_1,W_2,W_3,\dots \dots ,W_n\right)\ast \left[\begin{array}{cccc}{r}_{11}& r_{12}& \dots & r_{1m}\\ r_{21}& r_{22}& \dots & r_{2m}\\ r_{31}& r_{32}& \dots & r_{3m}\\ \dots & \dots & \dots & \dots \\ r_{m1}& r_{m2}& \dots & r_{km}\end{array}\right]$$

(5)

5. Results and Analysis

In photovoltaic power generation, there are three business models, and analyzing each one takes additional time and effort. Here, we chose the feed-in tariff business model as the simulation object and used MALTAB to conduct the simulation analysis.

5.1. Calculation of Fuzzy Decision Index Weights

Firstly, the fuzzy decision index weights were calculated, and the number of first-level indicators was four, which were represented by Z1, Z2, Z3, Z4, and their weights were represented by $w$. At this time, $M_i={{\displaystyle \prod }}_{j=1}^n{c}_{ij}$, which was further transformed into $w_i=\sqrt[n]{M_i}$, and at this time, substituting the importance values, we were able to complete the calculation of index weights. The index weights of economic profit, investment costs, power supply level and social benefits are $w_1, w_2, w_3, w_4$ respectively.

$$w_1=\sqrt[4]{1\times 1\times \left(1/2\right)\times \left(1/2\right)}=0.71$$

(6)

$$w_2=\sqrt[4]{1\times 1\times \left(1/3\right)\times \left(1/2\right)}=0.064$$

(7)

$$w_3=\sqrt[4]{3\times 2\times 1\times 1}=1.57$$

(8)

$$w_4=\sqrt[4]{2\times 2\times 1\times 1}=1.407$$

(9)

Then, by processing w = (w₁, w₂, w₃, w₄)T, using normalization means, it was possible to obtain ${\displaystyle \sum }_{i=1}^4{w}_i=0.71+0.64+1.57+1.41=4.33$, which belongs to the indicator feature vector, and then with the help of $W_i=\frac{w_i}{{{\displaystyle \sum }}_{i=1}^n{w}_i}$, the true weights of the different indicators were derived as follows,

$$W_1\frac{w_1}{{{\displaystyle \sum }}_{i=1}^4{w}_1}=\frac{0.71}{4.33}=0.164,W_2\frac{w_2}{{{\displaystyle \sum }}_{i=1}^4{w}_2}=\frac{0.064}{4.33}=0.148,W_3\frac{w_3}{{{\displaystyle \sum }}_{i=1}^4{w}_3}=\frac{1.57}{4.33}=0.363,W_4\frac{w_4}{{{\displaystyle \sum }}_{i=1}^4{w}_4}=\frac{1.407}{4.33}=0.325$$

(10)

According to the similar method mentioned above, the calculation of the weights of the secondary fuzzy decision indicators corresponding to each primary fuzzy decision indicator can be realized by simulation with the help of MATLAB software.

5.2. Fuzzy Decision Simulation and Analysis

Once the decision indicators and their weights at all levels of photovoltaic power generation were clarified, the comprehensive judgment matrix model was used to calculate the fuzzy judgment matrix and comprehensive judgment for the primary and secondary indicators.

The judgment matrix for the first-level indicators was calculated. The judgment matrix for economic profit, investment costs, power supply level and social benefits were $R_1, R_2, R_3, R_4$ respectively. Using MATLAB, the following matrix was obtained,

$$R_1=\left[\begin{array}{ccccc}0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 1& 0& 0\end{array}\right],$$

(11)

$$R_2=\left[\begin{array}{ccccc}0& 0.2& 0.7& 0.1& 0\\ 0& 0.7& 0.3& 0& 0\\ 0.1& 0.2& 0.5& 0.2& 0\\ 0& 0.1& 0.8& 0.1& 0\\ 0.1& 0.3& 0.4& 0.2& 0\\ 0& 0.1& 0.5& 0.3& 0.1\\ 0.1& 0.3& 0.4& 0.2& 0\end{array}\right]$$

(12)

$$R_3=\left[\begin{array}{ccccc}0.1& 0.3& 0.3& 0.2& 0.1\\ 0.2& 0.2& 0.3& 0.2& 0.1\\ 0.2& 0.3& 0.4& 0.1& 0\\ 0.1& 0.1& 0.3& 0.3& 0.2\end{array}\right]$$

(13)

and

$$R_4=\left[\begin{array}{ccccc}0.1& 0.1& 0.4& 0.3& 0.1\\ 0& 0.2& 0.5& 0.3& 0\\ 0.2& 0.4& 0.4& 0& 0\\ 0.1& 0.2& 0.4& 0.3& 0\end{array}\right]$$

(14)

After calculating the secondary index judgment matrix using MATLAB, the following matrix was obtained.

Secondary indicator economic profit B₁:

$$B_1=W_1\times R_1=\left(0.298,0.158,0.298,0.158,0.088\right)\times \left[\begin{array}{ccccc}0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 1& 0& 0\end{array}\right]=\left(0,0.596,0.404,0,0\right)$$

(15)

Secondary indicator investment cost B₂:

$$\begin{gathered} B_2=W_2\times R_2=\left(0.166,0.166,0.137,0.194,0.1,0.166,0.071\right)\times \left[\begin{array}{ccccc}0& 0.2& 0.7& 0.1& 0\\ 0& 0.7& 0.3& 0& 0\\ 0.1& 0.2& 0.5& 0.2& 0\\ 0& 0.1& 0.8& 0.1& 0\\ 0.1& 0.3& 0.4& 0.2& 0\\ 0& 0.1& 0.5& 0.3& 0.1\\ 0.1& 0.3& 0.4& 0.2& 0\end{array}\right] \\ =\left(0.0308,0.2641,0.5411,0.1474,0.0166\right) \end{gathered}$$

(16)

Secondary indicator power supply level B₃:

$$\begin{gathered} B_3=W_3\times R_3=\left(0.351,0.351,0.189,0.109\right)\times \left[\begin{array}{ccccc}0.1& 0.3& 0.3& 0.2& 0.1\\ 0.2& 0.2& 0.3& 0.2& 0.1\\ 0.2& 0.3& 0.4& 0.1& 0\\ 0.1& 0.1& 0.3& 0.3& 0.2\end{array}\right] \\ =\left(0.154,0.2431,0.3189,0.192,0.092\right) \end{gathered}$$

(17)

Secondary indicators of social benefits B₄:

$$\begin{gathered} B_4=W_4\times R_4=\left(0.326,0.124,0.193,0.357\right)\times \left[\begin{array}{ccccc}0.1& 0.1& 0.4& 0.3& 0.1\\ 0& 0.2& 0.5& 0.3& 0\\ 0.2& 0.4& 0.4& 0& 0\\ 0.1& 0.2& 0.4& 0.3& 0\end{array}\right] \\ =\left(0.1069,0.206,0.4124,0.2421,0.0326\right) \end{gathered}$$

(18)

When the calculation of the judgment matrix of each indicator was completed, the calculation of the comprehensive evaluation was launched, and the following matrix was obtained through the calculation:

$$R=\left[\begin{array}{ccccc}0& 0.596& 0.404& 0& 0\\ 0.0308& 0.2641& 0.5411& 0.1474& 0.0166\\ 0.154& 0.2431& 0.3189& 0.192& 0.092\\ 0.1069& 0.206& 0.4124& 0.2421& 0.0326\end{array}\right]$$

(19)

$$\begin{array}{c}B=W\times R=\left(0.164,0.148,0.363,0.325\right)\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\times \left[\begin{array}{ccccc}0& 0.596& 0.404& 0& 0\\ 0.0308& 0.2641& 0.5411& 0.1474& 0.0166\\ 0.154& 0.2431& 0.3189& 0.192& 0.092\\ 0.1069& 0.206& 0.4124& 0.2421& 0.0326\end{array}\right]\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}=\left(0.096,0.292,0.396,0.17,0.046\right)\end{array}$$

(20)

As a result, a comprehensive evaluation of the final PV feed-in tariff business model could be obtained,

$$V=0.096\times 1+0.292\times 0.8+0.396\times 0.6+0.17\times 0.4+0.046\times 0.2=0.6444$$

(21)

Similarly, according to the above-mentioned process, MATLAB calculated and simulated the self-consumption business model and the net power settlement business model, respectively, and the result of the comprehensive evaluation of the net power settlement business model was 0.7102 and the result of the comprehensive evaluation of the self-consumption business model was 0.7798. Thus, in today’s socio-economic environment, the business model of PV power generation with the highest stakeholder relevance is the self-consumption business model. In other words, the self-consumption business model has the highest value in terms of promotion and application in all aspects. In fact, the biggest advantage of this model is that it is self-generating and self-consuming, and there is no transaction process, which saves state funds and government subsidies.

6. Conclusions

This study explored the business models of photovoltaic power generation under integrated energy services. This paper used fuzzy mathematical theory to construct the decision model and MATLAB to perform simulations. It was concluded from the analysis and calculations that among the various business models of PV power generation, the one with the highest correlation with stakeholders, namely, investors and government, was the self-consumption business model. In other words, this model provided the best overall performance.

Investors are recommended to use a self-consumption business model for distributed PV power generation projects with third-party involvement. As soon as the PV price reaches a level that is equal to or below the grid retail electricity rate, the state ceases to provide subsidies. As long as the annual electricity consumption is greater than the PV power generation, there is no power trading. The power company charges as before, with no additional services added and no transaction costs. As PV power generation progresses, the self-consumption business model can shift PV power customers from traditional power companies to the usual commercial and industrial enterprises as well as to the end consumer. This is the most adaptable business model for the development of the distributed PV power generation market and the direction of future PV power generation business models.

At the government level, it is necessary to reduce the cost of PV power generation on the one hand, and to ensure the stability of PV power generation costs on the other. In terms of policy development, the government can provide relevant technical and financial support at the national level. For example, by promoting research programs or incentive policies to promote better and faster development of PV technology. This will reduce the cost of PV power generation and promote the rapid expansion of the PV power generation market domestically, which will ultimately benefit the national energy layout. Promoting market-oriented reforms in the electricity sector will separate the revenues from to the power companies from the transactions and revenues and expenses of the companies and the electricity users. On the one hand, it will ensure total revenues for grid companies, and on the other hand, create a market for electricity between users and power generators. As a result, it is possible to create a pattern of large-scale construction and the use of solar energy.

In recent years, the PV power generation industry has experienced rapid growth under integrated energy services worldwide. The business models are unique and innovative. Consequently, to understand and integrate the operation of various business models of PV power generation, it is necessary to continually explore and analyze new models in addition to the existing three types. Business models are typically selected based on the focus of different stakeholders, and the benefits of the business models also vary considerably. To achieve the most value chain benefits for each stakeholder, it is essential that the business model is selected carefully and that the choice is scientific enough. The PV industry should make its investment decisions based on its stakeholder value chain so the sector can develop rapidly and make new breakthroughs. By optimizing business models, it is possible to maintain and coordinate stakeholders’ needs. A crucial challenge for the future development of the PV industry is to choose the appropriate business model.