Evaluating the Feasibility of Low-Carbon Technology in the Sustainable Economic Development of Electricity Using Generative Adversarial Networks

Electricity occupies a central position in energy strategy, and the low-carbon development of the power industry is an important guarantee for achieving sustainable development and carbon dioxide reduction goals. Orderly carrying out low-carbon technology transformation and innovation in electricity, ultimately achieving the goal of sustainable development of the power industry, has strong practical significance. Therefore, this article conducts a feasibility analysis of using generative adversarial networks to evaluate low-carbon technologies in the sustainable economic development of electricity. Firstly, the background of the theory of technological and economic evaluation was analysed and used as a theoretical basis. Secondly, an adversarial network was generated for the study of low-carbon emissions in the power industry, and an evaluation system for low-carbon technology in the sustainable economic development of electricity was constructed. The study showed that the preliminary economic evaluation index system and comprehensive benefit evaluation model for low-carbon technology transformation projects in electricity established in this article have feasibility and reliability for systematic analysis and evaluation of actual projects.


Introduction
The issue of energy and climate change is one of the most concerning issues for countries around the world today.Both lower-middle and high-income countries fully recognize the necessity of transforming economic growth patterns, and "green environmental protection" and "low-carbon economy" have become hot topics [1-2].Studying the low-carbon development capacity of the power grid can provide reference opinions for the power supply and grid construction of power grid companies, as well as the formulation of scientific and reasonable low-carbon emission reduction measures, thereby improving the current imbalance between power load and consumption level, and achieving the goal of continuously enhancing power supply capacity, improving power supply quality, and maintaining stable power supply safety as soon as possible [3][4].After years of development, scholars from various countries have achieved relatively more research results in the field of technological economy.At present, research on the technical and economic evaluation of power projects mainly focuses on new and renewable energy (such as solar, wind, nuclear, etc.) power projects, as well as the economic aspects of the application of new technologies in power plants.Part of the content is the practical application and analysis of a specific evaluation method [5][6][7][8].Maeda H. Murakami S applied the fuzzy evaluation method to the decision selection problem and analysed the case [9].SattyT L et al. respectively studied the practical application of Analytic Hierarchy Process in investment decision-making and conducted extended research [10] Athanasios evaluated the technical, economic, and feasibility of power plant projects using the AHP analysis method [11].Charoenngam C and Yeh C. Y conducted a systematic analysis of the risks in the investment and construction evaluation of hydropower projects [12].MokhtariH et al. used ant colony analysis to study the time cost problem in project management processes [13].FedericoPasin introduced a simple model for analysing project investment returns, which is analysed based on actual cases [14].It can be seen that there have been a large number of achievements in the research on technical and economic evaluation of projects, including research on technical and economic evaluation of power construction projects, power transformation projects, and thermal power technology transformation projects [15][16][17].However, based on the current adherence of countries around the world to the new goal of "low-carbon development" and the new requirements of "energy conservation and environmental protection", there is relatively little research on the economic benefits evaluation of low-carbon technology transformation in thermal power projects.Against the backdrop of a significant increase in energy demand in China, it is more practical to use low-carbon energy as the main driving force for adjusting the energy structure and meeting the rapid growth of energy demand.Due to the common and unique characteristics of low-carbon technology renovation projects in thermal power, it is necessary to correctly identify the actual impact range and degree of different low-carbon technologies on the renovation project when conducting comprehensive benefit evaluation.In order to achieve low-carbon development of the power grid, power grid enterprises need to introduce low-carbon concepts, conduct research on low-carbon development models, tap into the low-carbon potential of the power grid, widely apply low-carbon power technology, and ultimately achieve a "low-carbon power grid" [18].This article will start with the formation mechanism of lowcarbon power grid benefits, and summarize the experience of low-carbon power.Firstly, it will analyse the low-carbon potential and benefits of the power generation side, grid side, and electricity consumption side.Secondly, it will analyse the low-carbon development technology path of the power grid, and then establish a low-carbon technology development benefit evaluation index system for the power grid.

Discriminative Method Based On Generative Adversarial Networks
Through the idea of zero sum game, the discriminative network reaches Nash equilibrium, which means that the difference between the generated sample and the real sample cannot be distinguished, thus obtaining enough false samples to confuse the real ones [19].Generative adversarial networks can automatically learn the data distribution of real samples, that is, complex mapping laws between input and output, and discriminative networks can automatically learn a good discriminative method.
The objective loss function consists of two terms, the first being the loss for discriminating between real data and the latter being the loss for discriminating between false data generated [20].According to the formula, it can be seen that the purpose of the objective function is to maximize the differentiation of D and minimize the difference between the distribution of G and the real data.Figure 1 depicts a schematic diagram of the structure of the generated adversarial network [21].
We can guide the generation of data by inputting y.In the generator, the input noise z and condition y are combined in the hidden representation.

Case Analysis of Carbon Emissions from Electricity
The data collection of this article mainly involves two parts.One part is the carbon emission data of electricity that has been successfully evaluated or reasonably evaluated, and the other part is the indicator data of low-carbon evaluation.A total of 240 pieces of data were collected, and Figure 2 shows the classification of the sample set.
wFigure 2: Classification of training sample sets using survival adversarial network method The carbon emission data of electricity is influenced by a series of factors, and has been divided into three dimensions based on their respective characteristics.But before determining the weights of each indicator, it is necessary to consider whether there is a strong correlation between indicators in the same or different dimensions, in order to provide a judgment basis for the existence of redundant indicators.Figure 3 depicts the sample distribution of carbon emissions from different categories of electricity.As mentioned above, in traditional methods, 4 out of 10 carbon emissions from electricity have incorrect evaluation results, with an accuracy rate of approximately 60%.In this model, only 2 out of 10 electricity carbon emission data have incorrect evaluation results, with an accuracy rate of almost 80%.In addition, among the two inaccurate evaluation results of carbon emissions from electricity, only one evaluation result deviated from the actual evaluation by two value levels, while the other evaluation result remained within one value level deviation.On this basis, we will conduct further testing on the platform.Figure 4 shows the system flow test results of the power sustainability assessment platform.In order to achieve the operation of the business layer on the database, it is necessary to first establish an access connection to the database, and each time the data operation is performed, this connection is taken as a parameter.Therefore, it is necessary to store the connection in the cache during the initial operation of the system.At the same time, according to the system pattern, the structure will be compared and updated with the database during the initial connection.Based on the construction of a sustainability assessment model for electricity, a corresponding value assessment system is established through big data from historical transactions.After inputting relevant information on the carbon emissions of the electricity to be evaluated in the system, the sustainability evaluation value is obtained by the system.During the continuous use of the system, based on new electricity carbon emission data and practical experience, the system has been continuously optimized and improved.Figure 4 shows the system flow test results of the power sustainability assessment platform.

Conclusion
In today's increasingly perfect marketization of the power industry, energy conservation and environmental protection issues are gradually receiving close attention from the country and society.As a major energy producer and consumer, power companies have begun to pay attention to the practical application of new energy-saving and low-carbon environmental protection technologies.Therefore, how to conduct a scientific and reasonable technical and economic evaluation of lowcarbon technology renovation projects for thermal power units has become an important research topic.This article combines the theory of technical and economic evaluation, comprehensive analysis and evaluation methods, and the actual situation of power plant production and operation, and conducts a comprehensive study on the technical and economic evaluation of low-carbon technology transformation projects in thermal power plants.Finally, based on the above research, corresponding improvement measures have been proposed for the low-carbon construction of the future power grid, mainly from the perspective of policy guidance and specific implementation.It should be pointed out that this article only analyses the economic benefits of some low-carbon technologies in evaluating specific low-carbon benefits.The influencing factors considered and the indicator system covered may not be comprehensive, and these aspects will be supplemented and improved in subsequent research.

Figure 1 :
Figure 1: Schematic diagram of the structure of the generated adversarial network

Figure 3 :
Figure 3: Sample distribution of carbon emissions from different categories of electricity Based on the above analysis, we have compiled and collected data on the distribution of carbon emissions different types of electricity, referred to relevant literature and consulted expert opinions, and cited secondary indicators under technical factors.Table 1 presents the analysis results of low-carbon technology and electricity sustainability.

Figure 4 :
Figure 4: the system flow test results of the power sustainability assessment platform The selection of various indicators in the technical and economic evaluation index system for lowcarbon thermal power technology renovation projects, such as social benefit indicators, still needs further improvement and optimization, especially based on the constantly increasing requirements for energy conservation and environmental protection in various countries around the world.

Table 1
Analysis Results of Low Carbon Technology and Electricity Sustainability Table 1 presents the analysis results of low-carbon technology and electricity sustainability.