Multi-scenario driven water supply carbon emission accounting system optimization method and system

Abstract

The present invention relates to the field of carbon emission accounting technology, and discloses a multi-scenario driven water supply carbon emission accounting system optimization method and system. The method constructs a three-dimensional dynamic entropy weight classification model according to the process topology characteristics, water source heterogeneity and spatiotemporal load fluctuations of the water supply system, divides the scenario clusters and calculates the carbon accounting gradient; collects heterogeneous carbon source pulse data to generate a dynamic carbon accounting density function and a heterogeneous entropy weight deviation; constructs and solves a carbon accounting gradient multi-objective optimization model to obtain the optimal carbon accounting gradient and output the optimization result; the present invention solves the problem that the existing carbon emission accounting methods in the water supply industry are inconsistent, inaccurate and not applicable to different scenarios, and provides strong support for urban low-carbon development and energy conservation and emission reduction in the water supply industry.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission accounting, and more specifically, to a multi-scenario driven water supply carbon emission accounting system optimization method and system. Background Art

[0002] As a critical urban infrastructure, the water supply industry generates carbon emissions across multiple stages, including raw water intake, water plant production, and water transmission and distribution. Accurately clarifying the carbon emissions profile of these stages and establishing a scientific accounting system are crucial for low-carbon urban development and energy conservation and emission reduction within the water supply industry. However, the water supply industry currently faces numerous challenges in carbon emissions accounting, and effective solutions are urgently needed.

[0003] Although existing carbon emission accounting related technologies have applications in other fields, they have adaptability issues in the water supply industry. For example, the Chinese patent application with publication number CN114723134A proposes a multi-scenario building carbon emission prediction method and device, which mainly aims at predicting carbon emissions during the building operation period. This method analyzes the current status of building energy consumption and carbon emissions, and constructs a building LEAP scenario model to predict carbon emissions, carbon peak time and peak value. However, there are significant differences between the water supply industry and the construction industry in terms of energy consumption structure, process flow, etc. The water supply industry involves complex water treatment processes, and its energy consumption includes not only electricity, but also indirect carbon emissions caused by the use of chemicals, etc.; while the construction industry mainly consumes energy from various energy-consuming equipment. Therefore, the method of this prior art is difficult to be directly applied to the water supply industry, and it is impossible to accurately calculate the carbon emissions of each link in the water supply process, and cannot meet the water supply industry's demand for accurate quantification of carbon emissions from different treatment processes.

[0004] The Chinese patent application with publication number CN115796340A discloses a method and system for predicting carbon emissions in multiple fields of a city. By acquiring historical terminal energy consumption data, the LEAP model of urban energy consumption is constructed to calculate the total energy carbon emissions of different fields and the city. However, this method focuses on the calculation of energy consumption and carbon emissions at the macro level in multiple fields of the city, and does not deeply consider the impact of the unique water source heterogeneity, process topology characteristics and time-space load fluctuations of the water supply industry on carbon emissions. For example, differences in water quality in different water sources will lead to different energy and chemical consumption in the water treatment process, which in turn affects carbon emissions. However, this existing technology does not analyze such factors that are unique to the water supply industry, making it difficult to achieve refined accounting of carbon emissions in the water supply industry, and cannot meet the requirements of the water supply industry for accurate carbon emission accounting under different scenarios.

[0005] Currently, there is a lack of a unified, accurate, and scenario-specific carbon emissions accounting method for the water supply sector. Existing technologies fail to fully account for the particularities of the water supply sector, making it difficult to comprehensively and accurately calculate its carbon emissions. This inability to meet the demand for accurate carbon emissions data for low-carbon urban development and energy conservation and emission reduction within the water supply sector is unsatisfactory. Summary of the Invention

[0006] This invention is applicable to water supply systems of all sizes and operating modes. Whether it's a large urban water supply network or a small town water supply facility, this method and system can accurately calculate carbon emissions during the water supply process. For example, in urban planning, the calculation results can be used to rationally layout water supply facilities, optimize water supply processes, and reduce carbon emissions. In the daily operations and management of water supply companies, the system can monitor carbon emissions in real time, allowing timely adjustments to operational strategies to achieve energy conservation and emission reduction goals.

[0007] To overcome the aforementioned shortcomings of the prior art, the present invention provides a multi-scenario driven water supply carbon emissions accounting system optimization method and system. By constructing a three-dimensional dynamic entropy weighted hierarchical model, it fully considers the process topology characteristics, water source heterogeneity, and spatiotemporal load fluctuations of the water supply system, achieving a refined analysis of water supply carbon emissions under different scenarios. This method can collect heterogeneous carbon source pulse data in real time, generate an accurate dynamic carbon accounting density function, and construct a carbon accounting gradient multi-objective optimization model to obtain the optimal carbon accounting gradient, thereby outputting optimized carbon emission accounting results, providing strong support for urban low-carbon development and energy conservation and emission reduction in the water supply industry.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A multi-scenario driven water supply carbon emissions accounting system optimization method includes:

[0010] According to the process topology characteristics, water source heterogeneity and spatiotemporal load fluctuations of the water supply system, a three-dimensional dynamic entropy weight classification model is constructed. According to the three-dimensional dynamic entropy weight classification model, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. , where N≥3;

[0011] Real-time collection of heterogeneous carbon source pulse data throughout the water supply process, mapping the heterogeneous carbon source pulse data to the N scenario clusters divided by the accounting system, and generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation ;

[0012] According to the dynamic carbon accounting density function and heterogeneous entropy weight bias , build a carbon accounting gradient multi-objective optimization model; solve the carbon accounting gradient multi-objective optimization model, and iteratively optimize the carbon accounting gradient of each scenario cluster , and obtain the optimal carbon accounting gradient , according to the optimal carbon accounting gradient , output multi-scenario carbon emission accounting optimization results.

[0013] Furthermore, the constructing of the three-dimensional dynamic entropy weight classification model includes:

[0014] Based on the process topology characteristics of the water supply system, the historical power consumption data and historical reagent dosage data of each treatment process are obtained, and the process entropy value E is obtained according to the historical power consumption data and historical reagent dosage data of each treatment process. p ;

[0015] Based on the water source heterogeneity of the water supply system, the raw water BOD concentration and the length of the water transmission network of each water source are obtained, and the water source coupling factor α is generated according to the raw water BOD concentration and the length of the water transmission network of each water source. s ;

[0016] Calculate the load fluctuation coefficient β based on the temporal and spatial load fluctuations of the water supply system t ;

[0017] According to the process entropy value E p , water source coupling factor α s and load fluctuation coefficient β t , construct a three-dimensional dynamic entropy weight classification model.

[0018] Furthermore, the process entropy value E is obtained p include:

[0019] According to the historical power consumption data of each treatment process, the power consumption entropy increase rate of each treatment process is calculated, and the power consumption entropy increase rate of each treatment process is subtracted from the power consumption entropy increase rate of the benchmark process to obtain the relative entropy increase rate;

[0020] The total dosage of each treatment process is obtained based on the historical dosage data of each treatment process, and the ratio of the total dosage of the reagent to the water inflow is calculated to obtain the reagent dissipation ratio;

[0021] The process entropy value E is obtained by linearly combining the relative entropy increase rate and the reagent dissipation ratio. p .

[0022] Furthermore, the generated water source coupling factor α s include:

[0023] The BOD concentration of the raw water of each water source is normalized by min-max to obtain the BOD heterogeneity index; the length of the water supply network of each water source is normalized by min-max to obtain the distance heterogeneity index; the BOD heterogeneity index and the distance heterogeneity index are weighted and summed to obtain the water source coupling factor α s.

[0024] Furthermore, the load fluctuation coefficient β is calculated t include:

[0025] Based on the spatiotemporal load fluctuations of the water supply system, historical monthly water production data and historical carbon emission intensity time series data are obtained;

[0026] Based on the historical monthly water production data, the coefficient of variation of the monthly average water production is calculated to obtain the water production fluctuation index;

[0027] Perform Fourier transform on the historical carbon emission intensity time series data, extract the dominant frequency, calculate the peak period, and obtain the carbon emission intensity periodicity index;

[0028] The load fluctuation coefficient β is obtained by multiplying the water production fluctuation index and the carbon emission intensity periodicity index and taking the square root. t .

[0029] Furthermore, the constructing of the three-dimensional dynamic entropy weight classification model includes:

[0030] Process entropy E p , water source coupling factor α s and load fluctuation coefficient β t Normalization is performed separately;

[0031] Assign a weight coefficient w to the normalized process entropy value p , assign a weight coefficient w to the normalized water source coupling factor s , assign a weight coefficient w to the normalized load fluctuation coefficient t ;W p 、w s 、w t Constitute the entropy weight vector [w p ,w s ,w t ];

[0032] By process entropy value E p , water source coupling factor α s , load fluctuation coefficient β t and entropy weight vector [w p ,w s ,w t ], construct a three-dimensional dynamic entropy weight classification model.

[0033] Furthermore, the carbon accounting gradient of each scenario cluster is calculated include:

[0034] Calculate the mean of the process entropy value for each scenario cluster , the mean of water source coupling factors and the mean of the load fluctuation coefficient ,Will 、 and According to the entropy weight vector [w p ,w s ,w t ] to perform weighted calculations to obtain the carbon accounting gradient for each scenario cluster =[ ],in, represents the carbon accounting gradient of the Nth scenario cluster.

[0035] Furthermore, the heterogeneous carbon source pulse data of the whole water supply process includes the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, the real-time agent addition data and agent transportation distance of the water supply system, the methane emission rate data of the sludge anaerobic fermentation of the sludge treatment unit in the sludge treatment link, and the real-time carbon emission factor of the water supply system. d (t), methane emission rate data r during the landfill process l (t), key sludge treatment parameters of sludge treatment units and key landfill parameters of landfills;

[0036] Generating a dynamic carbon accounting density function include:

[0037] Based on the heterogeneous carbon source pulse data of the whole water supply process, the power entropy pulse signal P is obtained. e , carbon emission chain correlation matrix Γ of the pharmaceutical link c and the carbon emission eddy current coupling coefficient Ω during sludge treatment and landfill s ; Set the fusion weight vector of the scenario cluster ,in is the fusion weight of the Nth scenario cluster; the fusion weight vector based on the scenario cluster , integrating the power entropy pulse signal P e , carbon emission chain correlation matrix Γ c and carbon emission eddy current coupling coefficient Ω s , generating a dynamic carbon accounting density function .

[0038] Furthermore, the heterogeneous carbon source pulse data of the entire water supply process includes real-time power consumption data of each link of the water supply system, real-time carbon emission factors of the regional power grid, real-time chemical addition data of the water supply system, and chemical transportation distance;

[0039] The power entropy pulse signal P is obtained e include:

[0040] Multiply the real-time electricity consumption data of each link in the water supply system with the real-time carbon emission factor of the regional power grid to obtain the real-time carbon emission intensity time series data of electricity consumption;

[0041] Perform pulse detection on real-time carbon emission intensity time series data, extract power carbon emission pulse events, and construct power entropy pulse signal P e .

[0042] Furthermore, the real-time drug addition data includes real-time drug addition amount and drug concentration;

[0043] The carbon emission chain correlation matrix Γ of the pharmaceutical link c include:

[0044] The real-time dosage, concentration and transportation distance of each agent are organized into a three-dimensional correlation matrix; each element in the three-dimensional correlation matrix is ​​quantified as a carbon emission equivalent to generate a carbon emission chain correlation matrix Γ c .

[0045] Furthermore, the construction of the carbon accounting gradient multi-objective optimization model includes:

[0046] Dynamic carbon accounting density function Integrate and calculate total carbon emissions ;

[0047] According to the carbon accounting gradient of each scenario cluster and the water supply volume of the water supply unit corresponding to each scenario cluster , calculate the theoretical total carbon emissions of the water supply system ;

[0048] according to and , defining cross-modal validation bias ;

[0049] Combine cross-modal validation bias and heterogeneous entropy weight bias , construct a carbon accounting gradient multi-objective optimization model.

[0050] A multi-scenario driven water supply carbon emission accounting system optimization system is used to implement the multi-scenario driven water supply carbon emission accounting system optimization method described above. The system includes:

[0051] Scenario classification module: Based on the process topology characteristics, water source heterogeneity and spatiotemporal load fluctuations of the water supply system, a three-dimensional dynamic entropy weight classification model is constructed. Based on the three-dimensional dynamic entropy weight classification model, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. , where N≥3;

[0052] Deviation calculation module: real-time collection of heterogeneous carbon source pulse data of the entire water supply process, mapping the heterogeneous carbon source pulse data to the N scenario clusters divided by the accounting system, and generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation ;

[0053] Carbon accounting optimization module: based on dynamic carbon accounting density function and heterogeneous entropy weight bias , build a carbon accounting gradient multi-objective optimization model; solve the carbon accounting gradient multi-objective optimization model, and iteratively optimize the carbon accounting gradient of each scenario cluster , and obtain the optimal carbon accounting gradient , according to the optimal carbon accounting gradient , output multi-scenario carbon emission accounting optimization results.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] The present invention comprehensively considers multiple factors such as the process topology characteristics, water source heterogeneity, and spatiotemporal load fluctuations of the water supply system, constructs a three-dimensional dynamic entropy weight classification model to divide scenario clusters, and realizes a refined analysis of carbon emissions under different scenarios. By collecting heterogeneous carbon source pulse data in real time, generating a dynamic carbon accounting density function and calculating the heterogeneous entropy weight deviation, the real-time carbon emissions of the system can be more accurately reflected. A carbon accounting gradient multi-objective optimization model is constructed and solved, and the optimal carbon accounting gradient obtained further optimizes the accounting results. This method provides a unified, accurate and applicable accounting solution for different scenarios, which helps water supply companies and urban managers to clearly understand the carbon emission status, formulate targeted emission reduction strategies, improve energy utilization efficiency, reduce carbon emissions, meet the needs of urban low-carbon development and energy conservation and emission reduction in the water supply industry, and promote the sustainable development of the water supply industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0057] Figure 1 This is a principle flow chart of the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention;

[0058] Figure 2 The process entropy value E is obtained in the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention. p Flowchart of the method;

[0059] Figure 3 Generate the water source coupling factor α in the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention sFlowchart of the method;

[0060] Figure 4 Calculate the load fluctuation coefficient β in the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention t Flowchart of the method;

[0061] Figure 5 A flowchart of a method for constructing a three-dimensional dynamic entropy weight classification model in the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention;

[0062] Figure 6 The power entropy pulse signal P is obtained in the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention. e Flowchart of the method;

[0063] Figure 7 A flow chart of a method for obtaining a candidate function set in the multi-scenario driven water supply carbon emission accounting system optimization method of the present invention;

[0064] Figure 8 This is a functional module diagram of the multi-scenario driven water supply carbon emission accounting system optimization system in the present invention. DETAILED DESCRIPTION

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0066] Example 1

[0067] See also Figure 1 As shown, this embodiment provides a multi-scenario driven water supply carbon emission accounting system optimization method, including:

[0068] Step S1000: Based on the process topology characteristics, water source heterogeneity, and spatiotemporal load fluctuations of the water supply system, a three-dimensional dynamic entropy weight classification model is constructed. Based on the three-dimensional dynamic entropy weight classification model, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. , where N≥3;

[0069] Furthermore, step S1000 includes:

[0070] Step S1100: Based on the process topology characteristics of the water supply system, the historical power consumption data and the historical reagent dosage data of each treatment process are obtained, and the process entropy value E is obtained according to the historical power consumption data and the historical reagent dosage data of each treatment process. p;

[0071] Furthermore, if Figure 2 As shown, step S1100 includes:

[0072] Step S1110, based on the process topology characteristics of the water supply system, obtaining historical power consumption data and historical reagent dosage data of each treatment process;

[0073] Step S1120, calculating the power consumption entropy increase rate of each process based on the historical power consumption data of each process, and subtracting the power consumption entropy increase rate of each process from the power consumption entropy increase rate of the benchmark process to obtain a relative entropy increase rate;

[0074] Step S1130, obtaining the total dosage of each treatment process based on the historical dosage data of each treatment process, calculating the ratio of the total dosage of the dosage to the water inflow, and obtaining the dosage dissipation ratio;

[0075] Step S1140: linearly combine the relative entropy increase rate and the reagent dissipation ratio to obtain the process entropy value E p .

[0076] Specifically, the process topology of a water supply system refers to the interconnections and layout of the various treatment processes within the system, as well as their position and role within the overall water supply process. It encompasses the various treatment processes involved, and their combinations, from raw water intake to finished water delivery. The process topology encompasses a variety of treatment methods, including conventional treatment, advanced treatment, and CO2 dosing. Data acquisition methods vary for different treatment processes. For example, for the coagulation and sedimentation phase of conventional treatment, historical power consumption data can be obtained by installing power meters on relevant equipment (such as mixers and pumps) to record their power consumption over time. Historical chemical dosage data can be obtained from the operating records of chemical dosing equipment, such as the time and dosage of each flocculant addition. This data serves as the primary basis for subsequent analysis of the carbon emission characteristics of each process. Its beneficial effect lies in providing a data foundation for the subsequent quantification of carbon emissions from each process. Carbon emission accounting in the water supply industry lacks a unified and accurate methodology, and acquiring this data is a critical step in achieving accurate accounting. By accurately understanding the power consumption and reagent dosage of each treatment process, it is possible to more accurately assess the carbon emissions performance of different processes. For example, if a water plant's advanced treatment process has high power consumption over a period of time, combined with reagent dosage data, further analysis can be conducted to determine whether the process is experiencing excessive carbon emissions. This can provide guidance for subsequent process optimization and carbon emission reduction, thereby helping to address the issue of accurate carbon emission accounting in the water supply industry.

[0077] The entropy increase rate of power consumption is an indicator to measure the disorder degree of power consumption changes of each treatment process. When calculating the entropy increase rate of power consumption, first analyze the energy consumption changes at different time points based on the historical power consumption data of each treatment process. The power consumption is According to the relevant theory of thermodynamic entropy, the entropy increase can be calculated by a specific formula (in is the Boltzmann constant, n' is the total number of process energy consumption states, i' is the index of the process energy consumption state, is the probability of the process under different energy consumption states, which can be calculated based on the proportion of energy consumption data in the total energy consumption data), and then the entropy increase rate of power energy consumption is obtained The energy consumption state of the treatment process refers to the different power consumption conditions exhibited by each treatment process during operation in the water supply system. During the operation of the treatment process of the water supply system, its power consumption is not fixed, but fluctuates due to a variety of factors. Differences in energy consumption at different times, changes in equipment performance, and external environmental factors will lead to different energy consumption states. For example, during the peak water use period, in order to meet the water supply demand, the water pump needs to operate at a higher power, and the power consumption at this time is higher; while during the low water use period, the water pump power is reduced, and the power consumption also decreases accordingly. This forms different energy consumption states. Assuming that the water pump consumes 100 degrees of electricity per hour during peak water use and 50 degrees per hour during the low water use period, these two different power consumption conditions are two different energy consumption states.

[0078] The relative entropy increase rate is calculated by subtracting the entropy increase rate of each treatment process from the entropy increase rate of a benchmark process (usually a representative process with relatively stable energy consumption is used as the benchmark). A higher relative entropy increase rate indicates a higher carbon emission intensity for that process. This is because electricity consumption and carbon emissions are closely correlated, with high energy consumption often associated with high carbon emissions. By calculating the relative entropy increase rate, we can visually compare the carbon emission intensity differences between each treatment process relative to the benchmark process. For example, a high relative entropy increase rate for a process indicates a high degree of disorder in its electricity consumption, potentially leading to inefficient energy utilization and increased carbon emissions. This step is beneficial in that it can quickly identify processes with relatively high carbon emission intensities, providing important clues for further process optimization and carbon emission reduction. This helps address the issue of process carbon emission intensity assessment in carbon emission accounting within the water supply industry, making subsequent emission reduction measures more targeted.

[0079] The ratio of the total dosage of the reagent to the water inflow is the reagent dissipation ratio. For example, in a certain treatment process, the total dosage of the reagent is M' and the water inflow is Q' within a certain period of time, then the reagent dissipation ratio is . The larger the reagent dissipation ratio, the more reagents the process consumes when treating the same amount of water. Since the production, transportation and use of reagents may generate carbon emissions, the larger the reagent dissipation ratio, the higher the carbon emission intensity of the process. Its beneficial effect is to evaluate the carbon emission intensity of the process from the perspective of reagent use. In the carbon emission accounting of the water supply industry, reagent use is an important source of carbon emissions. By calculating the reagent dissipation ratio, we can clearly understand the impact of each treatment process on carbon emissions in terms of reagent use. For example, for the disinfection process of a water plant, if its reagent dissipation ratio is too high, we can further analyze whether there is excessive reagent addition, and then reduce carbon emissions by optimizing the reagent addition amount, etc., which will help to achieve a comprehensive assessment and precise control of carbon emissions from various processes in the water supply system.

[0080] The linear combination method usually takes the form of weighted summation, that is, E p =q1×relative entropy increase rate+q2×agent dissipation ratio (where q1 and q2 are weight coefficients of relative entropy increase rate and agent dissipation ratio, respectively, and q1+q2=1. The weight coefficient can be determined based on actual conditions, through expert experience or machine learning algorithms, to balance the contribution of relative entropy increase rate and agent dissipation ratio to process entropy value). In this way, the process entropy value E p The two main sources of carbon emissions, energy and chemicals, are comprehensively considered. p It can more comprehensively evaluate the carbon emission characteristics of each process and more accurately reflect the overall carbon emission of a treatment process than considering only the power consumption or reagent dosage. For example, for a process with a low entropy increase rate of power consumption but a high reagent dissipation ratio, the carbon emission may be low if the power consumption is considered alone. However, after comprehensively considering the use of reagents, the process entropy value E p It can be found that its overall carbon emission intensity is not low, which helps to avoid misjudgment of process carbon emissions. Secondly, by introducing the process entropy value E p , we can quantitatively evaluate the difference in contribution of different treatment processes to carbon emissions. In the water supply system, different treatment processes have different contributions to carbon emissions. By comparing the E p The value can clarify which processes are the main sources of carbon emissions, provide a basis for the selection and optimization of process routes, and reveal the important impact of process route selection and optimization on carbon emission reduction, thereby promoting the water supply industry to develop in a low-carbon direction and solving the problem of comprehensively evaluating process carbon emissions and guiding carbon emission reduction in carbon emission accounting in the water supply industry.

[0081] Step S1200: Based on the water source heterogeneity of the water supply system, the raw water BOD concentration and the length of the water pipe network of each water source are obtained, and the water source coupling factor α is generated according to the raw water BOD concentration and the length of the water pipe network of each water source. s ;

[0082] Further, if Figure 3 As shown, step S1200 includes:

[0083] Step S1210: Based on the water source heterogeneity of the water supply system, the raw water BOD concentration and the length of the water transmission network of each water source are obtained;

[0084] Specifically, water source heterogeneity refers to the differences in water quality and water transmission distance between different water sources within a water supply system. The raw water BOD concentration, also known as biochemical oxygen demand (BOD), reflects the amount of dissolved oxygen consumed by microorganisms during the decomposition of oxidizable organic matter in water under specific conditions. It is typically expressed in milligrams per liter (mg / L). Higher BOD values ​​indicate greater amounts of oxidizable organic matter in the water and poorer water quality. The length of the water transmission network reflects the distance between the water source and the water supply area. In practice, raw water BOD concentration can be determined by establishing water quality monitoring points at each water source, regularly collecting water samples, and sending them to laboratories for testing and analysis. BOD concentration is measured using specialized water quality testing equipment and methods. For example, the standard dilution method involves diluting water samples and incubating them under specified conditions. The BOD concentration is calculated by measuring the difference in dissolved oxygen values ​​before and after incubation. Regarding the length of the water transmission network, Geographic Information Systems (GIS) technology, combined with a water supply system layout, can be used to accurately measure the actual length of the water pipelines from each water source to the water supply area. The beneficial effect of this step is that it provides basic data for subsequent analysis of the impact of water sources on carbon emissions. In the carbon emission accounting of the water supply industry, water source is one of the important influencing factors. Accurately obtaining the raw water BOD concentration and the length of the water transmission network helps to understand the water quality conditions and water transmission distance differences of different water sources. For example, if the raw water BOD concentration of a water source is high, it means that more energy and chemicals may be needed to purify the water quality during the subsequent water treatment process, thereby increasing carbon emissions; a longer water transmission network length will lead to increased energy consumption during the water transmission process, which will also increase carbon emissions. Through these data, the carbon emissions of different water sources can be more comprehensively evaluated, providing strong support for solving the problem of accurate carbon emission accounting in the water supply industry.

[0085] Step S1220, performing min-max normalization on the raw water BOD concentration of each water source to obtain a BOD heterogeneity index;

[0086] Specifically, min-max normalization is a data standardization method whose purpose is to map the data range to the interval [0,1], eliminating differences in dimensions and numerical ranges between different data sets and making the data comparable. The normalized raw water BOD concentration is used as the BOD heterogeneity index; a larger BOD heterogeneity index indicates relatively poorer water quality at that source. This is because, after normalization, the closer the value is to 1, the higher the raw water BOD concentration is among all water sources, indicating a higher content of oxidizable organic matter and poorer water quality. This step has the beneficial effect of converting raw water BOD concentration into a comparable dimensionless indicator, making it easier to accurately measure the impact of water quality differences between different water sources on carbon emissions in subsequent analyses. In water supply systems, varying water quality can lead to differences in water treatment processes and energy consumption, which in turn affects carbon emissions. The BOD heterogeneity index allows for a clear comparison of the potential impact of water quality on carbon emissions at different water sources. For example, when grading scenarios, the scenario clusters where water sources with poor water quality are located can be divided into categories with relatively high carbon emission intensity based on the size of the BOD heterogeneity index, providing a basis for achieving refined carbon emission accounting and management, and helping to solve the problem of assessing the impact of water quality on different water sources in carbon emission accounting in the water supply industry.

[0087] Step S1230, performing min-max normalization on the length of the water pipe network of each water source to obtain a distance heterogeneity index;

[0088] Specifically, the normalized length of the water supply network is used as the distance heterogeneity index. A larger distance heterogeneity index indicates a relatively longer water supply distance for that water source. The closer the normalized value is to 1, the longer the water supply network length is among all water sources. This step has the beneficial effect of converting water supply network length into a uniform and comparable metric, highlighting the impact of differences in water supply distances on carbon emissions across different water sources. Water supply distances are directly related to energy consumption during water supply. Longer water supply distances require more power to maintain water delivery, thereby increasing carbon emissions. The distance heterogeneity index allows for a visual comparison of carbon emission differences across water sources due to water supply distance. During scenario grading, the distance heterogeneity index can be used to categorize scenario clusters containing water sources with long water supply distances as having higher carbon emission intensity. This provides data support for accurate carbon emission accounting and the development of targeted emission reduction strategies, helping to address the issue of assessing the impact of water supply distance on carbon emission accounting within the water supply industry.

[0089] Step S1240: weighted sum of the BOD heterogeneity index and the distance heterogeneity index to obtain the water source coupling factor α s .

[0090] Specifically, the water source coupling factor α sTaking into account the two key factors of water quality and water transmission distance, it can more comprehensively and accurately reflect the comprehensive impact of water sources on carbon emissions. Compared with considering water quality or water transmission distance alone, α s It can more accurately assess the carbon emissions of different water sources. For example, for a water source with poor water quality but a short water delivery distance, and a water source with good water quality but a long water delivery distance, the α s Their carbon emission intensity can be comprehensively judged, avoiding the one-sidedness of single factor evaluation. s It provides a criterion for water source dimension for scenario classification. When constructing a three-dimensional dynamic entropy weight classification model, α s As an important indicator, it can be used to divide water supply units with similar water source characteristics into the same scenario cluster, achieve refined scenario division, make carbon accounting more in line with actual conditions, help solve the problem of inaccurate scenario division in carbon emission accounting in the water supply industry, and provide strong support for subsequent carbon accounting and emission reduction measures.

[0091] Step S1300: Calculate the load fluctuation coefficient β based on the temporal and spatial load fluctuation of the water supply system. t ;

[0092] Further, if Figure 4 As shown, step S1300 includes:

[0093] Step S1310: Based on the spatiotemporal load fluctuations of the water supply system, historical monthly water production data and historical carbon emission intensity time series data are obtained; the spatiotemporal load fluctuations include the monthly average water production and the carbon emission intensity peak period;

[0094] Specifically, spatiotemporal load fluctuations encompass aspects such as average monthly water production and peak periods of carbon emission intensity. Historical monthly water production data can be obtained through water metering devices installed in the water supply system, which record monthly water production figures. For example, smart water meters can be installed on the outlet pipes of water plants. The monthly reading changes represent the water production figures for that month. Historical carbon emission intensity time series data can be obtained by installing carbon emission monitoring devices at various points in the water supply system, such as energy-consuming equipment (such as pumps and motors). By combining energy consumption data with corresponding carbon emission factors, carbon emission intensity is calculated and recorded chronologically. This step aims to comprehensively understand the temporal load variations and corresponding carbon emissions of the water supply system. Its beneficial effect is to provide data support for accurate analysis of the dynamic characteristics of the water supply system. In the absence of a unified and accurate carbon emission accounting method for the water supply industry, this data serves as the foundation for subsequent calculations and analyses. By analyzing historical monthly water production data and carbon emission intensity time series data, it is possible to identify monthly load variations and patterns in carbon emissions within the water supply system. For example, it was found that water consumption is higher in the summer, resulting in increased water production. Carbon emissions intensity may also increase due to high equipment load. These patterns provide a deeper understanding of the carbon emissions behavior of water supply systems, providing a basis for developing targeted carbon accounting strategies and emission reduction measures. This solves the problem of collecting temporal data in carbon emissions accounting in the water supply industry, making subsequent analysis and calculations more accurate and targeted.

[0095] Step S1320: Calculate the coefficient of variation of the monthly average water production based on the historical monthly water production data to obtain a water production fluctuation index;

[0096] Specifically, the coefficient of variation is a statistic that measures the degree of data dispersion. It eliminates the influence of data dimension and can more accurately compare the fluctuations of different data sets. To calculate the coefficient of variation of the monthly average water production, we must first calculate the monthly average water production. Assume that the water production of a water supply system from January to December is , then the average monthly water production Then calculate the standard deviation of water production The coefficient of variation of the average monthly water production , the coefficient of variation is the water production fluctuation index. The larger the water production fluctuation index, the more intense the dynamic changes in the water supply load. For example, if the water production fluctuation index of a water supply system is 0.3 within a year, and the water production fluctuation index of another water supply system is 0.5, then the dynamic changes in the water supply load of the latter are more drastic, which means that the difference in its water production between different months is greater. The purpose of this step is to quantify the degree of dynamic changes in the water supply load. Its beneficial effect is to provide quantitative indicators for evaluating the stability of the water supply system and potential changes in carbon emissions. Fluctuations in water supply load will affect the operating efficiency and energy consumption of the equipment, and thus affect carbon emissions. The water production fluctuation index can provide an intuitive understanding of the changes in the water supply load. When the water production fluctuation index is high, it means that the load of the water supply system is unstable, which may cause the equipment to frequently adjust its operating status, increase energy consumption and carbon emissions. This helps water supply companies to promptly identify unstable factors in the system and take corresponding measures to optimize them, such as adjusting equipment operation strategies or optimizing water supply scheduling to reduce carbon emissions. It solves the problem of quantitative assessment of dynamic changes in water supply load in carbon emission accounting in the water supply industry, and provides an important basis for subsequent more accurate carbon emission accounting.

[0097] Step S1330: Perform Fourier transform on the historical carbon emission intensity time series data, extract the dominant frequency, calculate the peak period, and obtain the carbon emission intensity periodicity index;

[0098] Specifically, Fourier transform is a mathematical transformation that can convert time domain signals into frequency domain signals, revealing the distribution of different frequency components in the signal. For historical carbon emission intensity time series data, the carbon emission intensity data is converted from the time domain to the frequency domain through Fourier transform to obtain the spectrum. In the spectrum, the frequency component with a larger amplitude is the dominant frequency of the signal; the peak period refers to the time interval when the carbon emission intensity peaks. When calculating the peak period, first calculate the peak period based on the dominant frequency. , calculate the period , this period is the peak period of carbon emission intensity. The carbon emission intensity periodicity index can be obtained by performing some standardization on the peak period. For example, if a base period is set , Carbon Emission Intensity Cyclical Index . The larger the periodicity index, the more paroxysmal the carbon emissions. For example, if the carbon emission intensity periodicity index of one water supply system is 3 and that of another is 2, the carbon emission of the system with an index of 3 is more paroxysmal, that is, the change in its carbon emission intensity at the peak moment is more significant. The purpose of this step is to analyze the periodic characteristics of carbon emission intensity. Its beneficial effect is to help understand the temporal distribution pattern of carbon emissions. In the water supply system, the periodic changes in carbon emission intensity are related to many factors, such as seasonal changes in water demand and equipment maintenance cycles. By extracting the dominant frequency and calculating the peak period to obtain the carbon emission intensity periodicity index, these patterns can be clearly grasped. For example, it was found that during the peak period of electricity consumption in the summer, due to the high load operation of the equipment, the carbon emission intensity of a water supply system showed obvious periodic changes, and the periodicity index was relatively high. This is of great significance for rationally arranging equipment maintenance time and optimizing energy use strategies. It can perform equipment maintenance during off-peak periods based on the cyclical changes in carbon emissions, reducing the risk of equipment failure during high carbon emission periods. It also provides more accurate time dimension information for carbon accounting, solves the problem of cyclical analysis of carbon emissions in carbon emission accounting in the water supply industry, and makes carbon accounting more in line with actual conditions.

[0099] Step S1340: Take the square root of the product of the water production fluctuation index and the carbon emission intensity periodicity index to obtain the load fluctuation coefficient β t .

[0100] Specifically, β t The larger the β is, the stronger the spatiotemporal dynamics of carbon emissions are, requiring higher accounting frequency and time resolution. t It comprehensively reflects the dynamic changes of water supply load (reflected by the water production fluctuation index) and the cyclical changes of carbon emissions (reflected by the carbon emission intensity cyclical index). t When the value is large, it means that the load changes and carbon emissions changes of the water supply system in the time dimension are more drastic. The traditional low-frequency accounting method is difficult to accurately reflect its carbon emissions, so a higher accounting frequency and finer time resolution are required to capture these changes. The purpose of this step is to comprehensively consider the dynamic characteristics of water supply load and carbon emissions, and construct an indicator that can fully reflect the temporal and spatial load fluctuations. Its beneficial effect is to provide a time dimension criterion for scenario classification, so that carbon accounting can better match the dynamic changes of the water supply system. When constructing a three-dimensional dynamic entropy weight classification model, the load fluctuation coefficient β t As an important indicator, water supply units with similar spatiotemporal load fluctuation characteristics can be divided into the same scenario cluster. tScenario clusters with larger values ​​can adopt higher accounting frequencies and more refined accounting methods to improve the accuracy of carbon accounting. This helps address the problems of inaccurate scenario divisions and mismatch between accounting methods and actual conditions in carbon emissions accounting in the water supply industry, improves the adaptability and accuracy of the carbon accounting system, and provides stronger support for energy conservation, emission reduction, and low-carbon development in the water supply industry.

[0101] Step S1400, according to the process entropy value E p , water source coupling factor α s and load fluctuation coefficient β t , construct a three-dimensional dynamic entropy weight classification model;

[0102] Furthermore, if Figure 5 As shown, step S1400 includes:

[0103] Step S1410: process entropy value E p , water source coupling factor α s and load fluctuation coefficient β t Normalization is performed separately;

[0104] Step S1420: assign a weight coefficient w to the normalized process entropy value. p , assign a weight coefficient w to the normalized water source coupling factor s , assign a weight coefficient w to the normalized load fluctuation coefficient t ;W p 、w s 、w t Constitute the entropy weight vector [w p ,w s ,w t ];

[0105] Step S1430, the process entropy value E p , water source coupling factor α s , load fluctuation coefficient β t and entropy weight vector [w p ,w s ,w t ], construct a three-dimensional dynamic entropy weight classification model.

[0106] Specifically, the process entropy value E p , water source coupling factor α s and load fluctuation coefficient β tThe purpose of performing normalization processing separately is to lay the foundation for the subsequent construction of the entropy weight vector and the three-dimensional dynamic entropy weight classification model. When constructing the three-dimensional dynamic entropy weight classification model, after unifying the dimensions and numerical ranges, the impact of different indicators on carbon emission intensity can be compared and analyzed on the same scale. For example, when determining the weights of each indicator, the impact of certain indicators will not be excessively amplified or reduced due to differences in dimensions and numerical ranges, ensuring the scientific nature and accuracy of the model. When dividing the carbon emissions of the water supply system into scenarios, the normalized data can more reasonably reflect the relative importance of each factor, making the division results more consistent with the actual situation. This helps to solve the problem of inaccurate analysis caused by differences in data dimensions and numerical values ​​in the carbon emission accounting of the water supply industry, and improves the accuracy of subsequent scenario division and carbon accounting.

[0107] Weight coefficient w p ,w s ,w t Used to characterize different indicators (process entropy value E p , water source coupling factor α s and load fluctuation coefficient β t ) to carbon emission intensity. These weight coefficients can be determined based on expert experience or machine learning algorithms. Taking the expert experience method as an example, experts will evaluate each indicator based on the actual operation of the water supply system, historical data, and in-depth understanding of the impact of various factors on carbon emission intensity, and then determine the corresponding weight coefficient. If the process link in a water supply system has a more critical impact on carbon emissions, experts may assign a process entropy value E p Relatively high weight coefficient w p If the water quality and water transmission distance of the water source have a significant impact on carbon emissions, the water source coupling factor α s The weight coefficient w s will be relatively large. The machine learning algorithm automatically determines the weight coefficient that can best reflect the relationship between each indicator and carbon emission intensity through learning and analyzing a large amount of historical data. The purpose of step S1420 is to quantify the degree of influence of each indicator on carbon emission intensity and construct an entropy weight vector so that various factors can be comprehensively considered in subsequent models. Its beneficial effect is that it can flexibly adjust the importance of each indicator according to the characteristics and actual situation of the water supply system. By reasonably setting the weight coefficient, the role of different factors in carbon emissions can be more accurately reflected. For example, in a water supply system that is mainly based on deep treatment technology and has relatively concentrated water sources, by adjusting the weight coefficient and highlighting the impact of process entropy value on carbon emission intensity, carbon emissions can be more accurately assessed. This is of great significance for solving the problem of comprehensive evaluation of multiple factors in carbon emission accounting in the water supply industry, making the carbon accounting results more in line with actual conditions, and providing a more reliable basis for formulating targeted emission reduction strategies.

[0108] Step S1430 constructs a three-dimensional dynamic entropy weight classification model based on the normalized data and entropy weight vector obtained above. p , water source coupling factor α s and load fluctuation coefficient β t As three dimensions, the data from these three dimensions are weighted and integrated using an entropy weight vector. During the construction process, the data from each dimension is multiplied by the corresponding weight coefficient and then accumulated to obtain a value that comprehensively reflects the multidimensional heterogeneity of carbon emissions from the water supply system. The purpose of this step is to construct a model that can comprehensively consider the three core influencing factors of process, water source, and load to achieve a multidimensional characterization and scenario division of carbon emissions from the water supply system. The beneficial effects are significant. First, this model can comprehensively and systematically analyze the heterogeneity of carbon emissions from the water supply system. Different water supply units have differences in process, water source, and load. The three-dimensional dynamic entropy weight classification model can integrate these differences and more accurately reflect the carbon emission characteristics of each water supply unit. Second, scenario division based on this model can divide water supply units with similar carbon emission characteristics into the same scenario cluster, achieving refined and hierarchical scenario division. This will help to formulate more targeted carbon accounting methods and emission reduction strategies based on the characteristics of different scenario clusters, improve the accuracy and effectiveness of carbon accounting, provide strong support for energy conservation and emission reduction in the water supply industry, and solve the problem of inaccurate scenario division and inability to fully consider the impact of multiple factors in carbon emission accounting in the water supply industry.

[0109] Step S1500: Divide the accounting system into N scenario clusters based on the three-dimensional dynamic entropy weight classification model, and calculate the carbon accounting gradient of each scenario cluster. , where N≥3.

[0110] Furthermore, step S1500 includes:

[0111] Step S1510: using the process entropy value E in the three-dimensional dynamic entropy weight classification model p , water source coupling factor α s and load fluctuation coefficient β t For three dimensions, three-dimensional spatial mapping is performed on each water supply unit of the water supply system to form a sample point set;

[0112] Step S1520, performing fuzzy C-means clustering on the sample point set to obtain N scenario clusters;

[0113] Step S1530: Calculate the mean of the process entropy value of each scenario cluster , the mean of water source coupling factors and the mean of the load fluctuation coefficient ,Will 、 and According to the entropy weight vector [w p ,w s ,w t ] to perform weighted calculations to obtain the carbon accounting gradient for each scenario cluster =[ ],in, represents the carbon accounting gradient of the Nth scenario cluster.

[0114] Specifically, each water supply unit of the water supply system in step S1510 refers to a part of the water supply system that has an independent water supply function or has obvious differences in carbon emission characteristics, such as different water plants, different water supply areas, etc. Taking a city water supply system as an example, it contains three water plants A, B, and C. Each water plant has its own characteristics in terms of process, water source, and load. These three water plants can be regarded as three different water supply units. For each water supply unit, its process entropy value E p , water source coupling factor α s and load fluctuation coefficient β t As the coordinate value in three-dimensional space, a point is determined in three-dimensional space. The points corresponding to all water supply units are collected together to form a sample point set. For example, the process entropy value E of water plant A is p is 0.7, the water source coupling factor α s is 0.5, the load fluctuation coefficient β t is 0.6, then in three-dimensional space, the coordinates of the point corresponding to Water Plant A are (0.7, 0.5, 0.6). After determining the points corresponding to all water plants in the city's water supply system, a set of sample points is obtained. The purpose of this step is to present the carbon emission-related characteristics of each water supply unit in the water supply system in the form of three-dimensional space points, providing a data basis for subsequent cluster analysis. Its beneficial effect is that it can intuitively display the distribution of each water supply unit in three-dimensional space, making it easier to discover the similarities and differences between different water supply units. By observing the distribution of the sample point set, it is possible to preliminarily determine which water supply units have similar carbon emission characteristics in terms of process, water source and load, providing an intuitive basis for subsequent scenario division. This helps to solve the problem of visualizing and preliminarily classifying the carbon emission characteristics of water supply units in carbon emission accounting in the water supply industry, making subsequent scenario division more targeted and reasonable.

[0115] Fuzzy C-means clustering is a commonly used clustering algorithm. Its principle is to determine the optimal cluster centers and the membership of each sample point to each cluster center by minimizing an objective function. In this step, fuzzy C-means clustering is performed on the sample point set obtained in step S1510, dividing the sample points into N scenario clusters (N ≥ 3). The N cluster centers obtained thus determine the N scenario clusters, and each sample point is assigned to a corresponding scenario cluster based on its membership. During the clustering process, the closer a sample point is to a cluster center, the higher its membership to that cluster center. For example, assume that N = 4 and fuzzy C-means clustering is performed on the sample point set of a water supply system. After multiple iterations, the objective function converges, resulting in four cluster centers, C1, C2, C3, and C4. Sample point x'1 has the highest membership to C1, so x'1 is assigned to the scenario cluster centered on C1. In this way, all sample points are divided into four scenario clusters, completing the scenario division. The purpose of step S1520 is to cluster the water supply units according to their carbon emission characteristics, group the water supply units with similar characteristics into the same scenario cluster, and realize scenario grading. Its beneficial effect is that it can classify complex water supply systems according to their carbon emission characteristics, making subsequent carbon accounting and management more efficient. Different scenario clusters have different carbon emission characteristics, and for each scenario cluster, a carbon accounting method and emission reduction strategy that is more in line with its characteristics can be formulated. For example, for a scenario cluster dominated by high-energy consumption processes, we can focus on process optimization to reduce carbon emissions; for scenario clusters with poor water quality at the water source and long water transmission distances, we can consider optimizing water source selection or improving water transmission methods. This helps to solve the problem of unclear scenario division and inability to manage in a targeted manner in carbon emission accounting in the water supply industry, and improves the accuracy of carbon accounting and the effectiveness of emission reduction measures.

[0116] The purpose of step S1530 is to quantify the carbon emission intensity of each scenario cluster and provide key indicators for subsequent carbon accounting and emission reduction analysis. Its beneficial effect is that by calculating the carbon accounting gradient, the carbon emission intensity level of each scenario cluster can be clearly understood. The larger the carbon accounting gradient, the higher the carbon emission intensity level of the scenario cluster, and a finer accounting granularity is required in subsequent carbon accounting. For example, for scenario clusters with higher carbon accounting gradients, more accurate carbon emission calculation methods and stricter monitoring measures can be used to ensure the accuracy of carbon emission accounting. At the same time, this also helps to formulate differentiated emission reduction strategies based on the carbon accounting gradients of different scenario clusters, focusing on emission reduction for scenario clusters with high carbon emission intensity, improving emission reduction efficiency, solving the problem of quantitative assessment and targeted management of carbon emission intensity of different scenario clusters in carbon emission accounting in the water supply industry, and providing strong support for achieving the carbon emission reduction goals of the water supply industry.

[0117] Step S2000: Real-time collection of heterogeneous carbon source pulse data of the entire water supply process, mapping the heterogeneous carbon source pulse data to the N scenario clusters divided by the accounting system, and generating a dynamic carbon accounting density function. , and calculate the heterogeneous entropy weight deviation The heterogeneous carbon source pulse data of the whole water supply process includes the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, the real-time agent addition data and agent transportation distance of the water supply system, the methane emission rate data of the sludge anaerobic fermentation of the sludge treatment unit in the sludge treatment link, and the real-time carbon emission factor of the water supply system. d (t), methane emission rate data r during the landfill process l (t), key sludge treatment parameters of sludge treatment units and key landfill parameters of landfills;

[0118] Furthermore, step S2000 includes:

[0119] Step S2100: collect the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, calculate the real-time carbon emission intensity time series data based on the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, and obtain the power entropy pulse signal P e ;

[0120] Furthermore, if Figure 6 As shown, step S2100 includes:

[0121] Step S2110, collecting real-time power consumption data of each link of the water supply system;

[0122] Step S2120, obtaining the real-time carbon emission factor of the regional power grid;

[0123] Step S2130: multiply the real-time power consumption data of each link of the water supply system by the real-time carbon emission factor of the regional power grid to obtain the real-time carbon emission intensity time series data of power consumption;

[0124] Step S2140: perform pulse detection on the real-time carbon emission intensity time series data, extract the power carbon emission pulse event, and construct the power entropy pulse signal P e .

[0125] Specifically, the water supply system encompasses multiple stages, including water intake, water transmission and distribution, and water treatment. Each of these stages consumes electricity. For example, water pumps in the water intake stage require electricity to extract raw water; during water transmission and distribution, the equipment in the booster pump station relies on electricity to operate; and during water treatment, the various equipment involved in processes such as coagulation, sedimentation, filtration, and disinfection also consume electricity. To collect this real-time data, online power consumption monitoring devices must be deployed on the relevant equipment. These devices, such as smart meters and power monitoring sensors, record the equipment's power consumption in real time and transmit this data to a data collection center via wired or wireless communication. Given the current lack of a unified and accurate carbon emissions accounting method for the water supply industry, obtaining real-time power consumption data for each stage is crucial for achieving accurate accounting. Accurately understanding power consumption at each stage allows for a more precise assessment of the relationship between energy consumption and carbon emissions. For example, if the electricity consumption of the water treatment link of a water plant suddenly increases during a certain period of time, subsequent calculations can further analyze the changes in carbon emissions in this link, providing a basis for determining whether there is energy waste or equipment abnormality in this link, thereby helping to solve the problem of accuracy in carbon emission accounting in the water supply industry.

[0126] The real-time carbon emission factor of a regional power grid reflects the carbon emissions generated per unit of electricity consumed by the regional power grid. This factor is not a fixed value and fluctuates with factors such as the power grid's energy mix (such as the proportion of different energy sources, such as thermal power, hydropower, and wind power), the level of power generation technology, and the operating efficiency of power generation equipment. The real-time carbon emission factor is typically obtained through real-time data integration with the regional power grid's data interface, or by obtaining real-time updated data from relevant energy management departments and professional data platforms. For example, energy management departments in some regions monitor the power generation and corresponding carbon emissions of various power generation methods in the power grid in real time. After calculation and compilation, they provide real-time carbon emission factor data for the regional power grid. Step S2120 establishes a direct link between the water supply system's electricity consumption and carbon emissions. When calculating the water supply system's carbon emissions, simply knowing the electricity consumption data is not enough; the carbon emission factor of the regional power grid must also be combined to determine the corresponding carbon emissions. For example, in a regional power grid dominated by thermal power generation, the carbon emission factor is relatively high, so the water supply system will generate more carbon emissions for the same amount of electricity consumed. In contrast, in a regional power grid with a higher proportion of clean energy, the carbon emission factor is lower, and the carbon emissions corresponding to the same amount of electricity consumed are also lower. Accurately obtaining real-time carbon emission factors helps to accurately calculate the carbon emissions of water supply systems, improve the accuracy of carbon emission accounting, and solve the problem of inaccurate accounting in the water supply industry due to the inability to accurately link electricity consumption and carbon emissions.

[0127] After obtaining the real-time electricity consumption data for each link in the water supply system and the real-time carbon emission factor of the regional power grid, the two are multiplied together to obtain the real-time carbon emission intensity time series data of the electricity energy consumption. Step S2130 converts the electricity consumption data into practical carbon emission intensity data, providing direct data support for the subsequent analysis of the dynamic changes in the carbon emissions of the water supply system. By generating real-time carbon emission intensity time series data, it is possible to intuitively understand the changes in the carbon emission intensity of the water supply system at different time points and different links. For example, analyzing this data may reveal that during the peak water consumption period in summer, due to the high load operation of water supply equipment, electricity consumption increases. At the same time, the carbon emission factor of the regional power grid increases due to the change in energy structure caused by the increase in electricity demand, resulting in a significant increase in the real-time carbon emission intensity of the water supply system. This information helps water supply companies to promptly identify the peak periods and key links of carbon emissions, provide a basis for formulating targeted emission reduction strategies, and solve the problem of monitoring the dynamic changes of carbon emissions in carbon emission accounting in the water supply industry.

[0128] Pulse detection is a signal processing technology used to identify and extract pulse features in signals. For real-time carbon emission intensity time series data, appropriate thresholds and detection algorithms are set to determine which parts of the data constitute pulse events. For example, when the carbon emission intensity at a certain moment is significantly higher than the average level of the period before and after it, and exceeds a pre-set threshold, the data at that moment and a nearby period with obvious change characteristics can be identified as a power carbon emission pulse event. Then, based on these detected pulse events, the power entropy pulse signal P is constructed. e During the construction process, the pulse amplitude reflects the intensity of carbon emissions. The larger the pulse amplitude, the higher the carbon emission intensity at that moment. The pulse frequency reflects the dynamic change characteristics of carbon emissions. The higher the pulse frequency, the more frequent the change of carbon emissions. e It can describe the dynamic carbon footprint of the energy consumption of the water supply system, perceive the suddenness and intermittency of carbon emissions, and enrich the description of the time series characteristics of carbon emissions. For example, in a certain link of the water supply system, there may be a sudden increase in power consumption due to equipment failure, which in turn causes an instantaneous increase in carbon emission intensity. This sudden change will be reflected in the power entropy pulse signal P e It is manifested as a pulse with a large amplitude. By analyzing the power entropy pulse signal P e Water supply companies can promptly detect these abnormal carbon emission changes and take appropriate measures, such as checking equipment operation and optimizing equipment scheduling, to reduce carbon emissions. This signal also provides important foundational data for the subsequent construction of a dynamic carbon accounting density function, facilitating the refined accounting of carbon emissions from water supply systems and addressing the issue of insufficient description of the dynamic characteristics of carbon emissions in carbon emission accounting within the water supply industry.

[0129] Step S2200: Obtain the real-time drug addition data and drug transportation distance of the water supply system, and obtain the carbon emission chain correlation matrix Γ of the drug link based on the real-time drug addition data and drug transportation distance of the water supply system. c ;

[0130] Further, if Figure 7 As shown, step S2200 includes:

[0131] Step S2210, obtaining real-time drug addition data and drug transportation distance, wherein the real-time drug addition data includes real-time drug addition amount and drug concentration;

[0132] Step S2220, organizing the real-time dosage, concentration, and transportation distance of each drug into a three-dimensional correlation matrix;

[0133] Step S2230: quantify each element in the three-dimensional correlation matrix into carbon emission equivalents to generate a carbon emission chain correlation matrix Γ c .

[0134] Specifically, steps S2210 and S2220 aim to collect real-time data related to chemical usage in the water supply system. This data forms the basis for subsequent analysis of carbon emissions from the chemical process. Real-time chemical addition data covers the real-time dosage and concentration of various chemical types, while chemical transportation distance records the distance from the chemical production site to the point of use in the water supply system. Obtaining this data is crucial for accurately assessing carbon emissions from chemical use. In practice, real-time chemical addition data can be obtained using an automatic chemical dosing system. This system typically features precise metering devices that monitor and record parameters such as the dosage and concentration of various chemical types in real time. For example, in the flocculation treatment process at a water plant, the automatic chemical dosing system can accurately determine the specific dosage and concentration of each flocculant added. Chemical transportation distance can be updated through a real-time connection to the logistics information system interface provided by the chemical supplier. The supplier's logistics system can track the transportation trajectory of chemical products in real time, thereby obtaining accurate transportation distance data. Suppose a water plant uses disinfectants provided by a specific supplier. By connecting to the supplier's logistics information system, the water plant can obtain real-time information on the disinfectant's transportation distance. This step has the beneficial effect of providing accurate data support for quantifying carbon emissions from the use of disinfectants. In carbon emission accounting in the water supply industry, chemical use is a significant source of carbon emissions. Accurate chemical dosage data and transportation distance data help accurately calculate the carbon emissions of chemicals during production, transportation, and use. This data can reveal the relationship between chemical usage and carbon emissions for different chemicals, different treatment stages, and over time. For example, if a water plant experiences a significant increase in the dosage of a certain chemical during a certain period, combined with transportation distance data, the impact of this change on carbon emissions can be further analyzed. This provides a basis for optimizing chemical usage strategies and reducing carbon emissions, solving the problem of inaccurate data acquisition during chemical usage in carbon emission accounting in the water supply industry.

[0135] In step S2220, the data acquired in step S2210 is integrated to construct a three-dimensional correlation matrix. The three dimensions of the matrix are agent type, treatment step, and time series. This three-dimensional structure comprehensively displays the usage of agents in the water supply system. For example, at a certain moment, the dosage of a coagulant in the coagulation and sedimentation treatment stage of a water plant is X kg, the concentration is Y%, and the transportation distance is Z km. These data constitute an element in the matrix. By filling the matrix with relevant data for different agents, different treatment steps, and different time points, a complete three-dimensional correlation matrix can be formed. The purpose of this step is to present the scattered data in a structured manner, facilitating subsequent systematic analysis of agent usage. Its benefits are reflected in multiple aspects. First, this structured data organization clearly displays the differences in usage of different agents at different treatment steps and time periods. By analyzing the matrix, it is possible to intuitively identify which agents are used most in which steps and time periods, facilitating targeted optimization of agent usage. Second, it provides a well-organized data foundation for subsequent quantification of carbon emissions. When converting this data into carbon emission equivalents, the three-dimensional correlation matrix structure helps accurately calculate carbon emissions under different circumstances, improving the accuracy of carbon emission accounting. For example, by analyzing the matrix, it can be found that the use of a certain disinfectant in the disinfection process increases during the high temperatures of summer. Combined with information such as transportation distance, it can more accurately assess the changes in carbon emissions caused by the use of disinfectants during this period. This solves the problem of disorganized data and inconvenient analysis in carbon emission accounting in the water supply industry.

[0136] Step S2230 converts each element in the three-dimensional correlation matrix into a corresponding carbon emission equivalent, thereby generating a carbon emission chain correlation matrix Γ c . The quantification process is calculated based on factors such as the carbon emission factors of different agents and transportation distances. For carbon emissions in the production process of agents, the carbon emission factors are determined according to the production processes of different agents. For example, during the production process of a certain flocculant, each kilogram produced will produce M'' kilograms of carbon dioxide carbon emissions, which serves as the carbon emission factor for agent production. For carbon emissions in the transportation process, they are calculated based on the carbon emission coefficient of the transportation distance and mode of transportation. Assuming that road transportation is used, the transportation of each kilogram of agent will produce N'' kilograms of carbon dioxide carbon emissions per kilogram, which serves as the transportation carbon emission coefficient. If an element represents the addition of A'' kilograms of a certain agent with a concentration of B% at a certain time in a certain treatment link, and the transportation distance is C'' kilometers, then the carbon emission equivalent corresponding to the element is: the carbon emissions from agent production (A''×M'') plus the carbon emissions from the transportation process (A''×N''×C''). By calculating all elements in the three-dimensional correlation matrix in this way, the carbon emission chain correlation matrix Γ can be obtained. c Carbon emission chain correlation matrix Γc It reflects the correlation of carbon emissions of various agents in different links and at different times. Through this matrix, we can clearly see the carbon emission contributions of different agents in various links of the water supply system and their mutual influence. For example, by analyzing the matrix, we can find that in the treatment process of a certain water plant, the use of coagulants not only produces carbon emissions in the coagulation link, but also has an association with the carbon emissions of other agents due to its transportation process and the impact on subsequent treatment links. This association may be manifested in that the change in the dosage of coagulants affects the amount of agents used in the subsequent disinfection link, thereby indirectly affecting the carbon emissions of the disinfection link. On the one hand, this step realizes the conversion from agent usage data to carbon emission data, closely links agent use with carbon emissions, and makes the carbon emission accounting of the water supply industry more comprehensive and accurate. By quantifying the carbon emission equivalent of each element, the specific contribution of agent use to carbon emissions can be accurately assessed, providing an accurate basis for formulating emission reduction measures. On the other hand, the carbon emission chain correlation matrix Γ c The networked nature of carbon emissions transmission is revealed, helping to identify chain reactions in carbon emissions within water supply systems. For example, if the amount of chemical used in one process changes, the matrix can be used to analyze the impact of this change on carbon emissions in other processes, allowing proactive optimization measures to achieve carbon reduction. This addresses the issue of insufficient quantification and correlation analysis of carbon emissions in the chemical use process in carbon emissions accounting within the water supply industry.

[0137] Step S2300: Obtain methane emission rate data r of anaerobic fermentation of sludge in the sludge treatment unit of the sludge treatment process. d (t), methane emission rate data r during the landfill process l (t), the key parameters of sludge treatment in the sludge treatment unit and the key parameters of the landfill, and calculate the carbon emission eddy current coupling coefficient Ω of the sludge treatment and landfill process s ;

[0138] Step S2300 aims to obtain relevant data of sludge treatment and landfill process through a series of data collection, calculation and analysis methods, and calculate the carbon emission eddy current coupling coefficient Ω s , in order to quantitatively characterize the carbon emission correlation characteristics of the two processes, and provide a key basis for the full-process accounting and control of sludge carbon emissions.

[0139] Furthermore, step S2300 includes:

[0140] Step S2310: Install online methane monitoring equipment in the sludge treatment unit and landfill to collect real-time methane emission rate data from anaerobic fermentation of sludge. d (t) and the methane emission rate data r during landfilling l (t);

[0141] Specifically, methane emission rate refers to the volume or mass of methane gas emitted per unit time from a sludge treatment unit or landfill. It is directly measurable raw data. In practice, to achieve real-time data collection, specialized online methane monitoring equipment must be installed at both the sludge treatment unit and the landfill. This equipment continuously monitors methane emissions and transmits the data in real time to a data collection system. For example, a high-precision methane sensor is installed in the anaerobic sludge fermentation tank at a sewage treatment plant. This sensor accurately measures the volume of methane gas emitted from the fermentation tank per hour. Similar equipment is also installed at the corresponding landfill to monitor methane emissions during the landfilling process. Methane emissions from sludge treatment and landfilling processes are significant sources of carbon emissions in the water supply industry. Accurately acquiring real-time emission rate data allows for timely monitoring of carbon emissions from these two processes. By analyzing this data, patterns in methane emissions can be identified. For example, during the initial stages of anaerobic sludge fermentation, methane emission rates may increase as microbial activity increases, then gradually stabilize or decline. Methane emissions from landfills exhibit specific trends influenced by factors such as landfill duration and landfill volume. The discovery of these patterns helps us gain a deeper understanding of the mechanisms of carbon emissions, providing a strong basis for developing targeted emission reduction measures and resolving the challenge of obtaining real-time data on sludge treatment and landfilling processes in carbon emission accounting within the water supply industry.

[0142] Step S2320, obtaining key sludge treatment parameters of the sludge treatment unit and key landfill parameters of the landfill;

[0143] Specifically, key parameters for sludge treatment include the feed rate, residence time, and operating temperature of the sludge treatment unit. The feed rate directly affects the total amount of materials involved in the reaction; a higher feed rate may result in increased methane production. Residence time influences the extent of microbial decomposition of organic matter. A longer residence time results in more complete organic matter decomposition, potentially altering methane production. Operating temperature significantly influences microbial activity; an appropriate temperature promotes microbial growth and metabolism, thereby affecting methane production. Key landfill parameters include landfill volume, depth, and cover type. A larger landfill volume represents a greater amount of potentially degradable organic matter and, consequently, a greater amount of methane production. Landfill depth influences environmental factors such as oxygen content and temperature within the landfill, thus affecting the rate and total amount of methane production. The cover type influences the path and rate of methane release. This step provides essential parameter information for accurately calculating and analyzing carbon emissions. These key parameters play a crucial role in calculating methane emission fluxes and studying the patterns of methane emissions from landfills. For example, in calculating methane emission fluxes, the feed rate is a key component of the calculation. By accurately obtaining these parameters, the carbon emissions of the sludge treatment and landfill processes can be more accurately assessed, providing data support for the subsequent formulation of reasonable emission reduction strategies, and solving the problem of obtaining key parameters affecting carbon emissions in carbon emission accounting in the water supply industry.

[0144] Step S2330, based on the methane emission rate data r of the sludge treatment unit d (t) and key parameters of sludge treatment, calculate the methane emission flux F of the sludge treatment unit d (t), obtain emission flux time series data;

[0145] Specifically, this step converts the methane emission rate data into emission flux data based on the material conservation principle and the empirical formula for methane yield. The methane emission flux represents the ratio of the volume or mass of methane gas emitted by the sludge treatment unit per unit time to the amount of feed sludge. Its calculation formula is: ,in is the feed rate to the sludge treatment unit. The empirical formula for methane yield estimates the methane yield of sludge based on sludge properties (such as organic matter content and biodegradability) and operating conditions (such as temperature and residence time), and then calculates the methane emission flux. This step converts methane emission rate data into emission flux data, facilitating comparative analysis with landfill emission flux data. Because emission flux takes feed rate into account, it more accurately reflects the methane emissions per unit feed volume, enabling comparability of carbon emissions between different sludge treatment units and between landfills. This provides a more scientific metric for assessing the carbon emission efficiency of sludge treatment units. By analyzing emission flux time series data, we can intuitively understand the changes in the carbon emission efficiency of sludge treatment units over time. A sudden increase in emission flux over a period of time may indicate an abnormality in the sludge treatment process, such as increased microbial activity or changes in feed composition. Further analysis and appropriate measures are needed. This helps optimize the sludge treatment process, improve energy efficiency, reduce carbon emissions, and solve the problem of carbon emission efficiency assessment of sludge treatment units in carbon emission accounting in the water supply industry.

[0146] Step S2340, based on the methane emission rate data r of the landfill l (t) and key landfill parameters, fitting landfill methane emission flux The time series model of the landfill emission flux time series data was obtained;

[0147] Specifically, the methane emissions from landfills have a certain hysteresis and attenuation law. This step uses the methane generation attenuation model to fit the time series model of the landfill methane emission flux. The commonly used first-order exponential decay model is ,in is the initial emission flux, k is the decay coefficient, and t is the landfill time. l (t) and key parameters such as landfill volume and landfill time, and the model parameters are calibrated by curve fitting method and For example, collect methane emission rate data of a landfill at different landfill times over a period of time, combine it with parameters such as landfill volume, use professional data analysis software or mathematical methods to perform curve fitting, and determine the initial emission flux. and attenuation coefficient The landfill emission flux is obtained by A time series model. This model can be used to predict and extrapolate the long-term methane emissions of landfills. By analyzing the model, we can understand the changing trend of methane emissions from landfills over time and plan landfill management and emission reduction measures in advance. Its beneficial effect is that it provides a powerful tool for carbon emission management of landfills. By predicting the future methane emissions of landfills, landfill managers can arrange corresponding emission reduction equipment and measures in advance, such as installing methane collection and utilization devices to convert methane into energy and reduce its direct emissions into the atmosphere. At the same time, this also helps to evaluate the carbon emissions of landfills throughout their life cycle, provide data support for urban carbon emission planning and management, and solve the problem of long-term carbon emission prediction of landfills in carbon emission accounting in the water supply industry.

[0148] Step S2350: calculate the methane emission flux F of the sludge treatment unit. d (t) and landfill methane emission flux Perform cross-correlation analysis and calculate F d (t) and The mutual correlation coefficient ρ(τ) and the time lag constant τ0 are used to obtain the carbon emission eddy current coupling coefficient Ω s .

[0149] Specifically, cross-correlation analysis is a method used to study the correlation between two time series data. First, it is necessary to obtain the methane emission flux F of the sludge treatment unit. d (t) and landfill methane emission flux For example, within a period of time, record F at a certain time interval (such as every hour, every day, etc.) d (t) and In this embodiment, based on the cross-correlation function in statistics, F is calculated. d (t) and The mutual correlation coefficient ρ(τ) and the time lag constant τ0, the mutual correlation coefficient represents the landfill emission flux Delayed sludge treatment unit discharge flux time The time lag constant τ0 is the time difference when ρ(τ) reaches its maximum value, which reflects the time lag of sludge treatment discharge on landfill discharge. After calculating a series of mutual correlation coefficients ρ(τ), by observing the curve of ρ(τ) changing with τ, the τ value that makes ρ(τ) reach its maximum value is found. This value is the time lag constant τ0. For example, if ρ(τ) reaches its maximum value when τ=3 days, then the time lag constant τ0=3 days. This means that the landfill discharge flux F l (t) Delayed sludge treatment unit discharge flux F d(t) At 3 days, the correlation between the two was the strongest.

[0150] Carbon emission eddy current coupling coefficient The definition of ,in The reference starting time of the sludge treatment and landfill process (i.e. the t0th day from the start of the process, t0≥0). It is obtained in the previous cross-correlation analysis that when the lag time is τ0, the methane emission flux F of the sludge treatment unit is d (t) and landfill methane emission flux F l The cross-correlation coefficient between the two emission fluxes is τ0, which reflects the strength of the correlation between the two emission fluxes when the time lag is τ0. It represents the methane emission flux of the landfill at the time after the reference starting time t0 plus the time lag constant τ0. This value is used to measure the methane emission flux F of the sludge treatment unit. d (t) The most relevant lag time, the methane emission flux from the landfill. For example, if t0 represents the 10th day after the start of the sludge treatment and landfill process, τ0 = 3 days, then F l (t0+τ0) is the methane emission flux of the landfill on the 13th day. It is used in the formula to reflect the actual emission level of the landfill emission flux after a specific lag time, and is related to F d (t0) are calculated together to reflect the influence of the proportional relationship between the two on the carbon emission eddy current coupling coefficient. It represents the methane emission flux of the sludge treatment unit at the reference starting time t0. It is the emission flux of the sludge treatment unit at a specific starting time and is used as a reference value to participate in the calculation of the carbon emission eddy current coupling coefficient. For example, in the above example, F d (t0) is the methane emission flux of the sludge treatment unit on the 10th day. l By comparing and calculating (t0+τ0), the correlation strength and transmission law between the sludge treatment process and landfill emissions can be quantified.

[0151] Ω s The correlation and proportional relationship between sludge treatment emissions and landfill emissions are combined. The larger the value, the stronger the impact of the sludge treatment process on landfill carbon emissions, and the more obvious the transfer amplification effect. By introducing the carbon emission eddy current coupling coefficient Ω s , we can quantitatively characterize the correlation strength and transmission law between sludge treatment and landfill carbon emissions. This helps to reveal the amplification effect of sludge carbon emissions on the entire carbon emission chain and provide a basis for optimizing sludge treatment processes and landfill management. For example, if Ω sThe larger the value, the more significant the impact of carbon emissions from sludge treatment on landfills. Priority should be given to optimizing the sludge treatment process to reduce the amount of organic matter entering the landfill, thereby reducing carbon emissions from the landfill and achieving carbon reduction and efficiency improvement. This solves the problem of quantitative analysis of the correlation between sludge treatment and landfill carbon emissions in carbon emission accounting for the water supply industry.

[0152] Step S2400: Setting the fusion weight vector of the scene cluster ,in is the fusion weight of the Nth scenario cluster; the fusion weight vector based on the scenario cluster , integrating the power entropy pulse signal P e , carbon emission chain correlation matrix Γ c and carbon emission eddy current coupling coefficient Ω s , generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation .

[0153] The setting of the fusion weight vector W of the scenario cluster is not arbitrary, but requires comprehensive consideration of multiple factors. For example, the carbon accounting gradient of each scenario cluster can be , the number of water supply units covered by the scenario cluster, and the degree of contribution of each scenario cluster to the overall carbon emissions. If the carbon accounting gradient of a scenario cluster is higher, it means that the carbon emission intensity of the water supply units in the scenario cluster is higher and has a greater impact on the overall carbon emissions. In this case, when setting the weight, the corresponding weight value can be appropriately increased. According to the characteristics of different scenario clusters, the weights of each scenario cluster in the overall accounting are reasonably allocated so that the accounting results can more accurately reflect the actual carbon emissions. Through precise weight setting, the impact of high-carbon emission scenario clusters can be highlighted, and important carbon emission sources can be avoided from being ignored in the accounting process. This will help solve the problem of unreasonable weight distribution of different scenario clusters in carbon emission accounting in the water supply industry and improve the accuracy and reliability of accounting.

[0154] Dynamic carbon accounting density function The mathematical form is:

[0155]

[0156] in, It represents the dynamic carbon accounting density function value at position x, that is, the comprehensive carbon emission intensity index of that position.

[0157] This represents the spatial location of the water supply system. In practical applications, its specific value can be determined through a geographic information system (GIS) or the distribution of monitoring points within the water supply system. Its significance lies in clearly defining the specific location for calculating carbon emission density, thereby accurately characterizing carbon emissions at different spatial locations within the water supply system.

[0158] Indicates the The fusion weight of the scenario cluster reflects the relative importance of the i-th scenario cluster in the calculation of the entire dynamic carbon accounting density function. The larger the weight, the greater the impact of the scenario cluster on the final result.

[0159] Indicates the number of scenario clusters.

[0160] It represents the local mean of the power entropy pulse signal at position x, reflecting the dynamic change characteristics of the power carbon emission intensity at that location.

[0161] It represents the local mean of the carbon emission chain correlation matrix at position x, reflecting the correlation characteristics of carbon emissions in the links such as agent addition at that position.

[0162] It represents the local mean of the eddy current coupling coefficient of sludge carbon emission at position x, reflecting the coupling characteristics of carbon emissions from sludge treatment and landfill processes at this location.

[0163] Represents the trace operation of the matrix, that is, the sum of the main diagonal elements of the matrix, which is used to link the carbon emission chain matrix Converted to a scalar indicator.

[0164] Indicates taking the maximum value of each position, which is used for the power entropy pulse signal , carbon emission chain correlation matrix trace, eddy current coupling coefficient of sludge carbon emission Normalization is performed to make the dimensions consistent to facilitate weighted fusion.

[0165] when When increases, if other terms remain unchanged, the contribution of the corresponding scenario cluster to the function value increases. will increase accordingly; When any value in increases and its ratio to the corresponding maximum value increases, Increases, and vice versa. This shows that when the intensity or correlation of each carbon source at the corresponding location increases, the dynamic carbon accounting density function value of that location will increase, reflecting that the carbon emissions are more significant. This formula integrates data from multiple heterogeneous carbon sources such as electricity, pesticide consumption, and sludge to comprehensively reflect the carbon emissions of the water supply system at different spatial locations, realizing integrated accounting at multiple scales in time and space. By comprehensively considering the contribution of each scenario cluster and the characteristics of different carbon sources, the dynamic adaptability and robustness of the accounting paradigm are improved, which helps to accurately assess the carbon emissions of the water supply system and provide a scientific basis for the formulation of targeted emission reduction strategies.

[0166] Heterogeneous entropy weight bias The calculation method is:

[0167] Fusion weight vector according to scenario cluster , calculate the arithmetic mean of the fusion weights of each scenario cluster ,according to and , calculate the heterogeneous entropy weight deviation .

[0168]

[0169] in, represents the arithmetic mean of the fusion weights of each scenario cluster, It reflects the heterogeneous entropy weight differences of each scenario cluster. The larger the value, the worse the adaptability of the fusion algorithm to different scenarios, and further optimization is needed.

[0170] When the fusion weight of each scenario cluster When the difference is large, The value of will increase, the numerator Increase in the denominator When the change is relatively small, will increase, indicating that the adaptability of the fusion algorithm to different scenarios is worse; on the contrary, when the fusion weights of each scenario cluster are closer, The smaller the value, the better the adaptability of the fusion algorithm. This formula is used to evaluate the rationality of the scenario cluster fusion weight vector W, reflecting the differences in the heterogeneous entropy weights of each scenario cluster. By calculating the heterogeneous entropy weight deviation, we can determine the adaptability of the current fusion algorithm to different scenarios. The smaller the deviation, the more reasonable the fusion weight distribution, the better the adaptability of the fusion algorithm to different scenarios, and the more reliable the calculation results.

[0171] Step S3000: Calculate the density function based on the dynamic carbon accounting and heterogeneous entropy weight bias , build a carbon accounting gradient multi-objective optimization model; solve the carbon accounting gradient multi-objective optimization model, and iteratively optimize the carbon accounting gradient of each scenario cluster , and obtain the optimal carbon accounting gradient , according to the optimal carbon accounting gradient , output multi-scenario carbon emission accounting optimization results.

[0172] Furthermore, step S3000 includes:

[0173] Step S3100: Calculate the density function based on the dynamic carbon accounting and heterogeneous entropy weight bias , construct a carbon accounting gradient multi-objective optimization model;

[0174] Furthermore, step S3100 includes:

[0175] Step S3110: Dynamic carbon accounting density function Integrate and calculate total carbon emissions ;

[0176] Specifically, the dynamic carbon accounting density function The carbon data from multiple sources in the water supply system, such as electricity, chemicals, and sludge, are integrated to reflect the carbon emissions of the water supply system at a spatial location x. x represents a specific spatial location in the water supply system, covering the areas where all links, from raw water intake, water treatment plants to water transmission and distribution, are located. The function is integrated over the spatial range X of the water supply system, and the formula is: The integration operation here is a mathematical method that converts the carbon emission density at different spatial locations x into Accumulate and get the overall carbon emission level of the entire water supply system .

[0177] For example, if the water supply system is divided into multiple small blocks, each small block has a corresponding The carbon emissions of these small blocks can be aggregated by integral calculation to obtain the total carbon emissions of the entire water supply system. The purpose of this step is to quantify the total carbon emissions of the entire water supply system and provide an overall indicator for subsequent evaluation and analysis. Its beneficial effect is that in the absence of a unified and accurate accounting method, the scale of carbon emissions of the water supply system can be comprehensively and accurately measured. This is crucial for urban low-carbon development planning. For example, city managers can use By analyzing the water supply industry's contribution to a city's total carbon emissions, the data provides a clear understanding of the water supply industry's contribution to the city's total carbon emissions, enabling the formulation of more reasonable carbon emission targets and reduction strategies. This data also provides a clear quantitative basis for the water supply industry's own energy conservation and emission reduction efforts, helping to assess changes in carbon emissions over time and across different water supply systems. This allows for targeted process improvements and management optimization, resolving the issue of inaccurate overall carbon emission quantification in the water supply industry's carbon emission accounting.

[0178] Step S3120: Based on the carbon accounting gradient of each scenario cluster and the water supply volume of the water supply unit corresponding to each scenario cluster , calculate the theoretical total carbon emissions of the water supply system ;

[0179] Specifically, the carbon accounting gradient of each scenario cluster It is determined in the early stage when the scenario clusters are divided according to factors such as process entropy value, water source coupling factor, load fluctuation coefficient, etc. It reflects the difference in carbon emissions per unit water supply under different scenario clusters. Represents the carbon accounting gradient of the i-th scenario cluster, where i represents the different scenario cluster numbers, from 1 to N, N is the number of scenario clusters, and N ≥ 3. The water supply volume of each scenario cluster corresponding to the water supply unit It is obtained by monitoring and statistics of the actual water supply conditions of each unit in the water supply system. The formula for calculating the total theoretical carbon emissions is: ,in, Represents the water supply of the water supply unit corresponding to the i-th scenario cluster; the meaning of this formula is to multiply the carbon accounting gradient of each scenario cluster by the water supply of the corresponding water supply unit, and then add the calculation results of all scenario clusters to obtain the theoretical carbon emission estimate based on scenario classification . The purpose of this step is to estimate the total carbon emissions of the water supply system based on the scenario grading method, taking into account the impact of differences in water supply units under different scenarios on carbon emissions. Its beneficial effect is that it can more carefully analyze the contribution of different scenarios to overall carbon emissions. Since the carbon accounting gradients of different scenario clusters are different, reflecting the carbon emission characteristics under different process, water source, load and other conditions, combined with the calculation of the theoretical total carbon emissions based on the water supply, it can be determined which scenario clusters are the main sources of carbon emissions. For example, if it is found through calculation that the carbon accounting gradient of a scenario cluster is higher and the water supply is larger, then this scenario cluster is the focus of carbon emissions, and special emission reduction measures can be formulated according to its characteristics. This helps to improve the refinement of carbon emission accounting and solves the problem that scenario differences cannot be considered in carbon emission accounting in the water supply industry.

[0180] Step S3130, according to and , defining cross-modal validation bias ;

[0181] Specifically, the cross-modal verification bias ε is calculated by the formula To define, Characterizes the degree of deviation between the two carbon accounting modes of dynamic accounting and scenario classification. The smaller ε is, the more consistent the two modes are, and the higher the reliability of the accounting system is. The purpose of this step is to measure the difference between the two different accounting methods, so as to evaluate the reliability of the accounting system. Its beneficial effect is that it provides a quantitative indicator for judging the accuracy of the accounting system. The smaller ε is, the closer the actual accounting results based on the dynamic carbon accounting density function are to the theoretical estimation results based on the scenario grading, which means that the consistency of the accounting system under different accounting methods is higher and the accounting results are more reliable. For example, if ε remains at a low level throughout multiple accounting, it means that the accounting system can more accurately reflect the carbon emissions of the water supply system; on the contrary, if ε is large, it indicates that there are large differences between the two accounting modes, and there may be problems such as unreasonable scenario division and inaccurate parameter calculation. The accounting system needs to be optimized, which will help improve the credibility of carbon emission accounting in the water supply industry and solve the problem of lack of quantitative standards for reliability assessment of the accounting system.

[0182] Step S3140: Combine cross-modality verification deviation and heterogeneous entropy weight bias , construct a carbon accounting gradient multi-objective optimization model.

[0183]

[0184] Where J is the carbon accounting gradient multi-objective optimization function, which is the objective to be minimized; To check for bias across modalities The weight coefficient of is the heterogeneous entropy weight deviation The weight coefficient of and Used to adjust the relative importance of the two deviations in the optimization objective; is the total water supply of the water supply system, and the constraint condition Ensure that the sum of water supply of each scenario cluster is equal to the total water supply. This is based on the principle of conservation of matter and ensures the rationality of water supply in the calculation process. Ensure that the carbon accounting gradient is non-negative, because the carbon accounting gradient represents the carbon emissions per unit water supply and cannot be negative in practical terms. The values ​​of the weight coefficients α and β will affect the focus of optimization. For example, if in practical applications, more attention is paid to the consistency of dynamic accounting and scenario-level accounting results, then the value of α can be appropriately increased so that the optimization process focuses more on reducing the cross-modal verification deviation ε; if more attention is paid to the adaptability of the fusion algorithm to different scenarios, that is, the heterogeneous entropy weight deviation is expected to be If α and β are smaller, the value of β can be increased. When determining α and β in practice, multiple experiments, expert experience, or machine learning algorithms can be used to find the value combination that best suits a specific water supply system.

[0185] The purpose of this model is to find the optimal carbon accounting gradient by minimizing the cross-modal verification bias and heterogeneous entropy weight bias, so as to balance the consistency and difference between dynamic accounting and scenario-based verification and improve the overall performance of the accounting system. Its beneficial effect is that it comprehensively considers the two important deviations in the accounting process, making the optimized carbon accounting gradient more consistent with the actual situation. For example, in an actual water supply system, after multiple iterative optimizations, when J reaches a small value, the optimal carbon accounting gradient obtained can make the dynamic accounting and scenario-based accounting results closer, and at the same time, the fusion algorithm is more adaptable to different scenarios, thereby achieving more accurate carbon emission accounting. This helps to solve the problems of unreasonable carbon accounting gradients and the need to improve the performance of the accounting system in carbon emission accounting in the water supply industry, and provides more reliable technical support for achieving low-carbon development in the water supply industry.

[0186] Step S3200: solve the carbon accounting gradient multi-objective optimization model and iteratively optimize the carbon accounting gradient of each scenario cluster. ;

[0187] Furthermore, step S3200 includes:

[0188] Step S3210: Initialize the carbon accounting gradient for each scenario cluster , let the number of iterations variable , set the maximum number of iterations and convergence threshold ;

[0189] Specifically, the number of iterations variable Initialized to 0, which means the optimization process starts counting from the initial state. Maximum number of iterations is a pre-set value that limits the number of iterations. This is because in actual calculations, the iterative process may not continue indefinitely due to various reasons (such as computing resource limitations or algorithm convergence characteristics). Setting this value ensures that the algorithm terminates within a reasonable time and resource range. The convergence threshold δ is the criterion for determining whether the iteration has converged. When the change in the cross-modal calibration deviation ε between two consecutive iterations is less than the convergence threshold δ, the iterative process is considered to have converged, meaning that a relatively stable result that meets the accuracy requirements has been achieved. This step aims to prepare the iterative optimization process by determining the starting conditions and termination rules. Its beneficial effect is that it provides a clear operational framework for the entire optimization algorithm. By properly setting the initial values ​​and constraints, the operating range and accuracy of the optimization process can be controlled, preventing the algorithm from entering an infinite loop or producing results that do not meet practical requirements. In carbon emission accounting in the water supply industry, this helps improve computational efficiency and result reliability, addressing the problems of lack of effective control in the optimization process and unclear starting conditions, and laying the foundation for subsequent accurate optimization of carbon accounting gradients.

[0190] Step S3220: fix the carbon accounting gradients of other scenario clusters and Carbon accounting gradients for each scenario cluster Optimize

[0191] make:

[0192]

[0193] In this formula, represents the carbon accounting gradient value of the i-th scenario cluster at the s-th iteration, which is the carbon accounting gradient state of the scenario cluster in the current iteration;

[0194] It represents the carbon accounting gradient value of the i-th scenario cluster at the s+1th iteration, which is the result after this optimization update.

[0195] The learning rate is a parameter that controls the optimization step size and determines the magnitude of the carbon accounting gradient adjustment at each iteration. If η is too large, the optimization process may jump back and forth around the optimal solution and fail to converge. If it is too small, the optimization process will become very slow, increasing the calculation time.

[0196] Gradient multi-objective optimization function for carbon accounting right The partial derivative of reflects the change in the current state. The degree of influence on the objective function J. By calculating the partial derivative, we can determine in which direction to adjust It can make the objective function J decrease faster.

[0197] By continuously adjusting the carbon accounting gradient of each scenario cluster, the objective function J is gradually reduced, thus approaching the optimal solution. Its beneficial effect is that it provides a specific calculation method for optimizing the carbon accounting gradient. In the carbon emission accounting of the water supply industry, this method can gradually improve the carbon accounting gradient, making dynamic accounting more consistent with scenario grading verification, and improving the accuracy of the accounting system. Each iteration is adjusted based on the current carbon accounting gradient and the rate of change of the objective function. It can fully consider the relationship between each scenario cluster and its impact on the overall accounting results, and solve the problem of how to optimize the carbon accounting gradient to improve the performance of the accounting system.

[0198] Step S3230: cyclically execute step S3220 until the carbon accounting gradients of all scenario clusters are optimized. ;

[0199] Specifically, by repeatedly executing step S3220, the carbon accounting gradient of each scenario cluster is optimized sequentially. In each iteration, the carbon accounting gradients of the other scenario clusters are fixed, and only the carbon accounting gradient of the currently selected scenario cluster is updated. This process continues until all scenario clusters have completed optimization. The iteration count variable, s, is then incremented by 1. This process continues until the termination condition in step S3240 is met. For example, for a water supply system with five scenario clusters, in the first iteration, the carbon accounting gradients of scenario clusters 1, 3, 4, and 5 are fixed, and the carbon accounting gradient of scenario cluster 2 is optimized. The other four scenario clusters are then fixed, and the carbon accounting gradient of scenario cluster 1 is optimized. This process continues, and so on. After all five scenario clusters have been optimized, s is updated from 0 to 1. A second iteration then occurs, repeating the above process to continuously adjust the carbon accounting gradients of each scenario cluster. The goal of this step is to gradually optimize the carbon accounting gradients of all scenario clusters through multiple iterations to achieve an overall optimal result. This beneficial effect is that it ensures the comprehensiveness and systematic nature of the optimization process. In carbon emissions accounting for the water supply industry, due to the interrelationships and influences between scenario clusters, optimizing a single scenario cluster alone may not achieve overall optimization. By iteratively considering the conditions of each scenario cluster, the carbon accounting gradient can be continuously improved, making the accounting system more consistent with actual carbon emissions, improving the accuracy and reliability of accounting, and addressing the issue of comprehensive optimization of carbon accounting gradients under multiple scenarios.

[0200] Step S3240: Calculate the optimized cross-modal verification deviation ,like or , then execute step S3250; otherwise, execute step S3220; wherein, represents the cross-modal calibration deviation after the sth iteration optimization, represents the cross-modal verification deviation after the s+1th iterative optimization;

[0201] Specifically, first The optimized carbon accounting gradient of each scenario cluster obtained after the iteration Substitute into the formula of step S3120 to calculate the optimized theoretical total carbon emissions ,Will Substitute the cross-modal verification deviation in step S3130 The calculation formula for Cross-modal calibration deviation after iterative optimization , then Compared with the previous iteration Compare and calculate the absolute value of their difference If this difference is less than the convergence threshold δ, it means that the iterative process has converged, and the optimized carbon accounting gradient has made the dynamic accounting and scenario-level accounting results close enough, achieving the expected accuracy requirements; or when the number of iterations s is greater than the maximum number of iterations Regardless of whether it converges or not, the iterative process is stopped and step S3250 is executed. If the above two conditions are not met, return to step S3220 to continue iterative optimization. The purpose of this step is to determine whether the iterative optimization process has achieved the expected results and decide whether to continue iterating. Its beneficial effect is to ensure that the optimization process ends when a certain accuracy requirement is met or within a reasonable number of calculations. In the carbon emission accounting of the water supply industry, unnecessary waste of computing resources is avoided, and at the same time, it is ensured that the final carbon accounting gradient can make the accounting system have higher reliability. By monitoring the cross-modal verification deviation and comparing it with the convergence threshold, the quality of the optimization results can be effectively controlled, and the problem of how to determine the termination conditions of the optimization process and ensure the accuracy of the optimization results is solved.

[0202] Step S3250: Output the optimal carbon accounting gradient ,in, represents the optimal carbon accounting gradient for the i-th scenario cluster.

[0203] Specifically, after the previous iterative optimization process, when the termination condition in step S3240 is met, the carbon accounting gradient of each scenario cluster obtained at this time is The optimal carbon accounting gradient . These optimal carbon accounting gradients are the results obtained through iterative optimization based on comprehensive consideration of factors such as dynamic carbon accounting density function, heterogeneous entropy weight deviation and cross-modal verification deviation. They can achieve a balance between consistency and difference between dynamic accounting and scenario grading verification, and are the key results of the optimization of the entire multi-scenario carbon emission accounting system. The purpose of this step is to output the optimized carbon accounting gradients and provide the final optimization results for multi-scenario carbon emission accounting. Its beneficial effect is to provide more accurate parameters for carbon emission accounting in the water supply industry. Based on these optimal carbon accounting gradients, carbon emissions under different scenarios can be more accurately assessed, providing strong support for the formulation of targeted emission reduction strategies. For example, when formulating emission reduction measures, we can focus on scenario clusters with higher carbon emission intensity based on the optimal carbon accounting gradients of different scenario clusters, and adopt more stringent emission reduction measures to achieve the energy conservation and emission reduction goals of the water supply industry, solving the problem that the carbon accounting gradients in the carbon emission accounting of the water supply industry are inaccurate and cannot meet actual needs.

[0204] Step S3300: Optimal carbon accounting gradient based on each scenario cluster , output multi-scenario carbon emission accounting optimization results.

[0205] Specifically, the optimal carbon accounting gradient is determined after comprehensively considering multiple factors such as the process topology characteristics of the water supply system, water source heterogeneity, spatiotemporal load fluctuations, as well as multiple indicators such as dynamic carbon accounting density function and heterogeneous entropy weight deviation. They can more accurately reflect the carbon emissions per unit water supply under different scenarios and are an accurate quantification of the carbon emission characteristics of the water supply system. Taking a water supply system containing three scenario clusters as an example, assuming that after iterative optimization, the optimal carbon accounting gradient of scenario cluster 1 is (Unit: tons of carbon dioxide / cubic meter of water supply), optimal carbon accounting gradient for scenario cluster 2 , optimal carbon accounting gradient for scenario cluster 3 These values ​​show that under Scenario Cluster 1, carbon emissions per cubic meter of water supply are approximately 0.4 tons; Scenario Cluster 2 has relatively low carbon emissions, with 0.3 tons of carbon dioxide produced per cubic meter of water supply; and Scenario Cluster 3 has a higher carbon emission intensity, with 0.5 tons of carbon dioxide produced per cubic meter of water supply.

[0206] The multi-scenario carbon emission accounting optimization results based on these optimal carbon accounting gradients contain more accurate carbon emission accounting data under different scenario clusters. This result can provide multiple support for carbon emission management in the water supply industry:

[0207] Accurately assess carbon emissions: By multiplying the optimal carbon accounting gradient for each scenario cluster by the water supply volume of the corresponding water supply unit, the carbon emissions of each scenario cluster can be accurately calculated, and then the total carbon emissions of the entire water supply system under different scenarios can be summarized. This allows water supply companies and relevant management departments to clearly understand the carbon emissions under different operating conditions. For example, scenario clusters with high carbon accounting gradients and large water supply volumes can be identified as key areas for carbon emission control, and resources can be concentrated on the development and implementation of energy-saving and emission reduction measures.

[0208] Guiding the formulation of emission reduction strategies: Targeted emission reduction strategies can be formulated based on the carbon emission characteristics of different scenario clusters. For scenario clusters with higher carbon accounting gradients, in-depth analysis can be conducted on the reasons for their high emissions, such as whether it is due to high energy consumption caused by a specific process, or whether poor water quality at the water source increases carbon emissions during the treatment process. Then, corresponding measures can be taken to address these reasons, such as optimizing processes and improving water source treatment methods, to achieve precise emission reductions. For example, if it is found that the high carbon emissions of a scenario cluster are mainly due to the low energy efficiency of a certain treatment process, resources can be invested in upgrading and transforming the process to reduce its carbon emission intensity.

[0209] Supporting Low-Carbon Development Planning: This provides a scientific basis for low-carbon development planning in cities or regions. Based on the results of multi-scenario carbon emissions accounting optimization, city managers can rationally plan the layout and operation of water supply systems. For example, in areas with high carbon emission intensities, consideration can be given to adjusting the allocation of water supply units, reducing water supply to high-emission areas, or introducing cleaner water sources. This can reduce carbon emissions in the water supply sector as a whole and promote the city's transition to low-carbon development.

[0210] The multi-scenario carbon emissions accounting optimization results output in step S3300 address the current lack of a unified, accurate, and scenario-specific carbon emissions accounting method for the water supply industry. By providing accurate carbon emissions data and targeted emission reduction evidence, this helps meet the needs of low-carbon urban development and energy conservation and emission reduction in the water supply industry. Furthermore, this result provides a solid data foundation and decision-making support for further research on the factors affecting carbon emissions in water supply systems, optimizing accounting methods, and evaluating emission reduction effectiveness, thereby promoting the sustainable development of the water supply industry in the context of climate change.

[0211] Example 2

[0212] This embodiment provides a multi-scenario driven water supply carbon emission accounting system optimization system based on embodiment 1, such as Figure 8 Shown, including:

[0213] Scenario classification module: Based on the process topology characteristics, water source heterogeneity and spatiotemporal load fluctuations of the water supply system, a three-dimensional dynamic entropy weight classification model is constructed. Based on the three-dimensional dynamic entropy weight classification model, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. , where N≥3;

[0214] Deviation calculation module: real-time collection of heterogeneous carbon source pulse data of the entire water supply process, mapping the heterogeneous carbon source pulse data to the N scenario clusters divided by the accounting system, and generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation ;

[0215] Carbon accounting optimization module: based on dynamic carbon accounting density function and heterogeneous entropy weight bias , build a carbon accounting gradient multi-objective optimization model; solve the carbon accounting gradient multi-objective optimization model, and iteratively optimize the carbon accounting gradient of each scenario cluster , and obtain the optimal carbon accounting gradient , according to the optimal carbon accounting gradient , output multi-scenario carbon emission accounting optimization results.

[0216] In the scenario grading module, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. include:

[0217] Step S1510: using the process entropy value E in the three-dimensional dynamic entropy weight classification model p , water source coupling factor α s and load fluctuation coefficient β t For three dimensions, three-dimensional spatial mapping is performed on each water supply unit of the water supply system to form a sample point set;

[0218] Step S1520, performing fuzzy C-means clustering on the sample point set to obtain N scenario clusters;

[0219] Step S1530: Calculate the mean of the process entropy value of each scenario cluster , the mean of water source coupling factors and the mean of the load fluctuation coefficient ,Will 、 and According to the entropy weight vector [w p ,w s ,w t ] to perform weighted calculations to obtain the carbon accounting gradient for each scenario cluster =[ ],in, represents the carbon accounting gradient of the Nth scenario cluster.

[0220] In the deviation calculation module, the dynamic carbon accounting density function is generated include:

[0221] Step S2100: collect the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, calculate the real-time carbon emission intensity time series data based on the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, and obtain the power entropy pulse signal P e ;

[0222] Step S2200: Obtain the real-time drug addition data and drug transportation distance of the water supply system, and obtain the carbon emission chain correlation matrix Γ of the drug link based on the real-time drug addition data and drug transportation distance of the water supply system. c ;

[0223] Step S2300: Obtain methane emission rate data r of anaerobic fermentation of sludge in the sludge treatment unit of the sludge treatment process. d (t), methane emission rate data r during the landfill process l (t), the key parameters of sludge treatment in the sludge treatment unit and the key parameters of the landfill, and calculate the carbon emission eddy current coupling coefficient Ω of the sludge treatment and landfill process s ;

[0224] Step S2400: Setting the fusion weight vector of the scene cluster ,in is the fusion weight of the Nth scenario cluster; the fusion weight vector based on the scenario cluster , integrating the power entropy pulse signal P e , carbon emission chain correlation matrix Γ c and carbon emission eddy current coupling coefficient Ω s , generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation .

[0225] In the carbon accounting optimization module, the carbon accounting gradient multi-objective optimization model is solved to iteratively optimize the carbon accounting gradient of each scenario cluster. include:

[0226] Step S3210: Initialize the carbon accounting gradient for each scenario cluster , let the number of iterations variable , set the maximum number of iterations and convergence threshold ;

[0227] Step S3220: fix the carbon accounting gradients of other scenario clusters and Carbon accounting gradients for each scenario cluster Optimize;

[0228] Step S3230: cyclically execute step S3220 until the carbon accounting gradients of all scenario clusters are optimized. ;

[0229] Step S3240: Calculate the optimized cross-modal verification deviation ,like or , then execute step S3250; otherwise, execute step S3220; wherein, represents the cross-modal calibration deviation after the sth iteration optimization, represents the cross-modal verification deviation after the s+1th iterative optimization;

[0230] Step S3250: Output the optimal carbon accounting gradient ,in, represents the optimal carbon accounting gradient for the i-th scenario cluster.

[0231] The methods and systems of the present application may be implemented in many ways. For example, the methods and systems of the present application may be implemented using software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps used in the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified.

[0232] In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0233] The above-described specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is merely a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A multi-scenario driven water supply carbon emission accounting system optimization method, characterized by: The method comprises: According to the process topology characteristics, water source heterogeneity and spatiotemporal load fluctuations of the water supply system, a three-dimensional dynamic entropy weight classification model is constructed. According to the three-dimensional dynamic entropy weight classification model, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. , where N≥3; Real-time collection of heterogeneous carbon source pulse data throughout the water supply process, mapping the heterogeneous carbon source pulse data to the N scenario clusters divided by the accounting system, and generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation ; The heterogeneous carbon source pulse data is mapped to the N scenario clusters divided by the accounting system to generate a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation include: Based on the heterogeneous carbon source pulse data of the whole water supply process, the power entropy pulse signal P is obtained. e , carbon emission chain correlation matrix Γ of the pharmaceutical link c and the carbon emission eddy current coupling coefficient Ω during sludge treatment and landfill s ; Set the fusion weight vector of the scenario cluster ,in is the fusion weight of the Nth scenario cluster; the fusion weight vector based on the scenario cluster , integrating the power entropy pulse signal P e , carbon emission chain correlation matrix Γ c and carbon emission eddy current coupling coefficient Ω s , generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation ; Dynamic carbon accounting density function The mathematical form is: ； in, represents the dynamic carbon accounting density function value at position x, Indicates the spatial location of the water supply system, Indicates the The fusion weight of the scene clusters, represents the number of scenario clusters, represents the local mean of the power entropy pulse signal at position x, represents the local mean of the carbon emission chain correlation matrix at position x, represents the local mean of the eddy current coupling coefficient of sludge carbon emission at position x, represents the trace operation of the matrix, Indicates taking the maximum value of each position; Heterogeneous entropy weight bias The calculation method is: Fusion weight vector according to scenario cluster , calculate the arithmetic mean of the fusion weights of each scenario cluster ,according to and , calculate the heterogeneous entropy weight deviation ; ； in, represents the arithmetic mean of the fusion weights of each scenario cluster; According to the dynamic carbon accounting density function and heterogeneous entropy weight bias , build a carbon accounting gradient multi-objective optimization model; solve the carbon accounting gradient multi-objective optimization model, and iteratively optimize the carbon accounting gradient of each scenario cluster , and obtain the optimal carbon accounting gradient , according to the optimal carbon accounting gradient , output multi-scenario carbon emission accounting optimization results.

2. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 1 is characterized in that: The construction of the three-dimensional dynamic entropy weight classification model includes: Based on the process topology characteristics of the water supply system, the historical power consumption data and historical reagent dosage data of each treatment process are obtained, and the process entropy value E is obtained according to the historical power consumption data and historical reagent dosage data of each treatment process. p ; Based on the water source heterogeneity of the water supply system, the raw water BOD concentration and the length of the water transmission network of each water source are obtained, and the water source coupling factor α is generated according to the raw water BOD concentration and the length of the water transmission network of each water source. s ; Calculate the load fluctuation coefficient β based on the temporal and spatial load fluctuations of the water supply system t ; According to the process entropy value E p , water source coupling factor α s and load fluctuation coefficient β t , construct a three-dimensional dynamic entropy weight classification model.

3. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 2 is characterized in that: The process entropy value E is obtained p include: According to the historical power consumption data of each treatment process, the power consumption entropy increase rate of each treatment process is calculated, and the power consumption entropy increase rate of each treatment process is subtracted from the power consumption entropy increase rate of the benchmark process to obtain the relative entropy increase rate; The total dosage of each treatment process is obtained based on the historical dosage data of each treatment process, and the ratio of the total dosage of the reagent to the water inflow is calculated to obtain the reagent dissipation ratio; The process entropy value E is obtained by linearly combining the relative entropy increase rate and the reagent dissipation ratio. p .

4. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 2 is characterized in that: The generated water source coupling factor α s include: The BOD concentration of the raw water of each water source is normalized by min-max to obtain the BOD heterogeneity index; the length of the water supply network of each water source is normalized by min-max to obtain the distance heterogeneity index; the BOD heterogeneity index and the distance heterogeneity index are weighted and summed to obtain the water source coupling factor α s .

5. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 2 is characterized in that: The calculated load fluctuation coefficient β t include: Based on the spatiotemporal load fluctuations of the water supply system, historical monthly water production data and historical carbon emission intensity time series data are obtained; Based on the historical monthly water production data, the coefficient of variation of the monthly average water production is calculated to obtain the water production fluctuation index; Perform Fourier transform on the historical carbon emission intensity time series data, extract the dominant frequency, calculate the peak period, and obtain the carbon emission intensity periodicity index; The load fluctuation coefficient β is obtained by multiplying the water production fluctuation index and the carbon emission intensity periodicity index and taking the square root. t .

6. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 2 is characterized in that: The construction of the three-dimensional dynamic entropy weight classification model includes: Process entropy E p , water source coupling factor α s and load fluctuation coefficient β t Normalization is performed separately; Assign a weight coefficient w to the normalized process entropy value p , assign a weight coefficient w to the normalized water source coupling factor s , assign a weight coefficient w to the normalized load fluctuation coefficient t ;W p 、w s 、w t Constitute the entropy weight vector [w p ,w s ,w t ]; By process entropy value E p , water source coupling factor α s , load fluctuation coefficient β t and entropy weight vector [w p ,w s ,w t ], construct a three-dimensional dynamic entropy weight classification model.

7. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 6 is characterized in that: The carbon accounting gradient for each scenario cluster is calculated as follows include: Calculate the mean of the process entropy value for each scenario cluster , the mean of water source coupling factors and the mean of the load fluctuation coefficient ,Will 、 and According to the entropy weight vector [w p ,w s ,w t ] to perform weighted calculations to obtain the carbon accounting gradient for each scenario cluster =[ ],in, represents the carbon accounting gradient of the Nth scenario cluster.

8. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 1 is characterized in that: The heterogeneous carbon source pulse data of the whole water supply process includes the real-time power consumption data of each link of the water supply system and the real-time carbon emission factor of the regional power grid, the real-time agent addition data and agent transportation distance of the water supply system, and the methane emission rate data of the anaerobic fermentation of sludge in the sludge treatment unit of the sludge treatment link. d (t), methane emission rate data r during the landfill process l (t), key sludge treatment parameters of the sludge treatment unit and key landfill parameters of the landfill.

9. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 8 is characterized in that: The heterogeneous carbon source pulse data of the entire water supply process includes real-time power consumption data of each link of the water supply system, real-time carbon emission factors of the regional power grid, real-time chemical addition data of the water supply system and chemical transportation distance; The power entropy pulse signal P is obtained e include: Multiply the real-time electricity consumption data of each link in the water supply system with the real-time carbon emission factor of the regional power grid to obtain the real-time carbon emission intensity time series data of electricity consumption; Perform pulse detection on real-time carbon emission intensity time series data, extract power carbon emission pulse events, and construct power entropy pulse signal P e .

10. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 9 is characterized in that: The real-time drug addition data includes real-time drug addition amount and drug concentration; The carbon emission chain correlation matrix Γ of the pharmaceutical link c include: The real-time dosage, concentration and transportation distance of each agent are organized into a three-dimensional correlation matrix; Quantify each element in the three-dimensional correlation matrix into carbon emission equivalents to generate the carbon emission chain correlation matrix Γ c .

11. The multi-scenario driven water supply carbon emission accounting system optimization method according to claim 1 is characterized in that: The construction of the carbon accounting gradient multi-objective optimization model includes: Dynamic carbon accounting density function Integrate and calculate total carbon emissions ; According to the carbon accounting gradient of each scenario cluster and the water supply volume of the water supply unit corresponding to each scenario cluster , calculate the theoretical total carbon emissions of the water supply system ; according to and , defining cross-modal validation bias ; Combine cross-modal validation bias and heterogeneous entropy weight bias , construct a carbon accounting gradient multi-objective optimization model.

12. A multi-scenario driven water supply carbon emission accounting system optimization system, which is used to implement the multi-scenario driven water supply carbon emission accounting system optimization method according to any one of claims 1 to 11, characterized in that: The system comprises: Scenario classification module: Based on the process topology characteristics, water source heterogeneity and spatiotemporal load fluctuations of the water supply system, a three-dimensional dynamic entropy weight classification model is constructed. Based on the three-dimensional dynamic entropy weight classification model, the accounting system is divided into N scenario clusters, and the carbon accounting gradient of each scenario cluster is calculated. , where N≥3; Deviation calculation module: real-time collection of heterogeneous carbon source pulse data of the entire water supply process, mapping the heterogeneous carbon source pulse data to the N scenario clusters divided by the accounting system, and generating a dynamic carbon accounting density function , and calculate the heterogeneous entropy weight deviation ; Carbon accounting optimization module: based on dynamic carbon accounting density function and heterogeneous entropy weight bias , build a carbon accounting gradient multi-objective optimization model; solve the carbon accounting gradient multi-objective optimization model, and iteratively optimize the carbon accounting gradient of each scenario cluster , and obtain the optimal carbon accounting gradient , according to the optimal carbon accounting gradient , output multi-scenario carbon emission accounting optimization results.

Images