Urban carbon peak reaching judgment method and system based on power data and fuzzy logic

Abstract

The invention discloses an urban carbon peak reaching judgment method and system based on electric power data and fuzzy logic. The method comprises the following steps: firstly, screening an index which has the greatest influence on urban carbon emission intensity in the electric power data; then, based on the screened indexes, a dynamic weight model and an extended unhooking model are constructed respectively to judge the urban carbon peak reaching condition, and carbon peak reaching judgment results and confidence coefficients of the two models are obtained; and finally, comprehensively analyzing the carbon peak reaching judgment results and confidence coefficients of the two models by using fuzzy logic to obtain a final judgment result of the urban carbon peak reaching condition. According to the method, comprehensive, real-time and accurate power data is taken as a core, urban carbon emission multi-dimensional influence factors are coupled, fuzzy logic processing is carried out on carbon peak reaching judgment results of a dynamic weight model and an extended unhooking model, one-sidedness errors possibly caused by a single index and a single judgment method are broken through, and the accuracy of the judgment result is improved. And a more accurate urban carbon peak reaching condition judgment result is obtained, systematic and dynamic monitoring of the urban carbon peak reaching process is realized, and a scientific basis is provided for urban carbon peak reaching path planning and policy making.

Description

Technical Field

[0001] This invention belongs to the field of new energy and energy conservation technology, specifically relating to a method and system for determining urban carbon peak based on power data and fuzzy logic. Background Technology

[0002] Currently, urban carbon peak monitoring and assessment still faces several bottlenecks that urgently need to be addressed: First, domestic monitoring and analysis work is mostly concentrated at the national and provincial levels. Only a small number of low-carbon pilot cities designated by the state are required to conduct targeted carbon emission monitoring, while the vast majority of non-pilot cities generally lack systematic monitoring requirements and assessment capabilities, resulting in incomplete monitoring coverage. Second, existing carbon peak monitoring and assessment rely heavily on macroeconomic energy statistics, which suffer from problems such as coarse data granularity, long update cycles, and insufficient timeliness, making it difficult to accurately capture the dynamic characteristics and peaking trends of urban carbon emissions. Third, research on the coupling and decoupling effects of key factors in urban carbon emissions and electricity data is relatively weak, and the lack of long-term series carbon emission data at the city level restricts the depth of empirical research on urban carbon peaking. Therefore, a mature and reliable dedicated monitoring and assessment system is urgently needed to monitor and assess urban carbon peaking. Summary of the Invention

[0003] The purpose of this invention is to address the aforementioned problems in the existing technology by providing a method and system for determining urban carbon peaking based on power data and fuzzy logic.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] In a first aspect, this invention proposes a method for determining urban carbon peaking based on electricity data and fuzzy logic, including:

[0006] S1. Select the indicators that have the greatest impact on urban carbon emission intensity from the electricity data;

[0007] S2. Based on the selected indicators, construct a dynamic weight model and an extended decoupling model to determine the carbon peaking status of the city, and obtain the carbon peaking determination results and confidence levels of the two models.

[0008] S3. Using fuzzy logic, a comprehensive analysis of the carbon peak determination results and confidence levels of the two models is conducted to obtain the final determination result of the city's carbon peak situation.

[0009] In S2, the dynamic weight model is as follows:

[0010] ;

[0011] In the above formula, For the target city Carbon intensity index at any given time This represents the total number of indicators for determining whether a city has reached its carbon peak. For the first Item judgment indicators in Dynamic weights at time points, For the first Item judgment indicators in Standardized value of time, This is an error correction term;

[0012] The carbon peak determination results from the dynamic weighting model include:

[0013] If the observation period for the target city is less than A years, the target city is determined to be "not at its peak". The observation period refers to the period from the start time when continuous historical data can be obtained to the current time.

[0014] like It has been declining for B consecutive months, and currently... If the value is ≤95% of the historical peak value and shows a significant trend through the Mann-Kendall trend test, then the target city is determined to have "reached its peak".

[0015] like Within ±5% of the historical peak, a state showing no significant trend according to the Mann-Kendall test persists for more than C months, and within this period... If the standard deviation or mean is ≤3%, it is considered a "plateau period" in the target city.

[0016] If the current >105% of historical peak, or If a city shows a clear and continuous upward trend, or does not meet the above criteria for "peak reached" or "plateau period", then the target city is determined to be "not yet at its peak".

[0017] The confidence level of the dynamic weight model is:

[0018] ;

[0019] ;

[0020] ;

[0021] In the above formula, The confidence level of the dynamic weighting analysis model. For the trend test results, This is the statistic for the Mann-Kendall trend test. To examine the total number of periods, , Each represents a different moment. , The target cities are respectively time, Carbon intensity index at any given time.

[0022] In S2, the extended decoupling model is as follows:

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] ;

[0028] ;

[0029] In the above formula, For the target city The decoupling index at any given moment. for Changes in urban carbon emission intensity over time for Changes in urban energy efficiency over time. for Changes in urban electricity productivity at any given time. for Changes in urban electricity consumption at any given time. The carbon emissions of the target city at the baseline time. The electricity consumption of the target city at the baseline time. The logarithmic average weighting factor is... for Urban carbon emission intensity at any given time The carbon emission intensity of the target city at the baseline time. for Urban energy efficiency at all times The energy efficiency of the target city at the baseline time. for Urban electricity productivity at any given time The electricity productivity of the target city at the baseline time. for Urban electricity consumption at any given time for Urban carbon emissions at any given moment;

[0030] The carbon peak determination results of the extended decoupling model include:

[0031] like If the value is less than -0.3 for D consecutive quarters, the target city is determined to have "peaked".

[0032] like If the value is greater than or equal to -0.3 and less than or equal to 0.2 for E consecutive quarters, it is determined to be a "plateau period" for the target city.

[0033] like For any F quarters, the value is greater than 0.2 and less than or equal to 0.8, or If the value is greater than 0.8 in any quarter, the target city is judged as "not reaching its peak";

[0034] The confidence level of the extended decoupling model is:

[0035] ;

[0036] In the above formula, To expand the confidence level of the decoupling model.

[0037] S3 includes:

[0038] S31. Input the carbon peak determination results and confidence levels of the dynamic weight model and the extended decoupling model, and convert the carbon peak determination results of the dynamic weight model and the extended decoupling model into three-dimensional state vectors respectively, and fuzzify the confidence levels of the dynamic weight model and the extended decoupling model.

[0039] S32. Based on the three-dimensional state vector and the fuzzy confidence level, the trigger intensity of each carbon peak determination result for the target city is calculated using the following formula:

[0040] ;

[0041] ;

[0042] ;

[0043] In the above formula, For the target city in the results The trigger strength, when The time has reached its peak, when During the plateau period, when It was before the peak was reached. , All are weights. Carbon peak determination results for dynamic weighted model The three-dimensional state vector, To extend the carbon peak determination results of the decoupling model The three-dimensional state vector, The confidence level of the dynamic weighting model's determination results corresponds to the membership level of higher-level members. To extend the confidence level of the decoupling model's determination results to higher-level membership;

[0044] S33. The following formula is used to correct the trigger intensity of the carbon peak determination results for each target city:

[0045] ;

[0046] ;

[0047] ;

[0048] ;

[0049] ;

[0050] ;

[0051] In the above formula, For the target city in the results Corrected trigger strength , , These represent the confidence levels of the dynamic weighting model's judgment results, and their membership degrees to the low, medium, and high levels, respectively. , , These represent the confidence levels of the extended decoupling model's determination results, and their membership degrees at the low, medium, and high levels, respectively.

[0052] S34. Based on the trigger intensity of the carbon peak determination results for each target city after correction, obtain the initial output vector:

[0053] ;

[0054] In the above formula, This is the initial output vector. The trigger strength after peak correction. This is the trigger strength after the plateau period adjustment. The trigger strength after correction for not reaching the peak;

[0055] S35. Normalize the initial output vector to obtain the final result of the target city's carbon peak status determination, including:

[0056] like If so, the target city is determined to have "reached its peak";

[0057] like If so, the target city is determined to be in a "plateau period";

[0058] like If so, the target city is determined to be "not yet at its peak".

[0059] S1 includes:

[0060] S11. Calculate the Pearson correlation coefficient between each indicator and carbon emission intensity. Filter out Indicators;

[0061] S12. Calculate the variance inflation factor of the selected indicators, and retain the indicators with a variance inflation factor of less than 5 to form an indicator system for determining urban carbon peaking.

[0062] Secondly, this invention proposes an urban carbon peak determination system based on power data and fuzzy logic, including an indicator screening module, a model building module, and a comprehensive analysis module.

[0063] The indicator screening module is used to screen the indicators that have the greatest impact on urban carbon emission intensity from the electricity data.

[0064] The model building module is used to construct a dynamic weight model and an extended decoupling model based on the selected indicators to determine the carbon peaking status of the city, and to obtain the carbon peaking determination results and confidence levels of the two models.

[0065] The comprehensive analysis module is used to perform a comprehensive analysis of the carbon peak determination results and confidence levels of the two models using fuzzy logic, so as to obtain the final determination result of the city's carbon peak status.

[0066] The model building module includes a dynamic weight model building unit, a dynamic weight model carbon peak determination unit, and a dynamic weight model confidence calculation unit.

[0067] The dynamic weight model construction unit is used to construct the following dynamic weight model:

[0068] ;

[0069] In the above formula, For the target city Carbon intensity index at any given time This represents the total number of indicators for determining whether a city has reached its carbon peak. For the first Item judgment indicators in Dynamic weights at time points, For the first Item judgment indicators in Standardized value of time, This is an error correction term;

[0070] The dynamic weighted model carbon peak determination unit is used to obtain the carbon peak determination result of the dynamic weighted model, including:

[0071] If the observation period for the target city is less than A years, the target city is determined to be "not at its peak". The observation period refers to the period from the start time when continuous historical data can be obtained to the current time.

[0072] like It has been declining for B consecutive months, and currently... If the value is ≤95% of the historical peak value and shows a significant trend through the Mann-Kendall trend test, then the target city is determined to have "reached its peak".

[0073] like Within ±5% of the historical peak, a state showing no significant trend according to the Mann-Kendall test persists for more than C months, and within this period... If the standard deviation or mean is ≤3%, it is considered a "plateau period" in the target city.

[0074] If the current >105% of historical peak, or If a city shows a clear and continuous upward trend, or does not meet the above criteria for "peak reached" or "plateau period", then the target city is determined to be "not yet at its peak".

[0075] The dynamic weight model confidence calculation unit is used to calculate the confidence of the dynamic weight model using the following formula:

[0076] ;

[0077] ;

[0078] ;

[0079] In the above formula, The confidence level of the dynamic weighting analysis model. For the trend test results, This is the statistic for the Mann-Kendall trend test. To examine the total number of periods, , Each represents a different moment. , The target cities are respectively time, Carbon intensity index at any given time.

[0080] The model building module also includes an extended decoupling model building unit, an extended decoupling model carbon peak determination unit, and an extended decoupling model confidence calculation unit.

[0081] The extended decoupling model construction unit is used to construct the following extended decoupling model:

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] ;

[0087] ;

[0088] In the above formula, For the target city The decoupling index at any given moment. for Changes in urban carbon emission intensity over time for Changes in urban energy efficiency over time. for Changes in urban electricity productivity at any given time. for Changes in urban electricity consumption at any given time. The carbon emissions of the target city at the baseline time. The electricity consumption of the target city at the baseline time. The logarithmic average weighting factor is... for Urban carbon emission intensity at any given time The carbon emission intensity of the target city at the baseline time. for Urban energy efficiency at all times The energy efficiency of the target city at the baseline time. for Urban electricity productivity at any given time The electricity productivity of the target city at the baseline time. for Urban electricity consumption at any given time for Urban carbon emissions at any given moment;

[0089] The extended decoupling model carbon peak determination unit is used to obtain the carbon peak determination result of the extended decoupling model, including:

[0090] like If the value is less than -0.3 for D consecutive quarters, the target city is determined to have "peaked".

[0091] like If the value is greater than or equal to -0.3 and less than or equal to 0.2 for E consecutive quarters, it is determined to be a "plateau period" for the target city.

[0092] like For any F quarters, the value is greater than 0.2 and less than or equal to 0.8, or If the value is greater than 0.8 in any quarter, the target city is judged as "not reaching its peak";

[0093] The extended decoupling model confidence calculation unit is used to calculate the confidence of the extended decoupling model using the following formula:

[0094] ;

[0095] In the above formula, To expand the confidence level of the decoupling model.

[0096] The comprehensive analysis module includes an input preprocessing unit, a trigger intensity calculation unit, a trigger intensity correction unit, an initial output vector formation unit, and a judgment result normalization unit.

[0097] The input preprocessing unit is used to input the carbon peak determination results and confidence levels of the dynamic weight model and the extended decoupling model, and converts the carbon peak determination results of the dynamic weight model and the extended decoupling model into three-dimensional state vectors, and fuzzifies the confidence levels of the dynamic weight model and the extended decoupling model.

[0098] The trigger intensity calculation unit is used to calculate the trigger intensity of each carbon peak determination result of the target city based on the three-dimensional state vector and the fuzzified confidence level, using the following formula:

[0099] ;

[0100] ;

[0101] ;

[0102] In the above formula, For the target city in the results The trigger strength, when The time has reached its peak, when During the plateau period, when It was before the peak was reached. , All are weights. Carbon peak determination results for dynamic weighted model The three-dimensional state vector, To extend the carbon peak determination results of the decoupling model The three-dimensional state vector, The confidence level of the dynamic weighting model's determination results corresponds to the membership level of higher-level members. To extend the confidence level of the decoupling model's determination results to higher-level membership;

[0103] The trigger intensity correction unit is used to correct the trigger intensity of each carbon peak determination result for the target city using the following formula:

[0104] ;

[0105] ;

[0106] ;

[0107] ;

[0108] ;

[0109] ;

[0110] In the above formula, For the target city in the results Corrected trigger strength , , These represent the confidence levels of the dynamic weighting model's judgment results, and their membership degrees to the low, medium, and high levels, respectively. , , These represent the confidence levels of the extended decoupling model's determination results, and their membership degrees at the low, medium, and high levels, respectively.

[0111] The initial output vector forming unit is used to obtain the initial output vector based on the trigger intensity of the corrected carbon peak determination results for each target city:

[0112] ;

[0113] In the above formula, This is the initial output vector. The trigger strength after peak correction. This is the trigger strength after the plateau period adjustment. The trigger strength after correction for not reaching the peak;

[0114] The result normalization unit is used to normalize the initial output vector to obtain the final result of the target city's carbon peak status, including:

[0115] like If so, the target city is determined to have "reached its peak";

[0116] like If so, the target city is determined to be in a "plateau period";

[0117] like If so, the target city is determined to be "not yet at its peak".

[0118] The indicator screening module includes a Pearson indicator judgment unit and a judgment indicator system formation unit;

[0119] The Pearson index judgment unit is used to calculate the Pearson correlation coefficient between each index and carbon emission intensity. Filter out Indicators;

[0120] The judgment index system forming unit is used to calculate the variance inflation factor of the selected indicators, and retain the indicators with a variance inflation factor of less than 5 to form the urban carbon peak judgment index system.

[0121] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0122] This invention proposes a method and system for determining urban carbon peaking based on electricity data and fuzzy logic. The method first filters the indicators with the greatest impact on urban carbon emission intensity from electricity data. Then, based on the selected indicators, a dynamic weighting model and an extended decoupling model are constructed to determine the urban carbon peaking situation, obtaining the carbon peaking determination results and confidence levels of the two models. Finally, fuzzy logic is used to comprehensively analyze the carbon peaking determination results and confidence levels of the two models to obtain the final determination result of the urban carbon peaking situation. On the one hand, this method couples multiple influencing factors of electricity data, overcoming the one-sidedness that may be caused by single-indicator determination. With comprehensive, real-time, and accurate data as the core, it solves the problems of limited scope and poor timeliness of traditional statistical methods, enabling real-time monitoring and determination of urban carbon peaking. On the other hand, this method processes the determination results of the dynamic weighting model and the extended decoupling model using fuzzy logic, which can eliminate the errors of single determination methods, obtaining more accurate determination results of urban carbon peaking, realizing systematic and dynamic monitoring of the urban carbon peaking process, and providing a scientific basis for urban carbon peaking path planning and policy formulation. Attached Figure Description

[0123] Figure 1 This is an overall flowchart of the method described in this invention.

[0124] Figure 2 This is a structural diagram of the system described in this invention. Detailed Implementation

[0125] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0126] This invention proposes a method and system for determining urban carbon peaking based on electricity data and fuzzy logic. The aim is to preprocess and filter relevant indicators of urban carbon emissions using big data on electricity as the main thread; based on the preprocessed and filtered indicators, a dynamic weighting model and an extended decoupling model are used to determine the urban carbon peaking situation; finally, based on the determination results of the dynamic weighting model and the extended decoupling model, fuzzy logic is used for comprehensive analysis to obtain the final determination result of the urban carbon peaking situation. This enables systematic and dynamic monitoring of the urban carbon peaking process, providing a scientific basis for urban carbon peaking path planning and policy formulation.

[0127] Example 1:

[0128] like Figure 1 As shown, the urban carbon peak determination method based on electricity data and fuzzy logic proceeds in the following steps:

[0129] 1. Screen the indicators that have the greatest impact on urban carbon emission intensity from the electricity data;

[0130] Taking into account multiple dimensions of power data, including power production, power transmission and distribution, and power consumption, the following indicators are considered: power production indicators include: total installed capacity and structure, total power generation, thermal / coal-fired power generation and its proportion, non-fossil / renewable energy power generation and its proportion, and power generation carbon emission intensity; power transmission and distribution indicators include: grid comprehensive line loss rate, grid carbon emission factor, and main transformer load rate; grid consumption indicators include: total social electricity consumption, electricity consumption by industry / sector, residential electricity consumption, non-fossil / renewable energy electricity consumption and its proportion, maximum electricity load and peak-valley difference, and the proportion of adjustable resources. These indicators are integrated into a unified format according to time series to form the original dataset.

[0131] First, outlier detection is performed on the original dataset, including:

[0132] use The initial principle is to screen outliers in each indicator's data: for example, for outliers in power data caused by equipment failure or metering errors, the mean of each data type is calculated. and standard deviation Marking out of range Data points.

[0133] Using box plots The outliers selected according to the principles are verified, and false positives are removed: Outliers selected according to the principle are sorted from smallest to largest, and the first quartile is calculated. , median Third and quartiles and interquartile range ; and then and Draw a box around the boundary, and label the median with horizontal lines inside the box. Straps extend from both ends of the box, with the upper end not exceeding [a certain value]. The maximum value, the lower end is not less than The minimum value is used to identify and mark outliers that exceed the range of the beard line as abnormal data.

[0134] The principle of box plots strictly relies on the assumption that the data follows or approximately follows a normal distribution, and is less sensitive to skewed distributions. For small datasets or data that does not follow a normal distribution, it may misclassify many normal data points as outliers. Box plots, on the other hand, perform better when dealing with skewed data, but may be overly conservative or overly sensitive under certain distributions. Therefore, combining the two, box plots can handle non-normally distributed cases and compensate for the shortcomings of box plots. The biggest drawback of this principle is that when the data is obviously non-normal, the results of the box plot can be used for correction, but when the data is close to normal, the two can corroborate each other.

[0135] Then, the identified anomalous data is corrected, including:

[0136] For the identified anomaly data in the original dataset, select the data before and after it as the center. Each data point constitutes a local neighborhood; The value is generally between 5 and 10, and is dynamically adjusted according to the fluctuation characteristics of the actual data. For example, for power data... );

[0137] The weights of each data point in the local neighborhood are calculated using a Gaussian weighting function:

[0138] ;

[0139] In the above formula, For the local neighborhood of the first The weight of each data point For the local neighborhood of the first Data points, This is abnormal data. The standard deviation is the Gaussian distribution, used to control the rate of weight decay. The value is usually determined based on the neighborhood. The standard deviation setting, or adjustment via cross-validation;

[0140] Solving within a local neighborhood Determine the regression coefficients , ,in For the local neighborhood of abnormal data;

[0141] Finally, abnormal data Substitution In the middle, the predicted value of outlier data Replace the identified outliers in the original dataset to form an initial dataset. (The initial dataset can be substituted into the new dataset.) (Verification is performed within the interval).

[0142] Further collaborative data governance is carried out on the initial dataset, including:

[0143] Cubic spline interpolation is employed for initial datasets of the target city at different time scales, such as annual GDP data and monthly per capita electricity consumption. The initial data are arranged chronologically to form an initial node set. A cubic polynomial is constructed between adjacent initial data nodes to ensure the spline function is a continuous and smooth cubic curve, thereby obtaining interpolations that best reflect the trend.

[0144] ;

[0145] In the above formula, For the first The interpolated data for the nth interval, the nth interval The interval refers to the first interval in the initial node set. Data The surrounding area For constant terms, The coefficient is a first-order coefficient. For the target time point where interpolation data is needed, For the first Initial data for each interval, The coefficient is a quadratic coefficient. It is a cubic coefficient;

[0146] By utilizing the continuity of function values, the continuity conditions of first and second derivatives, and natural boundary conditions (taking electricity data as an example, assuming the second derivative of electricity data changes is zero in the beginning and end months of the year, indicating that the trend of electricity data changes is relatively stable at the beginning and end of the year, without sudden acceleration or deceleration, which conforms to the normal characteristics of power system operation), a system of linear equations about the coefficient terms is established and solved, including:

[0147] Calculate intermediate quantities:

[0148] ;

[0149] ;

[0150] ;

[0151] ;

[0152] In the above formula, The interval step size represents the time interval between adjacent data points. , The weighting coefficient represents the influence of the length of the adjacent interval on the second derivative of the current node. The right-hand side term represents the weighted average rate of change of the first derivative at the nodes. express In data The value of is the constant term. ;

[0153] Establishing the second derivative The system of equations for the three diagonals:

[0154] ;

[0155] In the above formula, for The second derivative, The total number of intervals;

[0156] Solving the above system of equations yields The coefficients of each polynomial are as follows:

[0157] ;

[0158] ;

[0159] ;

[0160] ;

[0161] Based on the coefficients of each polynomial, at the target scale (e.g., unified to monthly), the time points are substituted into the polynomial function of the corresponding interval to generate interpolated data for each interval, thereby achieving time scale unification of multi-dimensional power data, forming a preliminary dataset, and providing adaptive data for subsequent indicator construction.

[0162] For initial datasets with coarse granularity, cubic spline interpolation constructs interpolation functions piecewise, strictly passing through all time series data points. The resulting curves are smooth, with controllable local errors, ultimately generating interpolations that best reflect the actual trend, thus refining the granularity of the original data.

[0163] Finally, the prepared dataset was standardized, and collinearity analysis was performed on the standardized dataset, including:

[0164] For indicator data in the preliminary dataset that conform to an approximately normal distribution, such as the installed capacity of renewable energy in some cities and the comprehensive line loss rate of the power grid, the Z-Score standardization method is used to transform the data into a standard form with a mean of 0 and a standard deviation of 1, eliminating differences in dimensions and making such indicators comparable between different cities.

[0165] ;

[0166] In the above formula, For the standardized dataset, To prepare the dataset, To prepare the average value of the dataset, The standard deviation of the prepared dataset;

[0167] For indicator data in the preliminary dataset with no specific pattern of distribution, Min-Max standardization is used to map the data to the [0,1] interval, highlighting the relative positional relationships of indicators in different cities:

[0168] ;

[0169] In the above formula, For the standardized dataset, , These are the minimum and maximum values ​​of the dataset, respectively.

[0170] Calculate the Pearson correlation coefficients between each indicator in the standardized preliminary dataset and carbon emission intensity. Filter out Indicators such as the proportion of renewable energy installed capacity and electricity consumption per unit of GDP are closely related to carbon emissions and will be retained for subsequent analysis.

[0171] Calculate the variance inflation factor of the selected indicators, and retain the data of indicators with a variance inflation factor of less than 5 to form an indicator system for judging urban carbon peaking.

[0172] The variance inflation factor is calculated using the following formula:

[0173] ;

[0174] In the above formula, For the first The variance inflation factor of the evaluation index. Indicates the first The fitting effect of the evaluation index with other indicators;

[0175] for The calculation uses the target indicator as the dependent variable and other indicators as independent variables to establish a regression model. ;in, arrive After preprocessing, it is aligned; this expression means that it is aligned. arrive This expression holds true for any value;

[0176] ;

[0177] ;

[0178] Ultimate goal factors Represented as:

[0179] ;

[0180] In the above formula, For the first Item judgment indicators (target indicators) , To exclude the first Other indicators for the determination of the item, , , These are all regression coefficients (which can be solved using the least squares method and the normal equation). This is the random error term, used to reflect nonlinear relationships that the regression model cannot capture (therefore for...). arrive Each set of values, Different, by (Solve) For the first The sum of squares of the criteria, For the first The first of the judgment indicators Each sample value For the first The sample mean of the evaluation index. For the first The sum of squared residuals of the criteria. The regression model established for the first The first of the judgment indicators Predicted values ​​for each sample;

[0181] After standardization and collinearity analysis, 12 indicators were selected, including electricity consumption per unit of GDP, the proportion of renewable energy installed capacity, and carbon emission intensity per unit of power generation. (Different cities will have different indicator systems that have the greatest impact on urban carbon emission intensity after screening. For example, for industrial cities, it may be the proportion of electricity consumption in high-energy-consuming industries and electricity consumption per unit of GDP; for service-oriented cities, it may be the growth rate of electricity consumption in the tertiary industry and residential electricity consumption; for energy resource-based cities, it may be the proportion of thermal power generation and renewable energy installed capacity.) Dynamic weight analysis and extended decoupling index were used to determine the carbon peaking status of cities, and then fuzzy logic was combined to analyze and determine the carbon peaking status of cities.

[0182] The vast majority of total carbon emissions originate from energy activities and industrial processes, and electricity consumption data, among other electricity-related big data, shows a strong correlation with carbon emissions. By using electricity big data as the main framework, a corresponding carbon peak monitoring system can be constructed to achieve urban carbon peak monitoring and analysis, thus contributing to the early realization of my country's dual-carbon goals.

[0183] 2. Based on the selected indicators, a dynamic weighting model and an extended decoupling model were constructed to determine the carbon peaking status of the city, and the carbon peaking determination results and confidence levels of the two models were obtained.

[0184] Urban carbon peaking based on dynamic weighting analysis includes:

[0185] First, calculate the initial weights of the indicators for determining urban carbon peaking;

[0186] Based on the selected urban carbon peak determination index system, an urban carbon peak determination index matrix is ​​constructed. ,in The total number of indicators for determining whether a city has reached its carbon peak depends on the specific analysis of the target city and may be 12 or fluctuate slightly above or below. The sample size is for a single indicator (the previous step has already performed collaborative governance and standardization on various indicators), so the specific data volume is the same for different indicators.

[0187] ;

[0188] ;

[0189] ;

[0190] In the above formula, For the first The initial weights of the evaluation indicators, For the first The entropy value of the judgment index, For the first The entropy value of the judgment index, for In all the The proportion of each judgment indicator For matrix The element in the text means the first The first of the judgment indicators One sample value (standardized value).

[0191] Then, based on the initial weights of the urban carbon peak determination indicators, the weights of the urban carbon peak determination indicators are dynamically adjusted.

[0192] Considering that the weights of various criteria in the process of achieving carbon peaking in cities are affected by policy guidance, event-driven factors, seasonal changes, etc., corresponding factors are introduced to dynamically adjust the initial weights of each criterion.

[0193] ;

[0194] In the above formula, For the first Item judgment indicators in time Dynamic weights, Indicates the policy orientation correction factor. For the first The judgment indicator is subject to the first The policy sensitivity coefficient for the impact of relevant policies is set to 0.3 by default, but can be adjusted based on the actual situation of the target city, experience, expert knowledge, etc. For the first The judgment indicator is subject to the first The policy intensity index for the impact of relevant policies can be used, for example, 0.4 for national-level policies, 0.3 for provincial-level policies, and 0.2 for municipal-level policies. This index can also fluctuate by 0.1 depending on the urgency of the policy and can be adjusted according to the actual situation. For the first The judgment indicator is subject to the first The time decay coefficient of the impact of a relevant policy indicates how much the policy's impact diminishes over time. The time of policy release For the first The attenuation coefficient for each policy can be selected based on the nature of the policy itself. For example, 30 days can be used for emergency policies, 90 days for routine policies, and 180 days for strategic policies. For the first The seasonal variation correction factor for the first determination indicator depends on the first... The selection of indicators should consider their specific nature and seasonal characteristics. For example, for "per capita electricity consumption in urban areas," a suitable indicator could be chosen during the summer when air conditioning load surges. Regarding "total electricity consumption," it is advisable to take advantage of peak industrial production seasons. wait, Indicates the event-driven correction factor. For the first The judgment indicator is subject to the first The event sensitivity coefficient affected by each related event. For the first The judgment indicator is subject to the first The intensity of the impact of a related event. For the first The judgment indicator is subject to the first The time decay coefficient of the impact of a related event indicates how much the impact of the event diminishes over time. For the first The time of occurrence of each event For the first The decay coefficient of an event; , , The value is similar to the policy guidance, and is selected based on the actual situation, experience, and expert knowledge of the target city, depending on the type of event.

[0195] (The above analysis is only an example and can be calculated and processed according to the actual situation of the target city);

[0196] In summary, the final revised dynamic weight vector for urban carbon emission assessment indicators is as follows:

[0197] ;

[0198] In the above formula, In time The dynamic weight vector, , , , For each judgment indicator in time The corrected dynamic weights, This is for normalization purposes.

[0199] Finally, based on the dynamic weight adjustment results of the urban carbon peak determination indicators, the dynamic weight model is constructed as follows:

[0200] ;

[0201] In the above formula, For the target city Carbon intensity index at any given time This represents the total number of indicators for determining whether a city has reached its carbon peak. For the first Item judgment indicators in Dynamic weights at time points, For the first Item judgment indicators in Standardized value of time, This is the error correction term.

[0202] The carbon peak determination results based on the dynamic weighting model include:

[0203] If the observation period for the target city is less than 5 years, the target city is determined to be "not yet at its peak". The observation period refers to the period from the start time when continuous historical data can be obtained to the present time, which must include at least the past 5 years. Monthly calculation results, if the data volume is less than 5 years, even if It has been showing a downward trend month by month, and is also judged as "not reaching the peak" because cities still need a certain amount of time to test and prove that they have not reached the peak falsely after carbon reaches the peak, and trend verification cannot be performed when the amount of data is too small;

[0204] like It has been declining for 12 consecutive months, and currently... ≤95% of historical peak value, showing a significant trend through the Mann-Kendall trend test, i.e., satisfying the standardized statistic. Significance If so, the target city is determined to have "reached its peak";

[0205] like Within ±5% of the historical peak, no significant trend was observed according to the Mann-Kendall test, thus satisfying the significance standard. The condition lasts for more than 6 months, and within the duration of the period If the standard deviation or mean is ≤3%, it is considered a "plateau period" in the target city.

[0206] If the current >105% of historical peak, or If a city shows a clear and continuous upward trend, or does not meet the above criteria for "peak reached" or "plateau period", then the target city is determined to be "not yet at its peak".

[0207] A detailed explanation of the three states: "Peak reached" is a complete process, not a point in time. It refers to the process in which carbon dioxide emissions reach their historical highest value, experience a plateau period, and then continue to decline. "Plateau period" indicates that carbon emissions have reached their peak, but have not yet formed a stable downward trend. They may still be affected by short-term economic fluctuations and emissions may rebound (but not exceed the peak). It is the "critical point" for low-carbon transformation, but the transformation is not yet complete. "Not peaked" means that emissions are generally on an upward trend with no obvious signs of decline, or although there are short-term fluctuations, a stable peak has not been formed. If there is a decline, it is mostly due to short-term factors such as economic recession, rather than structural changes.

[0208] After obtaining the judgment result, it is also necessary to calculate the confidence level of the judgment result using a dynamic weight model analysis:

[0209] ;

[0210] ;

[0211] ;

[0212] In the above formula, The confidence level of the dynamic weighting analysis model. For the trend test results, This is the statistic for the Mann-Kendall trend test. To examine the total number of periods, , Each represents a different moment. , The target cities are respectively time, Carbon intensity index at any given time;

[0213] When the confidence level is ≥0.75, the results are basically reliable; when the confidence level is ≥0.80, the results are highly reliable; when the confidence level is ≥0.85, the results can be used for decision-making applications; for confidence levels that do not meet the requirements, when 0.65≤confidence level<0.75, the data window can be extended or the data granularity can be refined; when the confidence level<0.65, sliding window verification or other processing methods can be used.

[0214] Urban carbon peaking based on extended decoupling analysis includes:

[0215] First, the multi-dimensional indicators for determining urban carbon peaking are broken down.

[0216] To determine the peak carbon emissions of cities using the comprehensive decoupling index, it is first necessary to decompose the factors affecting changes in urban carbon emissions:

[0217] ;

[0218] In the above formula, for Urban carbon emissions at any given time for Total urban energy consumption at any given time for The city's total GDP at any given time for Urban electricity consumption at any given time for Urban carbon emission intensity at any given time for Urban energy efficiency at all times for Urban electricity productivity at any given moment;

[0219] (The parameters in the above formula are for illustrative purposes only. The corresponding parameters can be expanded or reduced according to the actual situation of the target city, and the same applies to subsequent analysis.)

[0220] Based on the above analysis, the target city is then... The change in carbon emissions at time 1 compared to the baseline time can be decomposed into the combined effect of various factors:

[0221] ;

[0222] ;

[0223] ;

[0224] ;

[0225] ;

[0226] ;

[0227] In the above formula, for Changes in urban carbon emissions over time. The carbon emissions of the target city at the baseline time. for Changes in urban carbon emission intensity over time for Changes in urban energy efficiency over time. for Changes in urban electricity productivity at any given time. for Changes in urban electricity consumption at any given time. This refers to the logarithmic average weighting factor in the logarithmic average Dichotomy decomposition method, used to calculate the contribution of each factor while ensuring the completeness of the decomposition. The carbon emission intensity of the target city at the baseline time. The energy efficiency of the target city at the baseline time. The electricity productivity of the target city at the baseline time. The electricity consumption of the target city at the baseline time;

[0228] The final extended decoupling model is as follows:

[0229] ;

[0230] In the above formula, For the target city The decoupling index at any given moment. For urban carbon emission intensity at The decoupling index at any given moment. For urban energy efficiency The decoupling index at any given moment. For urban electricity productivity The decoupling index at any given moment. For urban electricity consumption in The decoupling index at any given moment.

[0231] The carbon peak determination results based on the extended decoupling model include:

[0232] like If a city's carbon emissions are less than -0.3 for four consecutive quarters, and the city's carbon emissions show a strong decoupling relationship with various influencing factors, then the target city is determined to have "reached its peak."

[0233] like If a city's carbon emissions are greater than or equal to -0.3 and less than or equal to 0.2 for six consecutive quarters, and the city's carbon emissions show a weak decoupling relationship with various influencing factors, then it is determined to be in a "plateau period" for the target city.

[0234] like For any two quarters, the value is greater than 0.2 and less than or equal to 0.8, or If a city's carbon emissions exceed 0.8 in any quarter, and the city's carbon emissions show an expanding negative decoupling relationship with various influencing factors, then the target city is determined to have "not reached its peak".

[0235] After obtaining the judgment result, it is also necessary to calculate the confidence level of the judgment result of the extended decoupling model:

[0236] ;

[0237] In the above formula, To expand the confidence level of the decoupling model, the number of strong decoupling factors was increased by... , , , The specific value will be used to determine this.

[0238] 3. Using fuzzy logic, a comprehensive analysis of the carbon peak determination results and confidence levels of the two models is conducted to obtain the final determination result of the city's carbon peak situation.

[0239] Carbon peak determination results from input dynamic weight model Confidence of dynamic weight model Carbon peak determination results of extended decoupling model Confidence of the extended decoupling model The carbon peak determination results of the dynamic weight model and the extended decoupling model are converted into three-dimensional state vectors, and the confidence of the dynamic weight model and the extended decoupling model is fuzzified.

[0240] The three-dimensional state vector of the carbon peak determination result of the dynamic weight model is:

[0241] , ;

[0242] , ;

[0243] , ;

[0244] The three-dimensional state vector of the carbon peak determination result of the extended decoupling model is:

[0245] , ;

[0246] , ;

[0247] , ;

[0248] The confidence level of the dynamic weight model, after fuzzification, is represented as:

[0249] ;

[0250] In the above formula, , , These represent the confidence level of the dynamic weight model's judgment results and their membership in the low, medium, and high levels, respectively.

[0251] Where: If , ;

[0252] like , ;

[0253] like , ;

[0254] like , ;

[0255] like , ;

[0256] like , ;

[0257] like , ;

[0258] like , ;

[0259] like , ;

[0260] like , ;

[0261] like , ;

[0262] The confidence fuzzification process for the extended decoupling model is the same, and will not be repeated here.

[0263] Based on the fuzzy confidence scores of the dynamic weighting model and the extended decoupling model, the confidence score coordination degree between the two models is calculated:

[0264] ;

[0265] ;

[0266] ;

[0267] ;

[0268] In the above formula, , , These represent the confidence levels of the extended decoupling model's determination results, and their membership degrees at the low, medium, and high levels, respectively.

[0269] when At that time, the confidence level consistency value is 0.5.

[0270] Fuzzy inference is performed on the carbon peak determination results of each target city, and the trigger intensity of each carbon peak determination result of the target city is calculated using the following formula:

[0271] ;

[0272] ;

[0273] ;

[0274] In the above formula, For the target city in the results The trigger strength, when The time has reached its peak, when During the plateau period, when It was before the peak was reached. , All are weights. The default value is 0.7. The default value is 0.3, meaning the state support weight is relatively high. Carbon peak determination results for dynamic weighted model The three-dimensional state vector, To extend the carbon peak determination results of the decoupling model The three-dimensional state vector, The confidence level of the dynamic weighting model's determination results corresponds to the membership level of higher-level members. To expand the confidence level of the decoupling model's judgment results to the higher-level membership degree, the logic here is to take the maximum value of the high-confidence membership degree of the two models, reflecting the model's degree of confidence in its own judgment.

[0275] When the judgments of the two models differ, a conflict is considered to exist. The handling strategy is to adopt a conservative approach and increase the trigger strength of the "plateau period" state.

[0276] ;

[0277] ;

[0278] In the above formula, To enhance the strength of the plateau phase trigger, the greater the degree of conflict, the more serious the divergence between models, and the more conservative the plateau phase determination should be.

[0279] The following formula is used to correct the trigger intensity of the carbon peak determination results for each target city:

[0280] ;

[0281] ;

[0282] In the above formula, For the target city in the results The corrected trigger strength.

[0283] Based on the trigger intensity of the carbon peak determination results for each target city after correction, the initial output vector is obtained:

[0284] ;

[0285] In the above formula, This is the initial output vector. The trigger strength after peak correction. This is the trigger strength after the plateau period adjustment. The trigger strength after correction for not reaching the peak;

[0286] The initial output vector is normalized to obtain the final result of the target city's carbon peak status assessment, including:

[0287] like If so, the target city is determined to have "reached its peak";

[0288] like If so, the target city is determined to be in a "plateau period";

[0289] like If so, the target city is determined to be "not yet at its peak";

[0290] If none of the three conditions are met, the target city is conservatively judged to be in a "plateau period," and a comprehensive judgment is made in conjunction with specific external indicators, including:

[0291] According to the structural transformation indicator, if the target city has new large-scale high-energy-consuming projects starting construction or being approved within the next 1-2 years, it tends to be "before peaking"; if the target city's large coal-fired power units are permanently shut down, or large-scale new energy bases are connected to the grid for power generation, it tends to be "peaking"; according to the policy intensity indicator, if the target city government issues a mandatory annual total energy consumption control target or a mandatory green electricity consumption target, accompanied by clear assessment measures, it tends to be "peaking".

[0292] Example 2:

[0293] like Figure 2 As shown, the urban carbon peak determination system based on electricity data and fuzzy logic includes an indicator screening module, a model building module, and a comprehensive analysis module.

[0294] The indicator screening module is used to screen the indicators that have the greatest impact on urban carbon emission intensity from the electricity data.

[0295] The model building module is used to construct a dynamic weight model and an extended decoupling model based on the selected indicators to determine the carbon peaking status of the city, and to obtain the carbon peaking determination results and confidence levels of the two models.

[0296] The comprehensive analysis module is used to perform a comprehensive analysis of the carbon peak determination results and confidence levels of the two models using fuzzy logic, so as to obtain the final determination result of the city's carbon peak status.

[0297] The indicator screening module includes a Pearson indicator judgment unit and a judgment indicator system formation unit;

[0298] The Pearson index judgment unit is used to calculate the Pearson correlation coefficient between each index and carbon emission intensity. Filter out Indicators;

[0299] The judgment index system forming unit is used to calculate the variance inflation factor of the selected indicators, and retain the indicators with a variance inflation factor of less than 5 to form the urban carbon peak judgment index system.

[0300] The model building module includes a dynamic weight model building unit, a dynamic weight model carbon peak determination unit, and a dynamic weight model confidence calculation unit.

[0301] The dynamic weight model construction unit is used to construct the following dynamic weight model:

[0302] ;

[0303] In the above formula, For the target city Carbon intensity index at any given time This represents the total number of indicators for determining whether a city has reached its carbon peak. For the first Item judgment indicators in Dynamic weights at time points, For the first Item judgment indicators in Standardized value of time, This is an error correction term;

[0304] The dynamic weighted model carbon peak determination unit is used to obtain the carbon peak determination result of the dynamic weighted model, including:

[0305] If the observation period for the target city is less than A years, the target city is determined to be "not at its peak". The observation period refers to the period from the start time when continuous historical data can be obtained to the current time.

[0306] like It has been declining for B consecutive months, and currently... If the value is ≤95% of the historical peak value and shows a significant trend through the Mann-Kendall trend test, then the target city is determined to have "reached its peak".

[0307] like Within ±5% of the historical peak, a state showing no significant trend according to the Mann-Kendall test persists for more than C months, and within this period... If the standard deviation or mean is ≤3%, it is considered a "plateau period" in the target city.

[0308] If the current >105% of historical peak, or If a city shows a clear and continuous upward trend, or does not meet the above criteria for "peak reached" or "plateau period", then the target city is determined to be "not yet at its peak".

[0309] The dynamic weight model confidence calculation unit is used to calculate the confidence of the dynamic weight model using the following formula:

[0310] ;

[0311] ;

[0312] ;

[0313] In the above formula, The confidence level of the dynamic weighting analysis model. For the trend test results, This is the statistic for the Mann-Kendall trend test. To examine the total number of periods, , Each represents a different moment. , The target cities are respectively time, Carbon intensity index at any given time.

[0314] The model building module also includes an extended decoupling model building unit, an extended decoupling model carbon peak determination unit, and an extended decoupling model confidence calculation unit.

[0315] The extended decoupling model construction unit is used to construct the following extended decoupling model:

[0316] ;

[0317] ;

[0318] ;

[0319] ;

[0320] ;

[0321] ;

[0322] In the above formula, For the target city The decoupling index at any given moment. for Changes in urban carbon emission intensity over time for Changes in urban energy efficiency over time. for Changes in urban electricity productivity at any given time. for Changes in urban electricity consumption at any given time. The carbon emissions of the target city at the baseline time. The electricity consumption of the target city at the baseline time. The logarithmic average weighting factor is... for Urban carbon emission intensity at any given time The carbon emission intensity of the target city at the baseline time. for Urban energy efficiency at all times The energy efficiency of the target city at the baseline time. for Urban electricity productivity at any given time The electricity productivity of the target city at the baseline time. for Urban electricity consumption at any given time for Urban carbon emissions at any given moment;

[0323] The extended decoupling model carbon peak determination unit is used to obtain the carbon peak determination result of the extended decoupling model, including:

[0324] like If the value is less than -0.3 for D consecutive quarters, the target city is determined to have "peaked".

[0325] like If the value is greater than or equal to -0.3 and less than or equal to 0.2 for E consecutive quarters, it is determined to be a "plateau period" for the target city.

[0326] like For any F quarters, the value is greater than 0.2 and less than or equal to 0.8, or If the value is greater than 0.8 in any quarter, the target city is judged as "not reaching its peak";

[0327] The extended decoupling model confidence calculation unit is used to calculate the confidence of the extended decoupling model using the following formula:

[0328] ;

[0329] In the above formula, To expand the confidence level of the decoupling model.

[0330] The comprehensive analysis module includes an input preprocessing unit, a trigger intensity calculation unit, a trigger intensity correction unit, an initial output vector formation unit, and a judgment result normalization unit.

[0331] The input preprocessing unit is used to input the carbon peak determination results and confidence levels of the dynamic weight model and the extended decoupling model, and converts the carbon peak determination results of the dynamic weight model and the extended decoupling model into three-dimensional state vectors, and fuzzifies the confidence levels of the dynamic weight model and the extended decoupling model.

[0332] The trigger intensity calculation unit is used to calculate the trigger intensity of each carbon peak determination result of the target city based on the three-dimensional state vector and the fuzzified confidence level, using the following formula:

[0333] ;

[0334] ;

[0335] ;

[0336] In the above formula, For the target city in the results The trigger strength, when The time has reached its peak, when During the plateau period, when It was before the peak was reached. , All are weights. Carbon peak determination results for dynamic weighted model The three-dimensional state vector, To extend the carbon peak determination results of the decoupling model The three-dimensional state vector, The confidence level of the dynamic weighting model's determination results corresponds to the membership level of higher-level members. To extend the confidence level of the decoupling model's determination results to higher-level membership;

[0337] The trigger intensity correction unit is used to correct the trigger intensity of each carbon peak determination result for the target city using the following formula:

[0338] ;

[0339] ;

[0340] ;

[0341] ;

[0342] ;

[0343] ;

[0344] In the above formula, For the target city in the results Corrected trigger strength , , These represent the confidence levels of the dynamic weighting model's judgment results, and their membership degrees to the low, medium, and high levels, respectively. , , These represent the confidence levels of the extended decoupling model's determination results, and their membership degrees at the low, medium, and high levels, respectively.

[0345] The initial output vector forming unit is used to obtain the initial output vector based on the trigger intensity of the corrected carbon peak determination results for each target city:

[0346] ;

[0347] In the above formula, This is the initial output vector. The trigger strength after peak correction. This is the trigger strength after the plateau period adjustment. The trigger strength after correction for not reaching the peak;

[0348] The result normalization unit is used to normalize the initial output vector to obtain the final result of the target city's carbon peak status, including:

[0349] like If so, the target city is determined to have "reached its peak";

[0350] like If so, the target city is determined to be in a "plateau period";

[0351] like If so, the target city is determined to be "not yet at its peak".

Claims

1. A method for determining urban carbon peaking based on electricity data and fuzzy logic, characterized in that, The method includes: S1. Select the indicators that have the greatest impact on urban carbon emission intensity from the electricity data; S2. Based on the selected indicators, construct a dynamic weight model and an extended decoupling model to determine the carbon peaking status of the city, and obtain the carbon peaking determination results and confidence levels of the two models. S3. Using fuzzy logic, a comprehensive analysis of the carbon peak determination results and confidence levels of the two models is conducted to obtain the final determination result of the city's carbon peak situation.

2. The urban carbon peak determination method based on electricity data and fuzzy logic according to claim 1, characterized in that, In S2, the dynamic weight model is: ； In the above formula, For the target city Carbon intensity index at any given time This represents the total number of indicators for determining whether a city has reached its carbon peak. For the first Item judgment indicators in Dynamic weights at time points, For the first Item judgment indicators in Standardized value of time, This is an error correction term; The carbon peak determination results from the dynamic weighting model include: If the observation period for the target city is less than A years, the target city is determined to be "not at its peak". The observation period refers to the period from the start time when continuous historical data can be obtained to the current time. like It has been declining for B consecutive months, and currently... If the value is ≤95% of the historical peak value and shows a significant trend through the Mann-Kendall trend test, then the target city is determined to have "reached its peak". like Within ±5% of the historical peak, a state showing no significant trend according to the Mann-Kendall test persists for more than C months, and within this period... If the standard deviation or mean is ≤3%, it is considered a "plateau period" in the target city. If the current >105% of historical peak, or If a city shows a clear and continuous upward trend, or does not meet the above criteria for "peak reached" or "plateau period", then the target city is determined to be "not yet at its peak". The confidence level of the dynamic weight model is: ； ； ； In the above formula, The confidence level of the dynamic weighting analysis model. For the trend test results, This is the statistic for the Mann-Kendall trend test. To examine the total number of periods, , Each represents a different moment. , The target cities are respectively time, Carbon intensity index at any given time.

3. The urban carbon peak determination method based on electricity data and fuzzy logic according to claim 1, characterized in that, In S2, the extended decoupling model is: ； ； ； ； ； ； In the above formula, For the target city The decoupling index at any given moment. for Changes in urban carbon emission intensity over time for Changes in urban energy efficiency over time. for Changes in urban electricity productivity at any given time. for Changes in urban electricity consumption at any given time. The carbon emissions of the target city at the baseline time. The electricity consumption of the target city at the baseline time. The logarithmic average weighting factor is... for Urban carbon emission intensity at any given time The carbon emission intensity of the target city at the baseline time. for Urban energy efficiency at all times The energy efficiency of the target city at the baseline time. for Urban electricity productivity at any given time The electricity productivity of the target city at the baseline time. for Urban electricity consumption at any given time for Urban carbon emissions at any given moment; The carbon peak determination results of the extended decoupling model include: like If the value is less than -0.3 for D consecutive quarters, the target city is determined to have "peaked". like If the value is greater than or equal to -0.3 and less than or equal to 0.2 for E consecutive quarters, it is determined to be a "plateau period" for the target city. like For any F quarters, the value is greater than 0.2 and less than or equal to 0.8, or If the value is greater than 0.8 in any quarter, the target city is judged as "not reaching its peak"; The confidence level of the extended decoupling model is: ； In the above formula, To expand the confidence level of the decoupling model.

4. The urban carbon peak determination method based on electricity data and fuzzy logic according to claim 1, characterized in that, S3 includes: S31. Input the carbon peak determination results and confidence levels of the dynamic weight model and the extended decoupling model, and convert the carbon peak determination results of the dynamic weight model and the extended decoupling model into three-dimensional state vectors respectively, and fuzzify the confidence levels of the dynamic weight model and the extended decoupling model. S32. Based on the three-dimensional state vector and the fuzzy confidence level, the trigger intensity of each carbon peak determination result for the target city is calculated using the following formula: ； ； ； In the above formula, For the target city in the results The trigger strength, when The time has reached its peak, when During the plateau period, when It was before the peak was reached. , All are weights. Carbon peak determination results for dynamic weighted model The three-dimensional state vector, To extend the carbon peak determination results of the decoupling model The three-dimensional state vector, The confidence level of the dynamic weighting model's determination results corresponds to the membership level of higher-level members. To extend the confidence level of the decoupling model's determination results to higher-level membership; S33. The following formula is used to correct the trigger intensity of the carbon peak determination results for each target city: ； ； ； ； ； ； In the above formula, For the target city in the results Corrected trigger strength , , These represent the confidence levels of the dynamic weighting model's judgment results, and their membership degrees to the low, medium, and high levels, respectively. , , These represent the confidence levels of the extended decoupling model's determination results, and their membership degrees at the low, medium, and high levels, respectively. S34. Based on the trigger intensity of the carbon peak determination results for each target city after correction, obtain the initial output vector: ； In the above formula, This is the initial output vector. The trigger strength after peak correction. This is the trigger strength after the plateau period adjustment. The trigger strength after correction for not reaching the peak; S35. Normalize the initial output vector to obtain the final result of the target city's carbon peak status determination, including: like If so, the target city is determined to have "reached its peak"; like If so, the target city is determined to be in a "plateau period"; like If so, the target city is determined to be "not yet at its peak".

5. The urban carbon peak determination method based on electricity data and fuzzy logic according to claim 1, characterized in that, S1 includes: S11. Calculate the Pearson correlation coefficient between each indicator and carbon emission intensity. Filter out Indicators; S12. Calculate the variance inflation factor of the selected indicators, and retain the indicators with a variance inflation factor of less than 5 to form an indicator system for determining urban carbon peaking.

6. A city carbon peak determination system based on electricity data and fuzzy logic, characterized in that, The system includes an indicator screening module, a model building module, and a comprehensive analysis module; The indicator screening module is used to screen the indicators that have the greatest impact on urban carbon emission intensity from the electricity data. The model building module is used to construct a dynamic weight model and an extended decoupling model based on the selected indicators to determine the carbon peaking status of the city, and to obtain the carbon peaking determination results and confidence levels of the two models. The comprehensive analysis module is used to perform a comprehensive analysis of the carbon peak determination results and confidence levels of the two models using fuzzy logic, so as to obtain the final determination result of the city's carbon peak status.

7. The urban carbon peak determination system based on electricity data and fuzzy logic according to claim 6, characterized in that, The model building module includes a dynamic weight model building unit, a dynamic weight model carbon peak determination unit, and a dynamic weight model confidence calculation unit. The dynamic weight model construction unit is used to construct the following dynamic weight model: ； In the above formula, For the target city Carbon intensity index at any given time This represents the total number of indicators for determining whether a city has reached its carbon peak. For the first Item judgment indicators in Dynamic weights at time points, For the first Item judgment indicators in Standardized value of time, This is an error correction term; The dynamic weighted model carbon peak determination unit is used to obtain the carbon peak determination result of the dynamic weighted model, including: If the observation period for the target city is less than A years, the target city is determined to be "not at its peak". The observation period refers to the period from the start time when continuous historical data can be obtained to the current time. like It has been declining for B consecutive months, and currently... If the value is ≤95% of the historical peak value and shows a significant trend through the Mann-Kendall trend test, then the target city is determined to have "reached its peak". like Within ±5% of the historical peak, a state showing no significant trend according to the Mann-Kendall test persists for more than C months, and within this period... If the standard deviation or mean is ≤3%, it is considered a "plateau period" in the target city. If the current >105% of historical peak, or If a city shows a clear and continuous upward trend, or does not meet the above criteria for "peak reached" or "plateau period", then the target city is determined to be "not yet at its peak". The dynamic weight model confidence calculation unit is used to calculate the confidence of the dynamic weight model using the following formula: ； ； ； In the above formula, The confidence level of the dynamic weighting analysis model. For the trend test results, This is the statistic for the Mann-Kendall trend test. To examine the total number of periods, , Each represents a different moment. , The target cities are respectively time, Carbon intensity index at any given time.

8. The urban carbon peak determination system based on electricity data and fuzzy logic according to claim 6, characterized in that, The model building module also includes an extended decoupling model building unit, an extended decoupling model carbon peak determination unit, and an extended decoupling model confidence calculation unit. The extended decoupling model construction unit is used to construct the following extended decoupling model: ； ； ； ； ； ； In the above formula, For the target city The decoupling index at any given moment. for Changes in urban carbon emission intensity over time for Changes in urban energy efficiency over time. for Changes in urban electricity productivity at any given time. for Changes in urban electricity consumption at any given time. The carbon emissions of the target city at the baseline time. The electricity consumption of the target city at the baseline time. The logarithmic average weighting factor is... for Urban carbon emission intensity at any given time The carbon emission intensity of the target city at the baseline time. for Urban energy efficiency at all times The energy efficiency of the target city at the baseline time. for Urban electricity productivity at any given time The electricity productivity of the target city at the baseline time. for Urban electricity consumption at any given time for Urban carbon emissions at any given moment; The extended decoupling model carbon peak determination unit is used to obtain the carbon peak determination result of the extended decoupling model, including: like If the value is less than -0.3 for D consecutive quarters, the target city is determined to have "peaked". like If the value is greater than or equal to -0.3 and less than or equal to 0.2 for E consecutive quarters, it is determined to be a "plateau period" for the target city. like For any F quarters, the value is greater than 0.2 and less than or equal to 0.8, or If the value is greater than 0.8 in any quarter, the target city is judged as "not reaching its peak"; The extended decoupling model confidence calculation unit is used to calculate the confidence of the extended decoupling model using the following formula: ； In the above formula, To expand the confidence level of the decoupling model.

9. The urban carbon peak determination system based on electricity data and fuzzy logic according to claim 6, characterized in that, The comprehensive analysis module includes an input preprocessing unit, a trigger intensity calculation unit, a trigger intensity correction unit, an initial output vector formation unit, and a judgment result normalization unit. The input preprocessing unit is used to input the carbon peak determination results and confidence levels of the dynamic weight model and the extended decoupling model, and converts the carbon peak determination results of the dynamic weight model and the extended decoupling model into three-dimensional state vectors, and fuzzifies the confidence levels of the dynamic weight model and the extended decoupling model. The trigger intensity calculation unit is used to calculate the trigger intensity of each carbon peak determination result of the target city based on the three-dimensional state vector and the fuzzified confidence level, using the following formula: ； ； ； In the above formula, For the target city in the results The trigger strength, when The time has reached its peak, when During the plateau period, when It was before the peak was reached. , All are weights. Carbon peak determination results for dynamic weighted model The three-dimensional state vector, To extend the carbon peak determination results of the decoupling model The three-dimensional state vector, The confidence level of the dynamic weighting model's determination results corresponds to the membership level of higher-level members. To extend the confidence level of the decoupling model's determination results to higher-level membership; The trigger intensity correction unit is used to correct the trigger intensity of each carbon peak determination result for the target city using the following formula: ； ； ； ； ； ； In the above formula, For the target city in the results Corrected trigger strength , , These represent the confidence levels of the dynamic weighting model's judgment results, and their membership degrees to the low, medium, and high levels, respectively. , , These represent the confidence levels of the extended decoupling model's determination results, and their membership degrees at the low, medium, and high levels, respectively. The initial output vector forming unit is used to obtain the initial output vector based on the trigger intensity of the corrected carbon peak determination results for each target city: ； In the above formula, This is the initial output vector. The trigger strength after peak correction. This is the trigger strength after the plateau period adjustment. The trigger strength after correction for not reaching the peak; The result normalization unit is used to normalize the initial output vector to obtain the final result of the target city's carbon peak status, including: like If so, the target city is determined to have "reached its peak"; like If so, the target city is determined to be in a "plateau period"; like If so, the target city is determined to be "not yet at its peak".

10. The urban carbon peak determination system based on electricity data and fuzzy logic according to claim 6, characterized in that, The indicator screening module includes a Pearson indicator judgment unit and a judgment indicator system formation unit; The Pearson index judgment unit is used to calculate the Pearson correlation coefficient between each index and carbon emission intensity. Filter out Indicators; The judgment index system forming unit is used to calculate the variance inflation factor of the selected indicators, and retain the indicators with a variance inflation factor of less than 5 to form the urban carbon peak judgment index system.

Images