Method, system and medium for constructing carbon peak and carbon neutral path evaluation model

Abstract

The present application relates to the technical field of model construction, and more particularly to a method and system for constructing a carbon peak and carbon neutral path evaluation model and a medium. The method comprises the following steps: obtaining regional energy consumption, industrial structure and carbon emission source monitoring data, constructing a basic data set, and forming a source-by-source carbon emission inventory through item-by-item measurement; correcting the inventory based on an emission factor library to obtain standard carbon emission data; constructing a parameter matrix including energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and performing coupled operation to generate multi-scenario carbon emission evolution results; determining the carbon peak time node and carbon neutral path through time series fitting analysis, and finally constructing a path evaluation model and outputting the evaluation results. Through source-by-source correction, disturbance modulation and multi-dimensional parameter coupling analysis, the present application realizes multi-scenario accurate simulation and path optimization evaluation of carbon emission evolution, thereby significantly improving the accuracy of carbon peak and carbon neutral path prediction and decision support capability.

Description

Technical Field

[0001] This invention relates to the field of model building technology, and in particular to a method, system and medium for constructing a carbon peaking and carbon neutrality pathway evaluation model. Background Technology

[0002] The goals of carbon peaking and carbon neutrality place higher demands on regional energy structure optimization and industrial transformation and upgrading. Current carbon emission management mainly relies on statistical accounting methods, estimating regional carbon emission levels by analyzing total energy consumption and emission factors. However, these methods are largely based on macro-level aggregated data, lacking refined decomposition and dynamic tracking of emission sources, making it difficult to reflect the changing characteristics of different emission entities over time, resulting in limited accuracy.

[0003] Meanwhile, in terms of path assessment, existing technologies typically employ single-scenario analysis or simple parameter adjustments to predict carbon emission evolution trends. However, they often fail to fully consider the coupling relationships between multiple factors such as energy substitution, industrial restructuring, and the penetration of emission reduction technologies. They lack multi-dimensional parameter collaborative analysis mechanisms, making it difficult to construct evolutionary models with multi-scenario comparison capabilities. Furthermore, traditional models often process carbon emission time series through smoothing or linear fitting, failing to effectively characterize the stage fluctuations and abrupt changes in the emission process, resulting in inaccurate identification of carbon peak time points. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, system, and medium for constructing a carbon peaking and carbon neutrality pathway evaluation model to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a method for constructing an evaluation model for carbon peaking and carbon neutrality pathways is provided, the method comprising the following steps: Step S1: Obtain the basic dataset, which includes regional energy consumption data, industrial structure data, and carbon emission source monitoring data; Step S2: Analyze the basic dataset for itemized measurement to construct a source-specific carbon emission inventory; Step S3: Correct the source carbon emission inventory based on the preset emission factor library to obtain standard carbon emission data; Step S4: Establish a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and input standardized carbon emission data into the parameter matrix for coupling calculation to generate multi-scenario carbon emission evolution data; Step S5: Fit and analyze the time series of carbon emission evolution data under multiple scenarios to determine the carbon peak time point and the path to achieve carbon neutrality; Step S6: Construct a path evaluation model based on the carbon peak time node and the carbon neutrality achievement path, and output the corresponding path evaluation results.

[0006] This invention provides a system for constructing a carbon peaking and carbon neutrality pathway evaluation model, used to execute the method described above for constructing such a model. The system for constructing the carbon peaking and carbon neutrality pathway evaluation model includes: The data acquisition module is used to acquire the basic dataset, which includes regional energy consumption data, industrial structure data, and carbon emission source monitoring data. The inventory building module is used for the basic dataset of itemized measurements to build a source-specific carbon emission inventory; The emission correction module is used to correct the source carbon emission inventory based on a preset emission factor library to obtain standard carbon emission data. The scenario generation module is used to establish a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and input standardized carbon emission data into the parameter matrix for coupled calculation to generate multi-scenario carbon emission evolution data. The path identification module is used to fit and analyze the time series of carbon emission evolution data under multiple scenarios in order to determine the carbon peak time node and the path to achieve carbon neutrality. The evaluation output module is used to construct a path evaluation model based on the carbon peak time node and the carbon neutrality achievement path, and output the corresponding path evaluation results.

[0007] The present invention also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above for constructing a carbon peaking and carbon neutrality pathway evaluation model.

[0008] The beneficial effects of this invention lie in achieving high-precision standardized processing of carbon emission data by performing source-specific measurement and emission factor correction on regional energy consumption, industrial structure, and carbon emission source data. Based on this, a carbon emission segment rearrangement and amplitude dynamic modulation mechanism based on path perturbation weights is introduced, combined with adaptive smoothing processing of transition windows, effectively characterizing the stage-specific fluctuations and structural changes in the carbon emission process. Simultaneously, a multi-dimensional parameter grid is constructed and a collaborative constraint relationship between parameters is introduced to achieve coupled analysis and reasonable screening of energy substitution rate, industrial adjustment coefficient, and emission reduction technology penetration rate, generating carbon emission evolution data with multi-scenario comparison characteristics. Furthermore, time series fitting accurately identifies the carbon peak time node and carbon neutrality achievement path, thereby significantly improving the constructed path evaluation model in terms of dynamic simulation capability, scenario coverage, result reliability, and decision guidance, providing more targeted and operable technical support for optimizing regional low-carbon development paths. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating the steps involved in constructing an evaluation model for carbon peaking and carbon neutrality pathways. Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S4. Figure 3 This is a multi-scenario carbon emission evolution surface diagram of a method for constructing a carbon peaking and carbon neutrality pathway evaluation model according to this application; Figure 4 This is a scaled diagram illustrating a method for constructing a carbon peaking and carbon neutrality pathway evaluation model according to this application. The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0010] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0011] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0012] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0013] To achieve the above objectives, please refer to Figures 1 to 4 A method for constructing an evaluation model for carbon peaking and carbon neutrality pathways, the method comprising the following steps: Step S1: Obtain the basic dataset, which includes regional energy consumption data, industrial structure data, and carbon emission source monitoring data; In this embodiment of the invention, centralized collection of multi-source data can be achieved by establishing data interfaces with regional energy management platforms, statistical information systems, and online monitoring equipment. Specifically, regional energy consumption data can be obtained from power dispatching systems or energy metering systems, including time-period data on electricity, gas, and other energy consumption; industrial structure data can be extracted from regional statistical databases, including information such as industry categories, output value share, and enterprise distribution; and carbon emission source monitoring data is obtained through online monitoring equipment installed by key emission units, including data such as flue gas emission concentration, emission flow rate, and corresponding time stamps.

[0014] During data acquisition, to ensure data consistency over time, a unified time collection period can be set for data from different sources, such as summarizing by hour or day, and data with time discrepancies can be aligned. Simultaneously, the collected raw data should be preliminarily screened to remove obvious outliers or severely missing data records; for example, data with consecutive missing time periods can be marked or supplemented.

[0015] After completing the multi-source data collection, regional energy consumption data, industrial structure data, and carbon emission source monitoring data are organized and stored according to a unified data structure. Specifically, a field-based storage method can be adopted, assigning corresponding data identifiers to different types of data and establishing relationships between data, such as linking various types of data through time identifiers and regional codes. The final result is a basic dataset with a unified structure, clear sources, and time-series characteristics.

[0016] Step S2: Analyze the basic dataset for itemized measurement to construct a source-specific carbon emission inventory; In this embodiment of the invention, the basic data is classified according to energy type and emission source. Regional energy consumption data is divided into categories such as electricity consumption, gas consumption, coal consumption, and other energy consumption. Simultaneously, carbon emission source monitoring data is classified according to the emission subject into types such as industrial emission sources, transportation emission sources, and residential emission sources. For industrial structure data, it is mapped to various emission source categories according to industry classification standards.

[0017] After data classification, energy consumption data for each type is measured and calculated separately. Specifically, the raw energy consumption data is converted to standard energy quantities based on the measurement units corresponding to different energy types. For example, electricity consumption is converted to standard coal equivalent, and gas and coal consumption are converted according to their calorific value. For key emission sources with online monitoring capabilities, carbon emissions for the corresponding time period are calculated directly using the emission flow rate and emission concentration from the monitoring data. For emission sources without monitoring equipment, carbon emissions are indirectly calculated based on energy consumption data and the industry average emission factor to obtain the corresponding carbon emission data for each emission source.

[0018] After obtaining the carbon emissions from various emission sources, the emission data is collected and organized according to preset source-specific statistical rules. In practice, the data can be summarized according to a multi-level structure of "energy type - industry category - emission subject," classifying and statistically analyzing carbon emissions for each time period, and generating corresponding sub-emission records. Each record must include at least the emission source type, industry, time period, and corresponding carbon emission value.

[0019] The above-mentioned sub-item measurement results are integrated in a unified format to form a source-specific carbon emission inventory.

[0020] Step S3: Correct the source carbon emission inventory based on the preset emission factor library to obtain standard carbon emission data; In this embodiment of the invention, after obtaining the carbon emission inventory by source, it undergoes unified correction processing to eliminate deviations caused by different data sources and measurement methods, resulting in more comparable standard carbon emission data. Specifically, a pre-organized emission factor library is first established or invoked. This emission factor library stores corresponding carbon emission factor information according to energy type, industry category, and emission process, along with information such as applicable scope, time version, and source identifier, to facilitate matching and selection in different scenarios.

[0021] During the calibration process, emission records in the source-specific carbon emission inventory are read one by one. Based on the energy type, industry category, and emission entity information, the corresponding emission factors are retrieved from the emission factor database. For records that can be directly matched, the data with the same category and the closest time version is selected as the calibration basis. For cases where multiple emission factors are available, the emission factors are screened or weighted based on the process characteristics or technological level reflected in the industrial structure data to improve matching accuracy.

[0022] After completing the emission factor matching, the carbon emissions from different sources are corrected. In practice, the emissions obtained from the original sub-item measurement are recalculated or corrected with the selected emission factors. For example, emissions obtained by indirect estimation are corrected to ensure consistency with the unified emission factor system. For emission data obtained through online monitoring, the rationality of the data can be verified in conjunction with the emission factors, and data that deviates from the normal range is corrected or marked.

[0023] Furthermore, consistency processing can be applied to data from different time periods during the calibration process. For example, historical data can be retrospectively corrected using a unified version of the emission factor, or data with significant fluctuations can be smoothed to reduce the impact of changes in statistical methods. For missing or abnormal data records, data from adjacent time periods or similar emission sources can be combined to supplement or replace them, thereby ensuring data continuity.

[0024] Finally, the emission data, after being corrected for emission factors, are re-aggregated and organized to generate standard carbon emission data.

[0025] Step S4: Establish a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and input standardized carbon emission data into the parameter matrix for coupling calculation to generate multi-scenario carbon emission evolution data; In this embodiment of the invention, a parameter matrix is ​​constructed around the factors influencing carbon emissions. This parameter matrix is ​​based on the time dimension, and at each time point, parameters such as energy substitution rate, industrial adjustment coefficient, and emission reduction technology penetration rate are set. The energy substitution rate characterizes the proportion of traditional high-carbon energy being converted to clean energy; the industrial adjustment coefficient reflects the degree of optimization of the industrial structure within a region; and the emission reduction technology penetration rate describes the application coverage of low-carbon technologies in various industries. Each parameter can be set to different values ​​at different time stages to form multiple parameter combinations.

[0026] After the parameter matrix is ​​established, standard carbon emission data is imported into it in chronological order and mapped according to emission sources and industry categories. In practice, standard carbon emission data can be used as a baseline emission level, and at each time point, it can be associated with the corresponding energy type and industry category, ensuring a one-to-one correspondence with each parameter in the parameter matrix. This method establishes a correlation between basic emission data and influencing factors.

[0027] Subsequently, standard carbon emission data is coupled and processed based on different parameter combinations. Specifically, for each parameter combination, the standard carbon emission data can be adjusted separately. For example, emissions corresponding to high-carbon energy sources can be reduced based on the energy substitution rate, while compensation is made for the portion replaced by clean energy; the emission ratios of high-emission and low-emission industries can be structurally adjusted based on the industry adjustment coefficient; and emissions from each industry can be reduced in stages based on the penetration rate of emission reduction technologies. All of these adjustments are executed concurrently at the same time point to form the carbon emission result under that parameter combination.

[0028] After completing the coupling process for a single parameter combination, the above process is repeated for different parameter combinations to gradually generate multiple sets of carbon emission change results. Each set of results corresponds to a specific development scenario, such as a high substitution rate scenario, an industry optimization scenario, or a technology-driven scenario. Finally, the results are organized in chronological order to form multi-scenario carbon emission evolution data.

[0029] Step S5: Fit and analyze the time series of carbon emission evolution data under multiple scenarios to determine the carbon peak time point and the path to achieve carbon neutrality; In this embodiment of the invention, carbon emission evolution data are classified and organized according to different scenario categories, and the annual or quarterly carbon emission data under each scenario are arranged in chronological order to form a corresponding time series dataset.

[0030] In one embodiment, an appropriate fitting method can be selected based on the characteristics of carbon emission changes. For example, piecewise trend fitting can be used for data showing a continuous increase followed by a slowdown, while curve fitting can be used for data with obvious inflection point characteristics. Through fitting, a continuously changing carbon emission trend curve can be obtained, thus more clearly reflecting the changing patterns of emission growth, plateau period, and decline phases.

[0031] Based on the fitting results, peak values ​​of carbon emission trends under various scenarios are identified. In practice, the inflection point where carbon emissions change from rising to falling is determined by comparing the fitted curves over time periods, and the time node corresponding to this point is designated as the carbon peak time node. Simultaneously, the rate and duration of emission decline are analyzed in conjunction with the subsequent downward trend to assess the effectiveness and stability of emission reduction measures.

[0032] Furthermore, based on identifying the carbon peak time point, a comprehensive analysis of emission evolution paths under different scenarios is conducted. In practice, the changes in peak time, peak level, and subsequent decline phases of each scenario can be compared to screen out the scenario path that can achieve both early peaking and a sustained decline trend. For the preferred path, its corresponding energy substitution level, degree of industrial restructuring, and application of emission reduction technologies can be further analyzed to clarify the key implementation path for achieving carbon neutrality.

[0033] Finally, the analysis results of each scenario are compiled and the peak carbon time points and corresponding carbon neutrality achievement paths are output.

[0034] Step S6: Construct a path evaluation model based on the carbon peak time node and the carbon neutrality achievement path, and output the corresponding path evaluation results.

[0035] In this embodiment of the invention, after determining the carbon peak time point and the carbon neutrality achievement path, different paths are quantitatively evaluated to select the optimal or preferred development scheme. Specifically, an indicator system for the path evaluation model is first constructed. This indicator system can set evaluation indicators around dimensions such as emission reduction effect, economic impact, and implementation feasibility. For example, it includes indicators such as the size of the carbon emission peak, the degree of advancement in the peak time, the carbon neutrality achievement cycle, energy substitution costs, the degree of impact on industrial adjustment, and the difficulty of implementing emission reduction technologies. Each indicator is assigned a corresponding weight to reflect the importance of different factors in the comprehensive evaluation.

[0036] After the indicator system is established, the carbon peak time nodes and carbon neutrality achievement path parameters corresponding to each scenario are input into the path evaluation model. In specific implementation, the corresponding indicator values ​​can be extracted based on the performance of each path on each evaluation indicator. For example, peak emissions and decline rates can be obtained from multi-scenario carbon emission evolution data, and information such as energy substitution rate and technology penetration level can be extracted from the parameter matrix. The above data are then uniformly mapped into the evaluation indicator system.

[0037] Subsequently, the evaluation indicators are standardized to eliminate the influence between different units of measurement. Specifically, positive and negative indicators are normalized separately to ensure their values ​​fall within a uniform range, thus achieving comparability. After standardization, the indicators are weighted and aggregated according to preset weights to obtain the comprehensive evaluation value for each carbon neutrality pathway.

[0038] After obtaining the comprehensive evaluation value, the different paths are ranked and analyzed. In practice, the paths can be ranked according to the order of their comprehensive evaluation values, and the best-performing path can be identified. Furthermore, a grading system can be established to classify the paths into different levels, such as excellent, good, average, or requiring optimization, to facilitate intuitive judgment by decision-makers.

[0039] Furthermore, to enhance the reliability of the evaluation results, sensitivity analysis can be performed on key parameters. For example, small-scale disturbances can be made to the energy substitution rate or technology penetration rate to observe the changes in the evaluation results, thereby judging the stability and risk resistance of the path.

[0040] Finally, the evaluation results are output to form the corresponding path evaluation result data.

[0041] As an example of the present invention, reference is made to Figure 2 As shown, step S4 in this example includes: Step S41: Read the time series corresponding to each emission source in the standardized carbon emission data, divide it into several carbon emission segments according to the continuity of time, and record the start and end time and emission change of each segment; Step S42: Calculate the change gradient of each carbon emission segment based on the emission difference between adjacent time points, and determine the corresponding path perturbation weight factor based on the magnitude of the change gradient; Step S43: Based on the distribution of path perturbation weight factors, rearrange the positions and modulate the amplitudes of each carbon emission segment to generate carbon emission perturbation sequence data; Step S44: Construct a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and generate a parameter combination set according to different parameter value ranges; Step S45: Map the carbon emission perturbation sequence data to the parameter combinations in the parameter matrix, and perform weighted coupling calculations accordingly to obtain multiple sets of carbon emission evolution results; Step S46: Summarize the carbon emission evolution results to obtain multi-scenario carbon emission evolution data.

[0042] In one embodiment, taking the annual carbon emission data of a certain city with a statistical period from 2020 to 2030 as an example, the data source is the standard carbon emission data formed in the aforementioned steps, wherein industrial coal-fired emission sources are selected as examples, and their annual emissions are recorded sequentially as 100, 108, 115, 123, 130, 128, 120, 110, 95, 80, and 60 (unit: 10,000 tons), and each data is marked with the corresponding year.

[0043] In actual processing, the above data is written into a contiguous storage area in chronological order, and the emission difference between adjacent years is calculated year by year. After the difference calculation is completed, segments are divided according to the principle of "consistent and continuous difference sign": consecutive years with positive differences are classified as the same segment, and the point where the difference changes from positive to negative or vice versa is used as the dividing point. According to this rule, in this example, 2020 to 2024 is classified as the first segment (continuous positive growth), 2024 to 2026 as the second segment (a slight decrease after the difference changes from positive to negative), and 2026 to 2030 as the third segment (continuous decrease with gradually increasing magnitude). Subsequently, the start year, end year, and total change are recorded for each segment. For example, the first segment is recorded as "2020–2024, change +30".

[0044] The difference in emissions between two consecutive years is directly used as the gradient value, and its absolute value is used for weight calculation. For ease of subsequent processing, the maximum value among all gradient values ​​is used as the benchmark; for example, in this case, the maximum absolute difference is 20 (corresponding to 2029 to 2030). Each gradient value is then divided by this maximum value to obtain a weight value between 0 and 1; for example, a year with a difference of 8 has a weight of approximately 0.4, and a year with a difference of 20 has a weight of 1.0. This weight serves as a path perturbation weight factor and is stored in a way that is uniquely linked to the corresponding year.

[0045] The original sequence is perturbed according to the aforementioned weights. In practice, not all data is adjusted uniformly, but rather processed hierarchically according to weight: when the weight is greater than 0.7, the change in the corresponding year is amplified, for example, multiplied by 1.1 to 1.2; when the weight is between 0.3 and 0.7, the original value is maintained or fine-tuned within ±1; when the weight is less than 0.3, no adjustment is made. Simultaneously, without disrupting the chronological order, the positions of individual high-weighted years can be shifted forward by one position (i.e., the downward trend is advanced by one year) to simulate early policy intervention. After processing according to the above rules, a perturbed sequence is obtained in this example, for example, adjusted to 100, 107, 116, 125, 132, 125, 112, 98, 85, 70, 55, and this sequence is saved as a carbon emission perturbation sequence.

[0046] You can refer to this. Figure 3 The program sets three parameter arrays: energy substitution rate (0.2, 0.4, 0.6), industrial adjustment coefficient (0.9, 0.8, 0.7), and emission reduction technology penetration rate (0.3, 0.5, 0.7). All combinations are generated through a three-level loop, resulting in 27 sets of parameters. Each set of parameters is assigned a unique number, and a corresponding relationship is established with subsequent calculation results.

[0047] The perturbation sequence is substituted into the above parameter combinations one by one for calculation. During the calculation, the data for each year is processed in a fixed order: first, emissions are subtracted based on the energy substitution rate (e.g., multiplied by (1 - substitution rate × 0.25)); then, structural adjustments are made based on the industrial adjustment coefficient (directly multiplied by this coefficient); finally, emissions are reduced based on the technology penetration rate (multiplied by (1 - penetration rate × 0.2)). For example, for an emission value of 125 in a certain year, with parameters of 0.4, 0.8, and 0.5, the calculation process is: 125 × (1 - 0.4 × 0.25) = 112.5, then × 0.8 = 90, and finally × (1 - 0.5 × 0.2) = 81. This calculation is repeated for all years in the sequence to obtain a complete evolution sequence under this parameter combination. All 27 sets of parameters are processed in the same way to generate multiple sets of carbon emission evolution results.

[0048] The perturbation sequence is substituted into the above parameter combinations one by one for calculation. During the calculation, the data for each year is processed in a fixed order: first, emissions are subtracted based on the energy substitution rate (e.g., multiplied by (1 - substitution rate × 0.25)); then, structural adjustments are made based on the industrial adjustment coefficient (directly multiplied by this coefficient); finally, emissions are reduced based on the technology penetration rate (multiplied by (1 - penetration rate × 0.2)). For example, for an emission value of 125 in a certain year, with parameters of 0.4, 0.8, and 0.5, the calculation process is: 125 × (1 - 0.4 × 0.25) = 112.5, then × 0.8 = 90, and finally × (1 - 0.5 × 0.2) = 81. This calculation is repeated for all years in the sequence to obtain a complete evolution sequence under this parameter combination. All 27 sets of parameters are processed in the same way to generate corresponding results, i.e., multi-scenario carbon emission evolution data.

[0049] Preferably, step S43 includes: The carbon emission segments are sorted according to the magnitude of the path perturbation weight factor to determine the priority sequence of high-weight segments and low-weight segments; Based on the priority sequence, adjust the arrangement of each carbon emission segment on the time axis so that high-weight segments are shifted to key time intervals; Based on the correspondence between the path disturbance weight factor and the preset modulation coefficient, the emission amplitude of each carbon emission segment is amplified or compressed. The carbon emission segments, after position adjustment and amplitude modulation, are reassembled in chronological order to generate a continuous carbon emission perturbation sequence data.

[0050] In one embodiment, step S43 is specifically processed based on the three carbon emission segments already obtained above. It is assumed that the path perturbation weight factors corresponding to each segment have been obtained in step S42, wherein the weight of the first segment (2020–2024) is 0.35, the weight of the second segment (2024–2026) is 0.55, and the weight of the third segment (2026–2030) is 0.85.

[0051] First, the carbon emission segments are sorted according to the magnitude of the path perturbation weight factor. Specifically, the weight values ​​of each segment are written into an array and sorted from largest to smallest, resulting in a priority sequence: third segment, second segment, and first segment. This sorting result serves as the basis for subsequent adjustments, with segments having larger weights identified as critical segments that significantly impact the overall emission path.

[0052] Subsequently, the positions of each carbon emission segment on the timeline are adjusted according to the aforementioned priority sequence. In practice, the original timeline is used as a baseline, and the critical time interval is selected as "two years before and after peak emission" (e.g., 2023–2026). The third segment, with the highest weight, is shifted forward by one year, changing its starting time from 2026 to 2025; the second segment remains in its original position; and the first segment is sequentially connected according to the remaining time interval, ensuring no time overlap between segments and continuity from beginning to end. During the adjustment process, only the time identifier corresponding to the segment is changed, without altering the data order within the segment.

[0053] Next, based on the correspondence between the path disturbance weighting factor and the preset modulation coefficient, the emission amplitude of each carbon emission segment is modulated. The specific modulation rules are as follows: when the weight is ≥0.8, the modulation coefficient is 1.15; when the weight is between 0.5 and 0.8, the modulation coefficient is 1.05; when the weight is <0.5, the modulation coefficient is 0.95. According to this rule, the annual emission changes in the third segment are multiplied by 1.15 for amplification, the second segment is multiplied by 1.05 for slight amplification, and the first segment is multiplied by 0.95 for compression. For example, in the third segment, the original annual decrease was -10, which is adjusted to -11.5 after processing.

[0054] Finally, the carbon emission segments, after position adjustment and amplitude modulation, are reassembled according to the new time order. During the assembly process, segments are written to the output sequence buffer sequentially, and the continuity of adjacent segments is checked. If a jump exceeds a preset threshold (e.g., ±10), a transition value is inserted at the connection point for smoothing. After all segments are assembled, a new continuous carbon emission sequence is generated, for example, the adjusted result is: 100, 106, 114, 122, 129, 121, 108, 95, 82, 68, 54.

[0055] Preferably, based on the correspondence between the path disturbance weighting factor and the preset modulation coefficient, the emission amplitude of each carbon emission segment is amplified or compressed as follows: The path disturbance weight factor is divided into several weight level intervals, and a corresponding basic modulation coefficient is preset for each weight level interval. Based on the position of each carbon emission segment in the time series, the stage attribute to which it belongs is determined, and the basic modulation coefficient is modified in stages to obtain the dynamic modulation coefficient. Based on the dynamic modulation coefficient, the emission amplitude of the corresponding carbon emission segment is scaled proportionally.

[0056] In one embodiment, the specific implementation process of step S43, "amplifying or compressing the emission amplitude of each carbon emission segment according to the correspondence between the path perturbation weight factor and the preset modulation coefficient," will be described. Taking the three carbon emission segments obtained above as an example, it is assumed that their path perturbation weight factors are 0.35, 0.55, and 0.85, respectively.

[0057] First, the path perturbation weighting factor is divided into multiple weighting level intervals. In actual processing, the grading rules can be pre-defined in the program, for example: weights less than 0.4 are classified as low-weight intervals, 0.4 to 0.7 as medium-weight intervals, and weights greater than or equal to 0.7 as high-weight intervals. Simultaneously, a basic modulation coefficient is set for each interval, for example, 0.9 for low weight, 1.05 for medium weight, and 1.15 for high weight. According to this rule, in this example, the first segment corresponds to the low-weight interval, the second segment to the medium-weight interval, and the third segment to the high-weight interval.

[0058] Subsequently, the stage attribute of each carbon emission segment is determined based on its position in the time series. In practice, the time series can be divided into "rising stage," "peak transition stage," and "declining stage" according to emission trends. For example, 2020 to 2023 can be defined as the rising stage, 2023 to 2025 as the peak transition stage, and the period after 2025 as the declining stage. Combining the time range of the segments, it can be seen that the first segment is mainly in the rising stage, the second segment is in the peak transition stage, and the third segment is in the declining stage.

[0059] After determining the stage attributes, the base modulation coefficient is adjusted in stages. Specifically, the following rules can be set: For the rising stage, to avoid excessively amplifying emission growth, the base modulation coefficient is moderately lowered, for example, reduced by 0.05; for the peak transition stage, the base modulation coefficient remains unchanged; for the falling stage, to strengthen the emission reduction trend, the base modulation coefficient is increased, for example, increased by 0.05. After processing according to these rules, in this example, the dynamic modulation coefficient of the first segment is corrected from 0.9 to 0.85, the second segment remains unchanged at 1.05, and the third segment is corrected from 1.15 to 1.20.

[0060] Finally, the emission amplitudes of each carbon emission segment are proportionally scaled based on the obtained dynamic modulation coefficients. In practice, instead of scaling the absolute emission values ​​as a whole, the changes in adjacent years within a segment are adjusted. For example, in the first segment, the original increase of +8 is adjusted to approximately +7 after a coefficient of 0.85; in the second segment, the original change of -2 is adjusted to approximately -2.1 after a coefficient of 1.05; and in the third segment, the original change of -15 is adjusted to approximately -18 after a coefficient of 1.20. After processing all the changes within a segment in sequence, the adjusted changes are accumulated year by year to obtain a new emission value sequence.

[0061] Preferably, scaling the emission amplitude of the corresponding carbon emission segment according to the dynamic modulation coefficient further includes: Select a transition window of a preset length at the junction of adjacent carbon emission segments, and extract the tail feature value of the previous segment and the head feature value of the next segment within the window. Using the tail feature value as the reference amplitude and the head feature value as the target amplitude, a progressive mapping sequence is constructed within the transition window, gradually shifting from the reference amplitude to the target amplitude. Based on the relative distance between each position within the transition window and the boundary point, a corresponding transition ratio coefficient is assigned, and the scaling result is locally replaced according to the ratio coefficient, so that the amplitude change is a unidirectional gradual shift. After completing the local replacement, the data outside the transition window boundary is kept at the original scaling ratio to form a local adaptive transition structure.

[0062] In one embodiment, reference Figure 4 After scaling each carbon emission segment based on the dynamic modulation coefficient, a transition process is applied to the segment boundaries to avoid abrupt changes at the splicing points of adjacent segments. The following example illustrates this process using the boundary between the second segment (2024–2026) and the third segment (2026–2030).

[0063] First, a transition window of a preset length is selected at the boundary between adjacent carbon emission segments. In actual implementation, the window length can be set to 2 to 3 time steps; in this example, a length of 2 is used, selecting 2025 and 2026 as the transition window. 2025 represents the tail region of the previous segment, and 2026 represents the head region of the next segment. Then, the corresponding feature values ​​are extracted. For example, after scaling, the emission value for 2025 is 121, and the emission value for 2026 is 112. Therefore, 121 is used as the tail feature value, and 112 is used as the head feature value.

[0064] Secondly, using the tail feature value as the baseline amplitude and the head feature value as the target amplitude, a progressive mapping sequence is constructed within the transition window. In practice, linear interpolation can be used to generate intermediate transition values. For example, for a window of length 2, the transition can be directly from 121 to 112; if the window length is 3, a progressive sequence similar to 121, 116.5, and 112 can be generated, making the values ​​transition smoothly.

[0065] Subsequently, transition ratio coefficients are assigned based on the relative distances between each position within the transition window and the boundary point. In practice, the positions within the window can be normalized by distance; for example, for a window of length 2, the ratio coefficient closer to the previous segment is 0.3, and the ratio coefficient closer to the next segment is 0.7. Then, the scaled values ​​are locally replaced according to the ratio coefficients, i.e., updated in the manner of "new value = original value × (1 - ratio) + transition value × ratio," thereby causing the amplitude to gradually shift from the baseline value to the target value in one direction. For example, a slight adjustment is made to the 2025 data, and a larger adjustment is made to the 2026 data, making the overall change smoother.

[0066] Finally, after replacing the data within the transition window, the data outside the window boundaries are scaled to their original scale without further adjustments. This method creates a local adaptive transition structure only at the segment boundaries, without affecting the existing trends within each segment.

[0067] Preferably, based on the relative distance between each position within the transition window and the boundary point, a corresponding transition ratio coefficient is assigned, and the scaling result is locally replaced according to the ratio coefficient, so that the amplitude change shows a unidirectional gradual shift, including: Determine the distance values ​​from each location point within the transition window to the boundary point, and mark them in order of distance from near to far; Construct a monotonically changing proportion sequence based on the distance value, so that the position closer to the boundary point corresponds to a larger transition proportion, and the position farther away from the boundary point corresponds to a smaller transition proportion; Using the scaled amplitude as the baseline, and introducing a target transition amplitude, the amplitudes at each position point are weighted and replaced according to the proportion sequence, so that the amplitudes at each position point gradually shift towards the target amplitude according to the proportion. Perform sequential consistency correction on the replaced position points to ensure that the amplitude changes of adjacent position points maintain the same trend. After the replacement is completed, boundary constraints are applied to the overall amplitude change within the transition window.

[0068] In one embodiment, after scaling the amplitude of carbon emission segments, to ensure smooth changes at the boundaries of adjacent segments, the data within the transition window are assigned a proportion coefficient based on distance and subjected to weighted replacement processing. The following explanation uses a segment boundary as an example.

[0069] First, determine the range of the transition window. Assume we select one time step before and after the boundary point as the window, containing three locations: t-1, t (the boundary point), and t+1. Calculate the distance from each location within the window to the boundary point t; for example, the distance to location t is 0, and the distance between locations t-1 and t+1 is 1. Then, label the locations sequentially from closest to furthest, marking t as location 1, and t-1 and t+1 as location 2.

[0070] Secondly, a monotonically changing transitional proportion sequence is constructed based on the aforementioned distance values. In practical implementation, the maximum proportion can be preset to 0.8 and the minimum proportion to 0.3, and the proportions can be allocated according to the distance size, so that the positions closer to the boundary points have a larger proportion. In this example, the proportion of position t can be set to 0.8, and the proportions of positions t-1 and t+1 can be set to 0.4 respectively, thus forming a proportion structure that decreases from the center to both sides.

[0071] Subsequently, the scaled amplitude is used as the baseline value, and a target transition amplitude is introduced for weighted replacement. For example, if the amplitude at the end of the segment before the boundary point is 121 and the amplitude at the beginning of the segment after the boundary point is 112, then 112 can be used as the target transition amplitude. The replacement is performed on each point within the window according to the formula: "New amplitude = benchmark value × (1 - percentage) + target amplitude × percentage".

[0072] For example, if the original value at position t is 118, then after replacement it becomes 118×(1-0.8)+112×0.8≈113; the original value at position t-1 is 121, then after replacement it becomes approximately 121×0.6+112×0.4≈117; the same applies to position t+1. In this way, the amplitude at each position gradually shifts towards the target value.

[0073] After completing the weighted replacement, the sequential consistency of each position point within the window is corrected. In practice, the direction of change between adjacent position points is checked sequentially. If a local reversal occurs (e.g., the previous time step is less than the next time step, but the overall trend should be downward), the position point is finely adjusted (e.g., adjusted within ±1) to ensure that the change within the entire window remains monotonically decreasing or increasing, thereby avoiding local fluctuations.

[0074] Finally, after the replacement is completed, boundary constraints are applied to the overall amplitude change within the transition window. Specifically, upper and lower limits can be set, such as limiting the adjusted value to not exceed the maximum value of the original segment or be lower than the minimum value of the next segment, while controlling the change amplitude between adjacent time points to not exceed a preset threshold (e.g., ±15). If the range is exceeded, the value is truncated according to the boundary value.

[0075] Preferably, step S44 includes: Extract the historical value ranges of energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate from the source carbon emission inventory, and determine the value range of each parameter; Divide the range of values ​​into several discrete value nodes to form a corresponding parameter scale sequence; A multidimensional parameter grid is constructed by using a parameter scale sequence, so that each parameter can form a combination relationship under different value nodes; Constraints are introduced into the multidimensional parameter grid, and parameter combinations that meet the coordinated change trend are selected based on the degree of matching between the energy substitution rate and the industrial adjustment coefficient. The filtered parameter combinations are numbered according to a preset order to generate a parameter combination set.

[0076] In one embodiment, based on statistical data of a certain region from 2015 to 2020, historical data on energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate are extracted from source carbon emission inventories and related statistical data to determine the value range of each parameter.

[0077] First, extract the historical value ranges for each parameter. In practice, organize the historical data chronologically. For example, statistics show that the energy substitution rate fluctuates between 0.15 and 0.55, the industrial adjustment coefficient varies between 0.75 and 0.95, and the emission reduction technology penetration rate varies between 0.20 and 0.65. To avoid the influence of extreme values, the data can be truncated by 5% above and below, ultimately determining the energy substitution rate range to be 0.20–0.50, the industrial adjustment coefficient to be 0.80–0.95, and the emission reduction technology penetration rate to be 0.25–0.60.

[0078] Subsequently, discrete value nodes are defined based on the aforementioned value range. In specific implementations, an equal-interval division method can be adopted. For example, the energy substitution rate can be divided into three nodes: 0.20, 0.35, and 0.50; the industrial adjustment coefficient can be divided into three nodes: 0.80, 0.875, and 0.95; and the emission reduction technology penetration rate can be divided into three nodes: 0.25, 0.425, and 0.60.

[0079] Each node is arranged in ascending order to form a corresponding parameter scale sequence, which is stored in an array structure.

[0080] Next, a multidimensional parameter grid is constructed using the aforementioned parameter scale sequence. In practice, a triple loop is used to traverse and combine the three parameter sequences, ensuring that each set of parameters forms a triplet. For example, (0.20, 0.80, 0.25), (0.20, 0.80, 0.425)... up to (0.50, 0.95, 0.60), resulting in a total of 27 original parameter combinations.

[0081] After obtaining the multidimensional parameter grid, constraints are introduced to filter parameter combinations. Specifically, matching rules are set between the energy substitution rate and the industrial adjustment coefficient. For example, when the energy substitution rate is high (≥0.40), the industrial adjustment coefficient is required to be no less than 0.85 to ensure a synergistic effect between energy structure optimization and industrial structure adjustment; when the energy substitution rate is low (<0.30), a lower industrial adjustment coefficient is allowed. Based on these constraints, each of the 27 combinations is evaluated, and combinations that do not meet the synergistic change trend are eliminated, such as mismatched combinations (0.50, 0.80, 0.25).

[0082] Finally, the selected parameter combinations are numbered according to a preset order. Specifically, they can be sorted in the order of energy substitution rate first, followed by industrial adjustment coefficient, and finally technology penetration rate, and then assigned numbers accordingly, such as C1, C2...Cn. A mapping relationship is established between these numbers and the corresponding parameter combinations, and this mapping is stored. This generates the final set of parameter combinations for subsequent carbon emission evolution calculations.

[0083] Preferred methods for obtaining multidimensional parametric meshes include: Based on the value nodes of energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, corresponding one-dimensional parameter sequences are constructed respectively; Based on each one-dimensional parameter sequence, an initial parameter permutation set is generated by superimposing dimensions one by one, so that different parameter values ​​form a complete combination relationship. The initial parameter arrangement set is structurally mapped, and the parameter combinations are filled into the regular array structure according to the dimensional correspondence to form a multidimensional parameter grid.

[0084] In one embodiment, based on the parameter value nodes determined in the previous step, it is assumed that the energy substitution rate value nodes are 0.20, 0.35, and 0.50, the industrial adjustment coefficient value nodes are 0.80, 0.875, and 0.95, and the emission reduction technology penetration rate value nodes are 0.25, 0.425, and 0.60.

[0085] First, construct one-dimensional parameter sequences based on the aforementioned value nodes. In the specific implementation, store each parameter node in an array according to its numerical value. For example, the energy substitution rate sequence is denoted as A=[0.20, 0.35, 0.50], the industrial adjustment coefficient sequence is denoted as B=[0.80, 0.875, 0.95], and the emission reduction technology penetration rate sequence is denoted as C=[0.25, 0.425, 0.60]. Record the length information of each sequence.

[0086] Subsequently, based on the one-dimensional parameter sequences, an initial parameter permutation set is generated using a step-by-step superposition method. In practice, a nested loop approach can be adopted: the energy substitution rate sequence is the first loop, the industrial adjustment coefficient is the second loop, and the emission reduction technology penetration rate is the third loop. In each loop, one element is sequentially selected from each of the three sequences to form a parameter triplet. For example, the first triplet is (0.20, 0.80, 0.25), the second triplet is (0.20, 0.80, 0.425), and this process is repeated until all combinations are generated, resulting in an initial parameter permutation set containing all combination relationships, totaling 27 sets.

[0087] Next, the initial parameter arrangement set is structurally mapped to form a regular multidimensional parameter grid. In actual processing, a three-dimensional array structure Grid[i][j][k] can be constructed, where i corresponds to the energy substitution rate sequence index, j corresponds to the industry adjustment coefficient sequence index, and k corresponds to the emission reduction technology penetration rate sequence index. Then, according to the index correspondence, each set of parameters is filled into the corresponding position. For example, Grid[0][0][0] corresponds to (0.20, 0.80, 0.25), Grid[0][0][1] corresponds to (0.20, 0.80, 0.425), and so on, until the entire three-dimensional array is constructed.

[0088] The above method forms a multidimensional parametric grid with a clear dimensional structure, so that each dimension corresponds to a parameter change direction and the nodes have a fixed positional relationship.

[0089] Of particular importance, step S2 includes: Time alignment is performed on the energy consumption data and carbon emission monitoring data in the basic dataset to form a raw data sequence arranged according to a unified time scale; Extract energy consumption change feature points from the original data sequence, and construct several continuous energy consumption segments with the feature points as boundaries. Based on the relationship between the energy consumption change trend within each energy consumption segment and the emission response during the corresponding time period, segment-level emission contribution indicators are determined. Using segment-level emission contribution identifiers as indexes, each energy consumption segment is mapped to its corresponding energy type and industry category, forming a dynamic source-specific metering unit; Within the dynamic source metering unit, the emissions contributions from multiple sources in the same time segment are superimposed and allocated to obtain the source carbon emissions with time segment identifiers. Carbon emissions from different sources are spliced ​​and aggregated according to time segments to generate a carbon emission inventory by source.

[0090] In one embodiment, energy consumption data and carbon emission monitoring data are time-aligned. In actual processing, data from different sources are resampled using a "month" as the unified time scale. For example, monitoring data with daily or hourly data is converted to monthly data through summation or averaging, and uniformly labeled as a time series from January 2022 to December 2022, thus forming an original data sequence arranged according to a unified time scale.

[0091] Subsequently, energy consumption change feature points are extracted from the original data sequence. In practice, the rate of change in energy consumption is calculated monthly, and when the rate of change for two consecutive months exceeds a preset threshold (e.g., ±10%), that time point is marked as a feature point. For example, if electricity consumption increases by 12% from March to April, then April is marked as a feature point. Using all feature points as boundaries, the annual data is divided into several continuous energy consumption segments, such as "stable segment from January to March", "growth segment from April to June", "high-level fluctuation segment from July to September", and "decline segment from October to December".

[0092] Next, based on the relationship between energy consumption change trends within each energy consumption segment and the corresponding emission response time period, segment-level emission contribution labels are determined. In the specific implementation, energy consumption and emission data within each segment are compared and analyzed, for example, calculating the consistency and correlation of their changing directions. When energy consumption increases and emissions increase simultaneously, it is marked as a "positive contribution"; when energy consumption decreases but emission changes are not significant, it is marked as a "weak contribution"; when the direction of energy consumption change and emission change is inconsistent, it is marked as a "lagging contribution". Each energy consumption segment is assigned a corresponding emission contribution label in this way.

[0093] Then, using the segment-level emission contribution identifier as an index, each energy consumption segment is mapped to its corresponding energy type and industry category. In actual processing, combining the energy category identifier (such as electricity, gas, and coal) and industry classification information (such as industry, construction, and transportation) in the original data, each segment is assigned to the corresponding "energy type + industry category" combination unit, such as "electricity-industry" and "gas-commercial", thereby forming a dynamic source-specific metering unit.

[0094] Furthermore, within each dynamic sub-source metering unit, the emissions contributions from multiple sources in the same time segment are superimposed and allocated. In specific implementation, for emissions from different energy sources within the same time segment, a weighted allocation is performed according to their proportion of total energy consumption. For example, if electricity accounts for 60%, coal for 30%, and natural gas for 10% in a certain segment, the total emissions are allocated to each energy source according to this proportion to obtain the corresponding sub-source carbon emissions, along with a time segment identifier.

[0095] Finally, the carbon emissions from each time segment are spliced ​​and aggregated in chronological order to form a complete carbon emission inventory by source. In practice, this can be stored in a three-dimensional structure of "time segment + energy type + industry category", for example, recorded as (January–March 2022, electricity-industry, emissions X), and arranged sequentially to form a continuous list.

[0096] Of particular importance, based on the relationship between energy consumption change trends within each energy consumption segment and emission response during the corresponding time period, segment-level emission contribution indicators are determined, including: Extract the direction and magnitude of energy consumption changes within each energy consumption segment, and form corresponding trend characteristics; Within the corresponding time period, carbon emission monitoring data is read to obtain the direction and magnitude of emission changes, forming emission response characteristics; Align and match trend features with emission response features, and identify time intervals with consistent change directions as effective response intervals; The degree of dominance of each energy consumption segment on emission changes is determined based on the relative proportion of the change range within the effective response interval. Based on the degree of dominance, each energy consumption segment is assigned a corresponding emission contribution label.

[0097] In one embodiment, taking the monthly energy consumption and carbon emission monitoring data of a certain city from January to June 2022 as an example, the specific implementation process of "determining the segment-level emission contribution identifier based on the relationship between the energy consumption change trend within each energy consumption segment and the emission response of the corresponding period" is explained. Assume that two energy consumption segments are obtained through the previous steps: segment one is January to March, and segment two is March to June.

[0098] First, extract the direction and magnitude of energy consumption changes within each energy consumption segment. In actual processing, the energy consumption data at each time point within a segment are differentially calculated. For example, in segment one, if the electricity consumption increases from 100 to 115, the direction of change is "increasing," and the magnitude is +15%; in segment two, if the consumption decreases from 115 to 105, the direction of change is "decreasing," and the magnitude is -8.7%. The direction and magnitude of change are combined to form trend characteristics, for example, denoted as (increasing, +15%) and (decreasing, -8.7%).

[0099] Subsequently, carbon emission monitoring data is read within the corresponding time period to obtain the direction and magnitude of emission changes. In specific implementation, the same differential method is used to calculate the changes in emission data. For example, if emissions increase from 200 to 230 in segment one, an emission response characteristic (increase, +15%) is formed; if emissions decrease from 230 to 220 in segment two, a (decrease, -4.3%) is formed.

[0100] Next, the trend characteristics and emission response characteristics are time-aligned and matched. In the specific implementation, the direction of change is compared segment by segment, and when the direction of energy consumption change is consistent with the direction of emission change, the time period is marked as a valid response interval. For example, in this case, both segment one and segment two meet the requirement of consistent direction, so they are both identified as valid response intervals; if there are some months with inconsistent directions within a segment, only the consistent sub-intervals are selected as valid response intervals.

[0101] After identifying the effective response range, the dominance level is determined based on the relative proportion of the changes within that range. In practice, the ratio of "emission change rate / energy consumption change rate" can be calculated as the response coefficient; for example, in segment one, it is 15% / 15% = 1.0, and in segment two, it is 4.3% / 8.7% ≈ 0.49. According to preset classification rules, for example, a ratio ≥ 0.8 is considered "strongly dominant," 0.4–0.8 is considered "mediumly dominant," and < 0.4 is considered "weakly dominant." Therefore, segment one can be classified as strongly dominant, and segment two as moderately dominant.

[0102] Finally, based on the degree of dominance, each energy consumption segment is assigned an emission contribution label. In practical applications, "strong dominance" can be labeled as Class I contribution, "medium dominance" as Class II contribution, and "weak dominance" as Class III contribution, and a correlation can be established with the corresponding segment. For example, in this case, segment one is labeled as Class I emission contribution, and segment two is labeled as Class II emission contribution.

[0103] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0104] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for constructing an evaluation model for carbon peaking and carbon neutrality pathways, characterized in that, Includes the following steps: Step S1: Obtain the basic dataset, which includes regional energy consumption data, industrial structure data, and carbon emission source monitoring data; Step S2: Analyze the basic dataset for itemized measurement to construct a source-specific carbon emission inventory; Step S3: Correct the source carbon emission inventory based on the preset emission factor library to obtain standard carbon emission data; Step S4: Establish a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and input standardized carbon emission data into the parameter matrix for coupling calculation to generate multi-scenario carbon emission evolution data; Step S5: Fit and analyze the time series of carbon emission evolution data under multiple scenarios to determine the carbon peak time point and the path to achieve carbon neutrality; Step S6: Construct a path evaluation model based on the carbon peak time node and the carbon neutrality achievement path, and output the corresponding path evaluation results.

2. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Read the time series corresponding to each emission source in the standardized carbon emission data, divide it into several carbon emission segments according to the continuity of time, and record the start and end time and emission change of each segment; Step S42: Calculate the change gradient of each carbon emission segment based on the emission difference between adjacent time points, and determine the corresponding path perturbation weight factor based on the magnitude of the change gradient; Step S43: Based on the distribution of path perturbation weight factors, rearrange the positions and modulate the amplitudes of each carbon emission segment to generate carbon emission perturbation sequence data; Step S44: Construct a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and generate a parameter combination set according to different parameter value ranges; Step S45: Map the carbon emission perturbation sequence data to the parameter combinations in the parameter matrix, and perform weighted coupling calculations accordingly to obtain multiple sets of carbon emission evolution results; Step S46: Summarize the carbon emission evolution results to obtain multi-scenario carbon emission evolution data.

3. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 2, characterized in that, Step S43 includes: The carbon emission segments are sorted according to the magnitude of the path perturbation weight factor to determine the priority sequence of high-weight segments and low-weight segments; Based on the priority sequence, adjust the arrangement of each carbon emission segment on the time axis so that high-weight segments are shifted to key time intervals; Based on the correspondence between the path disturbance weight factor and the preset modulation coefficient, the emission amplitude of each carbon emission segment is amplified or compressed. The carbon emission segments, after position adjustment and amplitude modulation, are reassembled in chronological order to generate a continuous carbon emission perturbation sequence data.

4. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 3, characterized in that, Based on the correspondence between the path disturbance weighting factor and the preset modulation coefficient, the emission amplitude of each carbon emission segment is amplified or compressed as follows: The path disturbance weight factor is divided into several weight level intervals, and a corresponding basic modulation coefficient is preset for each weight level interval. Based on the position of each carbon emission segment in the time series, the stage attribute to which it belongs is determined, and the basic modulation coefficient is modified in stages to obtain the dynamic modulation coefficient. Based on the dynamic modulation coefficient, the emission amplitude of the corresponding carbon emission segment is scaled proportionally.

5. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 4, characterized in that, Based on the dynamic modulation coefficient, scaling the emission amplitude of the corresponding carbon emission segment also includes: Select a transition window of a preset length at the junction of adjacent carbon emission segments, and extract the tail feature value of the previous segment and the head feature value of the next segment within the window. Using the tail feature value as the reference amplitude and the head feature value as the target amplitude, a progressive mapping sequence is constructed within the transition window, gradually shifting from the reference amplitude to the target amplitude. Based on the relative distance between each position within the transition window and the boundary point, a corresponding transition ratio coefficient is assigned, and the scaling result is locally replaced according to the ratio coefficient, so that the amplitude change is a unidirectional gradual shift. After completing the local replacement, the data outside the transition window boundary is kept at the original scaling ratio to form a local adaptive transition structure.

6. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 5, characterized in that, Based on the relative distance between each position within the transition window and the boundary point, a corresponding transition ratio coefficient is assigned, and the scaling result is locally replaced according to the ratio coefficient, so that the amplitude change shows a unidirectional gradual shift, including: Determine the distance values ​​from each location point within the transition window to the boundary point, and mark them in order of distance from near to far; Construct a monotonically changing proportion sequence based on the distance value, so that the position closer to the boundary point corresponds to a larger transition proportion, and the position farther away from the boundary point corresponds to a smaller transition proportion; Using the scaled amplitude as the baseline, and introducing a target transition amplitude, the amplitudes at each position point are weighted and replaced according to the proportion sequence, so that the amplitudes at each position point gradually shift towards the target amplitude according to the proportion. Perform sequential consistency correction on the replaced position points to ensure that the amplitude changes of adjacent position points maintain the same trend. After the replacement is completed, boundary constraints are applied to the overall amplitude change within the transition window.

7. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 6, characterized in that, Step S44 includes: Extract the historical value ranges of energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate from the source carbon emission inventory, and determine the value range of each parameter; Divide the range of values ​​into several discrete value nodes to form a corresponding parameter scale sequence; A multidimensional parameter grid is constructed by using a parameter scale sequence, so that each parameter can form a combination relationship under different value nodes; Constraints are introduced into the multidimensional parameter grid, and parameter combinations that meet the coordinated change trend are selected based on the degree of matching between the energy substitution rate and the industrial adjustment coefficient. The filtered parameter combinations are numbered according to a preset order to generate a parameter combination set.

8. The method for constructing an evaluation model for carbon peaking and carbon neutrality pathways according to claim 7, characterized in that, Methods for obtaining multidimensional parametric meshes include: Based on the value nodes of energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, corresponding one-dimensional parameter sequences are constructed respectively; Based on each one-dimensional parameter sequence, an initial parameter permutation set is generated by superimposing dimensions one by one, so that different parameter values ​​form a complete combination relationship. The initial parameter arrangement set is structurally mapped, and the parameter combinations are filled into the regular array structure according to the dimensional correspondence to form a multidimensional parameter grid.

9. A system for constructing an evaluation model for carbon peaking and carbon neutrality pathways, characterized in that, For performing the method of constructing a carbon peaking and carbon neutrality pathway evaluation model as described in claim 1, the system for constructing the carbon peaking and carbon neutrality pathway evaluation model includes: The data acquisition module is used to acquire the basic dataset, which includes regional energy consumption data, industrial structure data, and carbon emission source monitoring data. The inventory building module is used for the basic dataset of itemized measurements to build a source-specific carbon emission inventory; The emission correction module is used to correct the source carbon emission inventory based on a preset emission factor library to obtain standard carbon emission data. The scenario generation module is used to establish a parameter matrix that includes energy substitution rate, industrial adjustment coefficient and emission reduction technology penetration rate, and input standardized carbon emission data into the parameter matrix for coupled calculation to generate multi-scenario carbon emission evolution data. The path identification module is used to fit and analyze the time series of carbon emission evolution data under multiple scenarios in order to determine the carbon peak time node and the path to achieve carbon neutrality. The evaluation output module is used to construct a path evaluation model based on the carbon peak time node and the carbon neutrality achievement path, and output the corresponding path evaluation results.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method for constructing a carbon peaking and carbon neutrality pathway evaluation model as described in any one of claims 1-8.

Images