Urban low-carbon level evaluation method and device in data missing scenario

Abstract

The application discloses a kind of data missing scenario under city low carbon level evaluation method and device, method includes the following steps: step S1, the multi-source index data of the city to be evaluated is obtained, the completeness of the multi-source index data is analyzed, and the data completeness is calculated;Step S2, based on the data completeness corresponding evaluation mode is enabled and corresponding index set is called;Step S3, the index set called is judged to be missing, according to the index type of missing data from the proxy rule library corresponding estimation rule is selected, and the estimated value of missing index is generated based on estimation rule;Step S4, the real observation data in the index set called is fused with the estimated data generated based on estimation rule, and the complete evaluation data set is constructed;Step S5, the evaluation model of city low carbon development level is constructed, and the city low carbon level evaluation is carried out based on the evaluation data set and the evaluation model.The application can reasonably evaluate the city low carbon level when data is incomplete.

Description

Technical Field

[0001] This invention relates to the field of urban low-carbon development level assessment technology, and in particular to a method and apparatus for assessing urban low-carbon level under data-missing scenarios. Background Technology

[0002] In the process of assessing the level of urban low-carbon development, due to factors such as inconsistent statistical standards, imperfect monitoring systems, and differences in data collection cycles, data on urban low-carbon related indicators often suffer from missing, discontinuous, or incomplete information. Current technologies often employ simple data removal or mean imputation to handle missing data, which can easily introduce systematic biases, affecting the stability and comparability of evaluation results. This is particularly detrimental to comprehensive assessments and horizontal comparisons across multiple cities and years.

[0003] With the advancement of the "dual-carbon" strategy, urban low-carbon assessment has become a crucial link in urban development. However, existing methods for evaluating the level of low-carbon urban development still have some problems in practical applications. These include: 1. The problem of interrupted or distorted assessment processes under data gaps. Existing urban low-carbon assessment methods typically handle missing key indicators by removing samples or simple imputation, which can easily lead to interruptions in the assessment process or amplified result biases, making it difficult to guarantee the continuity and stability of assessment results across multiple cities and years. 2. The problem of assessment models lacking adaptability to data completeness. Existing assessment models usually use a single, fixed indicator system and calculation process, failing to flexibly adjust according to differences in the completeness of urban data. This results in decreased credibility of assessment results and insufficient model adaptability when data is scarce or of low quality. 3. The problem of a lack of linkage mechanism between missing data processing and indicator weights. In existing technologies, missing data processing and indicator weight calculation are usually independent, failing to dynamically adjust weight configurations based on the source and reliability of indicator data. This leads to some indicators having an unreasonable amplified impact on the evaluation results when data is missing or the proportion of estimated data is high. The evaluation model lacks dynamic correction and adaptive feedback capabilities. Existing low-carbon assessment methods are mostly based on static model structures, making it difficult to dynamically correct the assessment results by combining time series information, and also unable to use historical evaluation results to optimize model parameters. As a result, the model is difficult to adapt to changes in the stage of urban development during long-term application.

[0004] Therefore, there is an urgent need to propose a method for urban low-carbon assessment that can maintain the integrity of the assessment process and the stability of the results even in the case of missing data. Summary of the Invention

[0005] In view of this, it is necessary to provide a method and device for assessing the low-carbon level of cities under data-scarce scenarios, so as to effectively solve the technical problem of assessment interruption or distortion in existing low-carbon assessment methods under data-scarce scenarios.

[0006] This invention provides a method for assessing urban low-carbon levels under data-missing scenarios, comprising the following steps:

[0007] Step S1: Obtain multi-source indicator data of the city to be evaluated, perform integrity analysis on the multi-source indicator data, and calculate the data integrity.

[0008] Step S2: Based on the data completeness, enable the corresponding evaluation mode and call the corresponding indicator set;

[0009] Step S3: Perform a missing data determination on the called indicator set, select the corresponding estimation rule from the proxy rule library according to the indicator type of the missing data, and generate the estimated value of the missing indicator based on the estimation rule.

[0010] Step S4: Merge the real observation data from the index set with the estimated data generated based on the estimation rules to construct a complete evaluation dataset;

[0011] Step S5: Construct an evaluation model for the city's low-carbon development level, and assess the city's low-carbon level based on the evaluation dataset and the evaluation model.

[0012] Preferably, step S1 specifically comprises:

[0013] Acquire and input multi-source indicator data for the city to be evaluated, perform integrity analysis on the input multi-source indicator data, and calculate the data completeness:

[0014]

[0015] in, For data integrity, The actual number of available indicators. This represents the total number of indicators in the preset indicator system.

[0016] Preferably, step S2 specifically comprises:

[0017] The cases where the data integrity is less than the first threshold are classified as the basic mode, the cases where the data integrity is between the first and second thresholds are classified as the standard mode, and the cases where the data integrity is greater than the second threshold are classified as the fine mode.

[0018] The basic mode calls the basic indicator set, the standard mode calls the standard indicator set, and the refined mode calls the complete indicator set. The basic indicator set, the standard indicator set, and the complete indicator set have different numbers and structures of indicators, but the evaluation dimensions are the same.

[0019] Preferably, step S3 specifically comprises:

[0020] Determine whether the called indicator set has missing data. If not, directly assess the city's low-carbon level based on the called indicator set. If so, generate missing data based on estimation rules. The estimation rules include data interpolation rules based on historical time series, estimation rules based on matching similar city features, and mapping rules based on regression relationships of relevant indicators.

[0021] Preferably, step S4 specifically comprises:

[0022] The estimated data is corrected by introducing a confidence coefficient, and then fused with the actual observation data to obtain the fused data:

[0023]

[0024] in, The merged index value To estimate the data, For real observation data, The credibility coefficient corresponding to the indicator. Determined based on the missing percentage and correction method.

[0025] Preferably, step S5 specifically comprises:

[0026] Step S51: Preprocess the evaluation dataset, construct a projection pursuit model to optimize the projection direction, and project the evaluation data based on the optimal projection direction to obtain the optimal projection vector;

[0027] Step S52: Calculate the basic weight of each indicator using the CRITIC objective weighting method. Based on the source and reliability coefficient of the indicator data, dynamically adjust the basic weight to obtain the dynamic weight.

[0028] Step S53: Calculate the city's low-carbon development level index based on the projection vector and the dynamic weight.

[0029] Preferably, step S51 specifically includes:

[0030] The evaluation dataset is standardized to obtain normalized data, and the normalized data is linearly projected to obtain an initial projection vector.

[0031] Construct a projection objective function, use an accelerated genetic algorithm to optimize the projection direction to obtain the optimal projection direction, and then project the optimal projection vector based on the optimal projection direction.

[0032] Preferably, step S52 specifically includes:

[0033] The basic weight of each indicator is calculated using the CRITIC objective weighting method:

[0034] ,

[0035] in, The basic weights, For the number of indicators, The standard deviation of the comparative strength of the indicators. Conflict parameter;

[0036] The basic weights are adjusted based on the source and reliability coefficient of the indicator data:

[0037]

[0038] in, The adjusted indicator weights, Based on weights, This is the reliability coefficient of the indicator.

[0039] Preferably, step S53 specifically includes:

[0040] Based on the projection vector decomposition, the index values ​​of each indicator are obtained, and the city's low-carbon development level index is calculated by combining the index weights:

[0041]

[0042] in, This serves as an index for the city's low-carbon development level. As the indicator weight, This refers to the indicator value.

[0043] The present invention also provides a device for assessing the urban low-carbon level under data missing scenarios, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, it implements the method for assessing the urban low-carbon level under data missing scenarios.

[0044] Compared with existing technologies, the advantages of this invention are as follows: Before assessing the low-carbon level, this invention first evaluates the completeness of the data, and based on the data completeness, calls up indicator sets with different numbers and structures of indicators; establishes a proxy rule base, and uses the estimation rules in the proxy rule base to complete the missing data of the indicator set, thereby obtaining a complete dataset, and then conducts the low-carbon level evaluation based on the complete dataset. Therefore, this invention can complete a reasonable assessment of the city's low-carbon development level under the real condition of incomplete data through intelligent data compensation and mode adaptation mechanisms. According to the actual available data, the system automatically selects which assessment mode to use, and intelligently compensates for missing data through a predefined proxy rule base, thereby achieving the goal of three-level data adaptation and intelligent rule compensation. Attached Figure Description

[0045] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0046] Figure 1 is a flowchart of an embodiment of a method for assessing urban low-carbon levels under data-missing scenarios provided by the present invention;

[0047] Figure 2 yes Figure 1 The diagram shows an embodiment of a method for assessing urban low-carbon levels under data-missing scenarios. Detailed Implementation

[0048] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0049] Example 1

[0050] Please see Figure 1 The method for assessing urban low-carbon levels under a data-missing scenario, as described in this embodiment, specifically includes the following steps:

[0051] Step S1: Obtain multi-source indicator data of the city to be evaluated, perform integrity analysis on the multi-source indicator data, and calculate the data integrity.

[0052] Step S2: Based on the data completeness, enable the corresponding evaluation mode and call the corresponding indicator set;

[0053] Step S3: Perform a missing data determination on the called indicator set, select the corresponding estimation rule from the proxy rule library according to the indicator type of the missing data, and generate the estimated value of the missing indicator based on the estimation rule.

[0054] Step S4: Merge the real observation data from the index set with the estimated data generated based on the estimation rules to construct a complete evaluation dataset;

[0055] Step S5: Construct an evaluation model for the city's low-carbon development level, and assess the city's low-carbon level based on the evaluation dataset and the evaluation model.

[0056] The following is for reference. Figure 2 The specific steps are explained.

[0057] S1. Construction of low-carbon evaluation index system and data input, data integrity analysis

[0058] Initiate the urban low-carbon assessment process, obtain and input multi-source indicator data for the city to be assessed, and the indicators shall include at least low-carbon related indicators such as carbon emissions, energy utilization, industrial structure, transportation, ecological environment, resource utilization and technological support.

[0059] Perform a completeness analysis on the input urban basic data and calculate the data completeness index. :

[0060]

[0061] in, The actual number of available indicators. This represents the total number of indicators in the preset indicator system.

[0062] S2. Activate the corresponding evaluation mode and call the indicator set based on data completeness.

[0063] Based on data completeness The current assessment scenarios are divided into the following three categories: Limited data as the basic model; The data is in the standard mode. All data is in fine-grained mode. ;in, The first threshold, This is the second threshold.

[0064] Based on the data integrity analysis results, the corresponding evaluation mode is automatically activated: the basic mode calls the basic indicator set; the standard mode calls the standard indicator set; and the refined mode calls the complete indicator set.

[0065] Different indicator sets may differ in the number and structure of indicators, but they maintain consistency in evaluation dimensions.

[0066] S3. Missing Data Detection and Generation

[0067] The missing data is determined for the called indicator set: if no missing data is found, proceed directly to step S5; if missing data is found, generate the missing data based on the proxy rules first, and then proceed to step S5.

[0068] When data is missing, the pre-built proxy rule base is queried, and the corresponding estimation rule is selected according to the type of missing indicator to generate the estimated value of the missing indicator.

[0069] The proxy rules include, but are not limited to, the following forms: data interpolation rules based on historical time series, estimation rules based on matching similar city features, and mapping rules based on regression relationships of relevant indicators.

[0070] Interpolation rules based on historical time series data:

[0071]

[0072] in, This represents the missing values ​​for the current period's indicators. This is the indicator value from the previous period. This is the indicator value for the next period.

[0073] Estimation rules based on matching similar city features:

[0074]

[0075] in, These are estimated values ​​for the missing indicators. For similar cities, the corresponding indicator values ​​are... For similarity weights, and .

[0076] The mapping rules based on the regression relationship of relevant indicators are explained using three specific data examples.

[0077] a. Estimating carbon sequestration capacity using the vegetation index NDVI

[0078]

[0079] in, For near-infrared reflectivity, Reflectivity in the red light band; The range of values ​​is

[0080]

[0081] in, For net primary productivity, These are regional empirical parameters;

[0082]

[0083] in, Carbon sequestration This refers to the area covered by vegetation.

[0084] b. Estimating economic activity intensity using nighttime light data: Light intensity information within the urban built-up area is extracted using nighttime light remote sensing data. Through pixel brightness accumulation and spatial normalization, a proxy indicator reflecting the intensity of urban economic activity is constructed and used as one of the multi-source data in dynamic urban carbon efficiency modeling.

[0085]

[0086] in, For the first The intensity of economic activity in each city For the city The Middle The light brightness value of each pixel. This represents the number of pixels within the city limits.

[0087] c. Estimating the urban heat island effect using surface temperature: Utilizing surface temperature data obtained through remote sensing inversion, surface temperature information of the urban built-up area and surrounding areas is extracted. By calculating the surface temperature difference or relative heat island intensity index, environmental characteristic indicators reflecting the urban heat island effect are constructed and used as one of the multi-source data in the urban carbon efficiency intelligent assessment model.

[0088]

[0089] in, The average surface temperature of the urban built-up area; Average surface temperature of the surrounding suburbs or natural land surface.

[0090] S4. Fusion of Real and Estimated Data: This involves fusing real observation data with estimated data generated through proxy rules to construct a complete evaluation dataset. A confidence coefficient is then introduced into the estimated data. This results in the corrected index value:

[0091]

[0092] in, The merged index value For proxy data, For real data, The confidence coefficient is determined based on the missing percentage and correction method.

[0093] S5. Based on the preprocessed data, construct the projection pursuit model PPM to complete the city's low-carbon development.

[0094] S51. Data preprocessing: Standardize the fused dataset using deviation standardization to achieve dimensionless processing, resulting in a standardized matrix. After processing, all index values ​​are mapped to... The interval represents the normalized values ​​of the evaluation indicators for the development level of low-carbon cities.

[0095] Constructing projection index vectors After normalization Dimensional Data Perform a linear projection, projecting onto the projection direction. Above, set , For a unit length vector, the projection value for:

[0096]

[0097] in, For the first The projection vector of each sample. For the first The normalized sample number of the th sample A number, Projection vector The projection direction.

[0098] Constructing the projection objective function In order to find the structural combination features of data in multidimensional indicators, when performing comprehensive projection, the projected values ​​are required to be... Extract as much as possible The variation information in, that is, the requirement Standard deviation As large as possible, while projecting Local density To reach the maximum. Based on this, the projection objective function can be constructed as:

[0099]

[0100]

[0101]

[0102] in, The standard deviation of the projected values ​​is used to reflect the dispersion of the data. The local density of the projected values ​​is used to reflect the characteristics of the data structure. For local collections The average value, Let be the distance between the two samples in the projection space. , The window radius representing the local density is usually taken as 0.1. , The unit step function takes the following values:

[0103] .

[0104] An accelerated genetic algorithm based on real-number encoding is used to determine the projection direction. Optimization includes: encoding, real-number encoding of variables; fitness function; Genetic operators: selection, crossover, mutation; introduction of acceleration operators to improve convergence speed.

[0105] The optimal projection direction is obtained through iterative search: .

[0106] S52, Calculation of indicator weights.

[0107] The initial calculation of indicator weights utilizes the CRITIC objective weighting method to calculate the basic weights. First, the indicator values ​​are obtained based on projection vector decomposition. Then, indicator standardization is performed on the ... The first indicator in the Raw value of the year , and perform dimensionless processing.

[0108] Positive indicators:

[0109]

[0110] Contrarian indicator:

[0111]

[0112] Standard deviation of indicator contrast strength:

[0113]

[0114] in, For the length of time, As an indicator The mean, As an indicator In time The value at that time.

[0115] Indicator comparison strength, conflict calculation formula:

[0116]

[0117]

[0118] Formula for calculating objective basis weights: , .

[0119] The dynamic adjustment of indicator weights involves adjusting the indicator weights dynamically based on the source and reliability of the indicator data during the comprehensive evaluation process.

[0120]

[0121] in, The adjusted indicator weights, For the original weights, This is the reliability coefficient of the indicator.

[0122] S53. Calculate the comprehensive low-carbon score. Based on the corrected indicator values ​​and dynamic weights, calculate the city's comprehensive low-carbon score:

[0123]

[0124] in, This is an index of the city's low-carbon development level.

[0125] The assessment report is generated based on the city's low-carbon development level index, and the process is completed.

[0126] Based on the urban low-carbon development level index, cities are divided into high-level low-carbon cities, medium-level low-carbon cities, or low-level low-carbon cities using the natural breakpoint method or quantile method, thereby identifying the low-carbon construction stages and key directions of different types of cities.

[0127] Based on the city's low-carbon development level index, the assessment results of the city's low-carbon development level are output, along with the comprehensive evaluation results and corresponding analysis report, concluding the assessment process. Furthermore, the low-carbon development level index can be used to conduct dynamic comparative analysis and assessment of the low-carbon development levels of different cities and regions. By combining the weight structure of each evaluation indicator, key factors restricting low-carbon development in each city can be identified, providing quantitative basis for differentiated regulation and policy formulation of low-carbon cities at the provincial level.

[0128] This embodiment maintains the continuity of the assessment process even when indicator data is missing, improving the comprehensiveness and accuracy of urban low-carbon development evaluation results. It dynamically switches evaluation modes based on data completeness, enhancing the adaptability of the evaluation method to changes in urban development stages and timelines. Weighted calculation models are used to weight the indicators, and missing data correction and weight adjustments reduce estimation errors. The methodology is clear and the calculation steps are standardized, facilitating its application in different cities and regions. It possesses significant engineering and application value, providing quantitative reference for urban low-carbon development decision-making and related management work.

[0129] Example 2

[0130] This embodiment provides a device for assessing urban low-carbon levels under data-missing scenarios, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, it implements the urban low-carbon level assessment method under data-missing scenarios described in Embodiment 1.

[0131] The urban low-carbon level assessment device under data missing scenarios provided in this embodiment is used to implement the urban low-carbon level assessment method under data missing scenarios. Therefore, the technical effects of the urban low-carbon level assessment method under data missing scenarios are also possessed by the urban low-carbon level assessment device under data missing scenarios, and will not be described again here.

[0132] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of the present invention.

Claims

1. A method for assessing urban low-carbon levels under data-missing scenarios, characterized in that, Includes the following steps: Step S1: Obtain multi-source indicator data of the city to be evaluated, perform integrity analysis on the multi-source indicator data, and calculate the data integrity. Step S2: Based on the data completeness, enable the corresponding evaluation mode and call the corresponding indicator set; Step S3: Perform a missing data determination on the called indicator set, select the corresponding estimation rule from the proxy rule library according to the indicator type of the missing data, and generate the estimated value of the missing indicator based on the estimation rule. Step S4: Merge the real observation data from the index set with the estimated data generated based on the estimation rules to construct a complete evaluation dataset; Step S5: Construct an evaluation model for the city's low-carbon development level, and assess the city's low-carbon level based on the evaluation dataset and the evaluation model.

2. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S1 specifically involves: Acquire and input multi-source indicator data for the city to be evaluated, perform integrity analysis on the input multi-source indicator data, and calculate the data completeness: in, For data integrity, The actual number of available indicators. This represents the total number of indicators in the preset indicator system.

3. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S2 specifically involves: The cases where the data integrity is less than the first threshold are classified as the basic mode, the cases where the data integrity is between the first and second thresholds are classified as the standard mode, and the cases where the data integrity is greater than the second threshold are classified as the fine mode. The basic mode calls the basic indicator set, the standard mode calls the standard indicator set, and the refined mode calls the complete indicator set. The basic indicator set, the standard indicator set, and the complete indicator set have different numbers and structures of indicators, but the evaluation dimensions are the same.

4. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S3 specifically involves: Determine whether the called indicator set has missing data. If not, directly assess the city's low-carbon level based on the called indicator set. If so, generate missing data based on estimation rules. The estimation rules include data interpolation rules based on historical time series, estimation rules based on matching similar city features, and mapping rules based on regression relationships of relevant indicators.

5. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S4 specifically involves: The estimated data is corrected by introducing a confidence coefficient, and then fused with the actual observation data to obtain the fused data: in, The merged index value To estimate the data, For real observation data, The credibility coefficient corresponding to the indicator. Determined based on the missing percentage and correction method.

6. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S5 specifically involves: Step S51: Preprocess the evaluation dataset, construct a projection pursuit model to optimize the projection direction, and project the evaluation data based on the optimal projection direction to obtain the optimal projection vector; Step S52: Calculate the basic weight of each indicator using the CRITIC objective weighting method. Based on the source and reliability coefficient of the indicator data, dynamically adjust the basic weight to obtain the dynamic weight. Step S53: Calculate the city's low-carbon development level index based on the projection vector and the dynamic weight.

7. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 6, characterized in that, Step S51 specifically involves: The evaluation dataset is standardized to obtain normalized data, and the normalized data is linearly projected to obtain an initial projection vector. Construct a projection objective function, use an accelerated genetic algorithm to optimize the projection direction to obtain the optimal projection direction, and then project the optimal projection vector based on the optimal projection direction.

8. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S52 specifically involves: The basic weight of each indicator is calculated using the CRITIC objective weighting method: ， in, The basic weights, For the number of indicators, The standard deviation of the comparative strength of the indicators. Conflict parameter; The basic weights are adjusted based on the source and reliability coefficient of the indicator data: in, The adjusted indicator weights, Based on weights, This is the reliability coefficient of the indicator.

9. The method for assessing urban low-carbon levels under data-missing scenarios according to claim 1, characterized in that, Step S53 specifically involves: Based on the projection vector decomposition, the index values ​​of each indicator are obtained, and the city's low-carbon development level index is calculated by combining the index weights: in, This serves as an index for the city's low-carbon development level. As the indicator weight, This refers to the indicator value.

10. A device for assessing urban low-carbon levels under data-missing scenarios, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, which, when executed by the processor, implements the method for assessing urban low-carbon levels under data-missing scenarios as described in any one of claims 1-9.

Images