A sea area performance evaluation method and system based on geospatial analysis technology
By dividing the marine performance assessment into assessment units, obtaining multi-source indicator data, performing local information entropy correction and asymmetric processing, and combining local Moran index to optimize the resolution coefficient, the problems of neglecting spatial autocorrelation and heterogeneity in existing technologies are solved, and more accurate assessment results are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SOUTH CHINA SEA PLANNING & ENVIRONMENT RES INST SOA
- Filing Date
- 2026-03-03
- Publication Date
- 2026-06-12
AI Technical Summary
Existing marine performance assessment methods neglect the spatial interaction and correlation between assessment units, making it difficult to identify key areas and potential problems, thus affecting the scientific validity of the assessment results and the effectiveness of decision support.
By dividing the target sea area into assessment units, obtaining multi-source assessment indicators, calculating the first-order statistical moments of indicator data within the spatial neighborhood, using local information entropy for initial correction of the resolution coefficient, constructing an asymmetric processing function, calculating the comprehensive grey relational degree, and combining the local Moran index for secondary correction, the spatial heterogeneity and autocorrelation reflection capabilities of the assessment model are improved.
It has improved the scientific rigor and accuracy of marine area performance assessment, enhanced the ability to identify risks and shortcomings, and provided more reliable decision support.
Smart Images
Figure CN122198328A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of performance evaluation, and in particular relates to a method and system for evaluating marine performance based on geospatial analysis technology. Background Technology
[0002] As a vital strategic space for a nation, the effectiveness assessment of marine areas is crucial for marine resource development, environmental protection, rights and interests safeguarding, and national defense security. Most marine area effectiveness assessment methods are based on indicator systems and employ models such as the Analytic Hierarchy Process (AHP), fuzzy comprehensive evaluation, and data envelopment analysis (DEA) for calculation. While these traditional methods can assess marine conditions to some extent, they often overlook the universally present characteristics of "nearby similarity" or "nearby dissimilarity" in geographic space, i.e., spatial autocorrelation; they also ignore the uneven spatial distribution of different indicators, i.e., spatial heterogeneity. This neglect of spatial characteristics can lead to deviations between the assessment results and objective geographical reality, thus affecting the scientific validity of the assessment and the effectiveness of decision support. Grey relational analysis, due to its advantages of low sample size requirements and no need for typical distribution patterns, is applied to assessment problems involving multiple indicators and incomplete information. In standard grey relational analysis, the resolution coefficient, as a key parameter, is usually set to a fixed value based on experience and applied to all assessment units. Existing assessment models typically use a symmetrical approach when representing the deviation of a unit's state from its surrounding environment, i.e., treating positive and negative deviations equally. However, in practical applications, the impact of negative bias is often more prominent and requires greater attention. Therefore, how to endogenously integrate spatial heterogeneity, spatial autocorrelation, and the asymmetry of assessment needs into the assessment model, adjust key parameters, and construct a comprehensive assessment framework that can reflect spatial differentiation patterns and interaction effects is a technical challenge that urgently needs to be solved in the field of marine performance assessment. Summary of the Invention
[0003] This invention proposes a marine area performance assessment method based on geospatial analysis technology to address the problem that existing technologies neglect the spatial interaction and correlation between assessment units, making it difficult to identify key areas and potential problems, thus affecting the scientific validity of the assessment results and the effectiveness of decision support. The method includes: The target sea area is divided into assessment units, and multi-source assessment indicators for each assessment unit are obtained. For any assessment unit, the indicator data of all units in the spatial neighborhood are extracted, and the first-order statistical moments of the neighborhood indicator data are calculated. Calculate the spatial Gini coefficient of each evaluation index in the entire domain and the local information entropy of each evaluation unit; when the spatial Gini coefficient of a certain evaluation index exceeds a preset threshold, use the local information entropy of the evaluation unit to make an initial correction to the preset resolution coefficient corresponding to the index. Calculate the deviation sequence between the index value of each evaluation unit and the first-order statistical moments of the neighborhood index data, and set an asymmetric processing function based on the local information entropy of the evaluation unit: for negative deviations, amplify the absolute value through an exponential function; for positive deviations, compress the value through a logarithmic function. Based on the processed deviation sequence and the initially corrected resolution coefficient, the comprehensive grey relational degree of each evaluation unit is calculated to obtain the initial effectiveness value; based on the initial effectiveness value, the local Moran index of each evaluation unit is calculated, and the spatial clustering type is determined accordingly; based on the spatial clustering type, the initially corrected resolution coefficient is subjected to a second weighted correction, and based on the second corrected resolution coefficient and the processed deviation sequence, the comprehensive grey relational degree is recalculated to obtain the marine effectiveness.
[0004] Optionally, the step of dividing the target sea area into assessment units and obtaining multi-source assessment indicators for each assessment unit includes: The target sea area was divided into N×M square assessment units using the equidistant grid method under the projected coordinate system, with each unit having an actual size of 1km×1km. Data on water depth, bottom sediment, salinity, sea temperature, chlorophyll concentration, distance from major shipping routes, and distance from ports for each assessment unit were acquired and represented as multi-source assessment indicators.
[0005] Optionally, the step of extracting index data from all units within the spatial neighborhood and calculating the first-order statistical moments of the neighborhood index data includes: The eight evaluation units that share an edge or vertex with any evaluation unit are defined as a spatial neighborhood; the arithmetic mean of the index data of all nine units in the spatial neighborhood is calculated as the first-order statistical moment.
[0006] Optionally, when the spatial Gini coefficient of a certain evaluation index exceeds a preset threshold, the preset resolution coefficient corresponding to the index is initially corrected using the local information entropy of the evaluation unit, including: For any evaluation index j, when the spatial Gini coefficient is greater than a preset threshold, the initial correction resolution coefficient of the index in evaluation unit i is... Through formula Calculation, where To evaluate the local information entropy of unit i; otherwise, The value is taken as the global reference resolution coefficient r.
[0007] Optionally, setting the asymmetric processing function based on the local information entropy of the evaluation unit includes: For values The negative deviation is expressed using an exponential function. Magnification processing is performed; For values The positive deviation is expressed using a logarithmic function. Perform compression processing; Among them, parameters a and b are derived from local information entropy. Uniquely determined, the relationship is .
[0008] Optionally, the step of calculating the local Moran index of each evaluation unit based on the initial performance value and determining the spatial clustering type accordingly includes: When the local Moran index salience test p-value is less than 0.05 and the index value is positive, and the initial power value is higher than the neighborhood mean, it is identified as a high-high clustering region; when the p-value is less than 0.05 and the index value is positive, and the initial power value is lower than the neighborhood mean, it is identified as a low-low clustering region; when the p-value is less than 0.05 and the index value is negative, and the initial power value is higher than the neighborhood mean, it is identified as a high-low anomaly region; when the p-value is less than 0.05 and the index value is negative, and the initial power value is lower than the neighborhood mean, it is identified as a low-high anomaly region.
[0009] Optionally, the second weighted correction of the resolution coefficients after the initial correction based on the spatial clustering type includes: A secondary correction weight is set for each spatial clustering type; for each index j within evaluation unit i, the resolution coefficient of the index after the initial correction is... Multiplying the result by the second-correction weight corresponding to the unit yields the second-corrected resolution coefficient. .
[0010] Optionally, the step of recalculating the comprehensive grey relational degree based on the resolution coefficients after secondary correction and the processed deviation sequence includes: Using formula Recalculate the comprehensive grey relational degree of evaluation unit i on all k indicators, where The value of unit i in the processed deviation sequence on index j. The resolution coefficient is the result of the second correction.
[0011] Furthermore, this invention also relates to a marine performance evaluation system based on geospatial analysis technology, comprising: The calculation module is used to divide the target sea area into assessment units and obtain multi-source assessment indicators for each assessment unit; for any assessment unit, it extracts the indicator data of all units in the spatial neighborhood and calculates the first-order statistical moments of the neighborhood indicator data. The correction module is used to calculate the spatial Gini coefficient of each evaluation index in the whole domain and the local information entropy of each evaluation unit; when the spatial Gini coefficient of a certain evaluation index exceeds the preset threshold, the preset resolution coefficient corresponding to the index is initially corrected by using the local information entropy of the evaluation unit. The setting module is used to calculate the deviation sequence between the index value of each evaluation unit and the first-order statistical moment of the neighborhood index data, and to set an asymmetric processing function based on the local information entropy of the evaluation unit: for negative deviations, the absolute value is amplified by an exponential function; for positive deviations, the value is compressed by a logarithmic function. The evaluation module is used to calculate the comprehensive grey relational degree of each evaluation unit based on the processed deviation sequence and the initially corrected resolution coefficients to obtain an initial performance value; calculate the local Moran index of each evaluation unit based on the initial performance value, and determine the spatial clustering type accordingly; perform a second weighted correction on the initially corrected resolution coefficients based on the spatial clustering type, and recalculate the comprehensive grey relational degree based on the second corrected resolution coefficients and the processed deviation sequence to obtain the marine performance.
[0012] Preferably, the step of dividing the target sea area into assessment units and obtaining multi-source assessment indicators for each assessment unit includes: The target sea area was divided into N×M square assessment units using the equidistant grid method under the projected coordinate system, with each unit having an actual size of 1km×1km. Data on water depth, bottom sediment, salinity, sea temperature, chlorophyll concentration, distance from major shipping routes, and distance from ports for each assessment unit were acquired and represented as multi-source assessment indicators.
[0013] Preferably, the step of extracting the index data of all units within the spatial neighborhood and calculating the first-order statistical moments of the neighborhood index data includes: The eight evaluation units that share an edge or vertex with any evaluation unit are defined as a spatial neighborhood; the arithmetic mean of the index data of all nine units in the spatial neighborhood is calculated as the first-order statistical moment.
[0014] Preferably, when the spatial Gini coefficient of a certain evaluation indicator exceeds a preset threshold, the preset resolution coefficient corresponding to the indicator is initially corrected using the local information entropy of the evaluation unit, including: For any evaluation index j, when the spatial Gini coefficient is greater than a preset threshold, the initial correction resolution coefficient of the index in evaluation unit i is... Through formula Calculation, where To evaluate the local information entropy of unit i; otherwise, The value is taken as the global reference resolution coefficient r.
[0015] Preferably, the step of setting the asymmetric processing function based on the local information entropy of the evaluation unit includes: For values The negative deviation is expressed using an exponential function. Magnification processing is performed; For values The positive deviation is expressed using a logarithmic function. Perform compression processing; Among them, parameters a and b are derived from local information entropy. Uniquely determined, the relationship is .
[0016] Preferably, the step of calculating the local Moran index of each evaluation unit based on the initial performance value and determining the spatial clustering type accordingly includes: When the local Moran index salience test p-value is less than 0.05 and the index value is positive, and the initial power value is higher than the neighborhood mean, it is identified as a high-high clustering region; when the p-value is less than 0.05 and the index value is positive, and the initial power value is lower than the neighborhood mean, it is identified as a low-low clustering region; when the p-value is less than 0.05 and the index value is negative, and the initial power value is higher than the neighborhood mean, it is identified as a high-low anomaly region; when the p-value is less than 0.05 and the index value is negative, and the initial power value is lower than the neighborhood mean, it is identified as a low-high anomaly region.
[0017] Preferably, the second weighted correction of the resolution coefficients after the initial correction based on the spatial clustering type includes: A secondary correction weight is set for each spatial clustering type; for each index j within evaluation unit i, the resolution coefficient of the index after the initial correction is... Multiplying the result by the second-correction weight corresponding to the unit yields the second-corrected resolution coefficient. .
[0018] Preferably, the step of recalculating the comprehensive grey relational degree based on the resolution coefficients after secondary correction and the processed deviation sequence includes: Using formula Recalculate the comprehensive grey relational degree of evaluation unit i on all k indicators, where The value of unit i in the processed deviation sequence on index j. The resolution coefficient is the result of the second correction.
[0019] This invention enhances the scientific rigor and accuracy of marine performance assessment by deeply integrating spatial heterogeneity and spatial autocorrelation. It utilizes the spatial Gini coefficient and local information entropy to initially correct the resolution coefficients in grey relational analysis, ensuring these coefficients reflect the spatial imbalance of indicators and the information complexity of each assessment unit. By constructing an asymmetric processing function, the negative deviation of assessment units from their neighborhood mean is amplified, enhancing the model's ability to identify risk weaknesses. The local Moran's index is used to identify spatial clustering patterns, and the resolution coefficients are then corrected a second time, allowing the assessment results to fully reflect the impact of spatial clustering effects on unit performance. The resulting performance assessment results represent the spatial pattern and interactions of performance, providing a more reliable basis for marine spatial management decisions. Attached Figure Description
[0020] Figure 1 A flowchart of the first embodiment; Figure 2 A schematic diagram illustrating the definition and statistics of spatial neighborhood; Figure 3 A schematic diagram of the asymmetric processing function for the deviation value; Figure 4 This is a schematic diagram of the initial performance calculation process. Detailed Implementation
[0021] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.
[0022] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.
[0023] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."
[0024] In the first embodiment, the present invention proposes a method for evaluating marine effectiveness based on geospatial analysis technology, such as... Figure 1 ,include: S1, Divide the target sea area into assessment units and obtain multi-source assessment indicators for each assessment unit; For any assessment unit, extract the indicator data of all units in the spatial neighborhood and calculate the first-order statistical moments of the neighborhood indicator data. Specifically, using Geographic Information System (GIS) tools, a 1km × 1km square grid covering the entire target sea area is generated, with each grid serving as an evaluation unit and assigned a unique number. For each evaluation unit, environmental indicators such as chlorophyll concentration and sea surface temperature are extracted from satellite remote sensing imagery, human activity indicators such as ship density and sailing speed are extracted from Automatic Identification System (AIS) data, and hydrodynamic indicators such as wave height and ocean current speed are obtained from ocean numerical models. These indicator values are then applied to the corresponding evaluation unit using spatial averaging or interpolation methods, forming a multidimensional indicator vector for each unit.
[0025] For evaluation unit i, the eight neighboring evaluation units are defined as a spatial neighborhood. All indicator data for these eight neighborhood units are extracted. For a specific indicator, such as ship density, the arithmetic mean of the sum of the ship density values from the eight neighboring units and divided by eight is the first-order statistical moment of the neighborhood indicator data at unit i. This process is repeated for all indicators to obtain the neighborhood mean for each indicator for each evaluation unit.
[0026] In an optional embodiment, dividing the target sea area into assessment units and obtaining multi-source assessment indicators for each assessment unit includes: The target sea area was divided into N×M square assessment units using the equidistant grid method under the projected coordinate system, with each unit having an actual size of 1km×1km. Data on water depth, bottom sediment, salinity, sea temperature, chlorophyll concentration, distance from major shipping routes, and distance from ports for each assessment unit were acquired and represented as multi-source assessment indicators.
[0027] Assuming the target sea area is 100km × 100km, the universal transverse Mercator projection coordinate system is used to ensure the accuracy of distances and area. Using an equidistant grid method, the sea area is divided into 10,000 square evaluation units, each with a unique geographical extent and number. This forms a regular geographic analysis framework, providing a foundation for subsequent data overlay and spatial analysis.
[0028] For each of the 10,000 assessment units, seven indicator data points need to be acquired. For example, for unit number i, the average water depth is 50m obtained by consulting electronic nautical charts; the seabed is determined to be sandy based on geological data, represented by the value 2; the average sea surface temperature of this unit is 15℃, salinity is 34 PSU, and chlorophyll concentration is 0.8mg / m³, obtained by inversion from satellite remote sensing imagery; the distance from the center point of this unit to the nearest major shipping route is calculated to be 5km, and the distance to the nearest port is 20km, calculated using geographic information system software. This process is repeated for all units, forming a raw assessment indicator database of 10,000 rows and 7 columns.
[0029] To detect the local environmental characteristics of each evaluation unit, in an optional embodiment, the extraction of index data for all units within the spatial neighborhood and the calculation of the first-order statistical moments of the neighborhood index data include: The eight evaluation units that share an edge or vertex with any evaluation unit are defined as a spatial neighborhood; the arithmetic mean of the index data of all nine units in the spatial neighborhood is calculated as the first-order statistical moment.
[0030] To define the spatial neighborhood, taking a central cell i as an example, the spatial neighborhood consists of eight adjacent cells: top, bottom, left, right, upper left, upper right, lower left, and lower right, forming a 3×3 analysis window. This neighborhood definition method, also known as the Queens Adjacency Criterion, comprehensively considers the influence of the surrounding environment.
[0031] Assuming the water depth index is analyzed, the water depth of the central unit i is 50m, and the water depths of its eight neighboring units are 48, 49, 51, 52, 47, 50, 53, and 54m, respectively. The calculated arithmetic mean is 50.44m. This value is used as the first statistical moment of the central unit i on the water depth index. It smooths out the extreme values of a single unit and reflects the overall water depth of the area where that point is located. This calculation process is repeated for each index of each evaluation unit, generating a new set of index datasets that reflect the local average characteristics, such as... Figure 2 .
[0032] S2, calculate the spatial Gini coefficient of each evaluation index in the whole domain and the local information entropy of each evaluation unit; when the spatial Gini coefficient of a certain evaluation index exceeds the preset threshold, use the local information entropy of the evaluation unit to make an initial correction to the preset resolution coefficient corresponding to the index. Specifically, for the ship density index, the values of the index across all evaluation units are sorted, and the spatial distribution unevenness across the entire sea area is calculated using the Gini coefficient formula to obtain the spatial Gini coefficient. For any evaluation unit i, the values of each index in the unit are compared with the sum of the corresponding index values in neighboring units to calculate the contribution ratio of each index within the local range. This contribution is then summed using the information entropy formula to obtain the local information entropy of the unit. A global preset resolution coefficient, such as 0.5, and a Gini coefficient threshold, such as 0.4, are set. If the calculated spatial Gini coefficient of the ship density index is 0.6, exceeding 0.4, then for evaluation unit i, the resolution coefficient for the ship density index is no longer 0.5, but is corrected to 0.5 multiplied by 1 plus the local information entropy of unit i. If the Gini coefficient of another index, such as sea surface temperature, is 0.2, not exceeding the threshold, then the resolution coefficient remains 0.5 across all units.
[0033] In an optional embodiment, when the spatial Gini coefficient of a certain evaluation index exceeds a preset threshold, the preset resolution coefficient corresponding to the index is initially corrected using the local information entropy of the evaluation unit, including: For any evaluation index j, when the spatial Gini coefficient is greater than a preset threshold, the initial correction resolution coefficient of the index in evaluation unit i is... Through formula Calculation, where To evaluate the local information entropy of unit i; otherwise, The value is taken as the global reference resolution coefficient r.
[0034] For all seven evaluation indicators, their spatial Gini coefficients were calculated across the entire target sea area. It was assumed that the calculated spatial Gini coefficients for water depth were 0.7 and for sea surface temperature (SST) were 0.4. The preset Gini coefficient threshold was 0.6, and the preset global baseline resolution coefficient was 0.5. Since the Gini coefficient for water depth (0.7) is greater than the threshold of 0.6, it indicates that the spatial distribution of water depth is extremely uneven, thus requiring local correction of the water depth resolution coefficient. Conversely, the Gini coefficient for SST (0.4) is less than the threshold of 0.6, indicating a relatively uniform spatial distribution; therefore, the SST resolution coefficient was maintained at the baseline value of 0.5 across all evaluation units.
[0035] For the water depth index, the resolution coefficient of each evaluation unit i will be based on the local information entropy of that unit. Adjustments are needed. Local information entropy. This reflects the uncertainty of the index values in unit i and its neighborhood. Assume that for evaluation unit A, the calculated local information entropy... If the value is 0.9, then the initial correction resolution coefficient for the water depth index in unit A is... For another evaluation unit B with lower information entropy, the local information entropy... If the value is 0.2, then the resolution coefficient is In one embodiment, the local information entropy is calculated before performing the calculation. A normalization operation is performed. Regions with strong spatial heterogeneity receive higher resolution coefficients, thus being given greater weight in subsequent calculations.
[0036] S3, calculate the deviation sequence between the index value of each evaluation unit and the first-order statistical moment of the neighborhood index data, and set an asymmetric processing function according to the local information entropy of the evaluation unit: for negative deviations, amplify the absolute value through an exponential function; for positive deviations, compress the value through a logarithmic function. Specifically, for the ship density index of evaluation unit i, a deviation value is obtained by subtracting the neighborhood mean from its own index value. If the value is negative, it indicates that the ship density is lower than the surrounding average level. The processed value is calculated by adding a negative e to the local information entropy of unit i as the base, raising the absolute value of the deviation to the power of 1, and then subtracting 1. If the value is positive, it indicates that the ship density is higher than the surrounding area. The processed value is calculated by adding a negative e to the local information entropy of unit i as the base, raising the absolute value of the deviation to the logarithm of the true value. This calculation is performed on all indices to form a processed deviation sequence.
[0037] In an optional embodiment, setting the asymmetric processing function based on the local information entropy of the evaluation unit includes: For values The negative deviation is expressed using an exponential function. Amplification is performed; values are... The positive deviation is expressed using a logarithmic function. Compression processing is performed; where parameters a and b are determined by local information entropy. Uniquely determined, the relationship is .
[0038] A unique bias handling function is customized for each evaluation unit i, the form of which is determined by the local information entropy of that unit. Decision. Assume the local information entropy of evaluation unit i. The calculated value is 0.7. According to the formula, the value of parameter a is a = 1.7, and the value of parameter b is b = 1.3. The negative deviation handling function for this unit is... The positive deviation processing function is In regions with higher information entropy, a larger value of 'a' amplifies negative bias more strongly; conversely, a smaller value of 'b' compresses positive bias more strongly. In one embodiment, the local information entropy is also considered before calculation. Perform a normalization operation.
[0039] The deviation sequence is the difference between the comparison sequence and the reference sequence. Assuming that at evaluation unit i, the deviation value for the water depth index is -0.2, a negative deviation, and the deviation value for the sea surface temperature index is 0.3, a positive deviation, then, using a function customized for this unit, the processed water depth deviation is f(-0.2) ≈ -0.288, with the absolute value amplified. The processed sea surface temperature deviation is f(0.3) ≈ 0.329, with the absolute value compressed. This asymmetric processing method reflects the sensitivity to negative impacts and the approach to positive impacts, and the degree of sensitivity and caution varies with the complexity of the local environment, such as... Figure 3 .
[0040] S4. Based on the processed deviation sequence and the initially corrected resolution coefficient, calculate the comprehensive grey relational degree of each evaluation unit to obtain the initial effectiveness value; calculate the local Moran index of each evaluation unit based on the initial effectiveness value, and determine the spatial clustering type accordingly; perform a second weighted correction on the initially corrected resolution coefficient based on the spatial clustering type, and recalculate the comprehensive grey relational degree based on the second corrected resolution coefficient and the processed deviation sequence to obtain the sea area effectiveness.
[0041] Specifically, a sequence matrix is constructed from the deviation values of all evaluation units for each indicator after processing. An ideal reference sequence is determined, for example, by taking the maximum deviation value of each indicator after processing across all units. For each evaluation unit, the correlation coefficient between the processed deviation value of each indicator and the corresponding value in the reference sequence is calculated using the grey relational analysis formula. The resolution coefficient used in the calculation is the resolution coefficient after initial correction by local information entropy. The weights of each indicator are determined using methods such as the entropy weight method. The correlation coefficients of each indicator in a single unit are weighted and summed to obtain the comprehensive grey relational degree of that evaluation unit, which is the initial effectiveness value. Figure 4 .
[0042] The initial performance values of all evaluation units are treated as a spatial variable. For each evaluation unit, a local Moran index is calculated using its own initial performance value, the initial performance values of its neighboring units, and the global mean performance value. Specifically, a spatial weight matrix is constructed, where each unit's neighborhood consists of eight units sharing edges or vertices. The initial performance values of all evaluation units, i.e., the comprehensive grey relational degree, are treated as a spatial dataset. For any unit i, the calculation of the local Moran index combines the deviation of unit i's performance value from its global average performance value, the weighted average of the deviations of the performance values of all its neighboring units from their global average performance value, and the variance of the entire dataset, using the formula... The calculation yielded that, It is the evaluation value of the i-th evaluation unit. It is the calculated average of the initial performance values of all evaluation units; For spatial weights, preferably, if j is the spatial neighborhood of i, then it is 1, otherwise it is 0. In one embodiment, the weight vector is standardized so that the sum of all weights of each unit i is 1. The variance of all initial performance values is used for evaluation. In an optional embodiment, the evaluation value is the initial performance value, i.e., the overall gray-scale correlation degree.
[0043] Based on the sign of the calculated local Moran index and the difference between the unit's own efficiency value and the global mean, the spatial clustering type is determined to be one of four types: high-high clustering areas surrounded by high-efficiency values, low-low clustering areas surrounded by low-efficiency values, high-low anomaly areas surrounded by low-efficiency values, or low-high anomaly areas surrounded by high-efficiency values.
[0044] Based on the spatial clustering types determined above, the resolution coefficients are adjusted. If an evaluation unit is identified as a high-high or low-low cluster, indicating consistency with the surrounding environment, the initial revised resolution coefficients of all indicators are multiplied by a weight less than 1, such as 0.8, to enhance the influence of the neighborhood. If it is identified as a high-low or low-high anomaly, indicating a heterogeneous point, the resolution coefficients are multiplied by a weight greater than 1, such as 1.2, to weaken the influence of the neighborhood. For units that do not exhibit spatial clustering, the resolution coefficients remain unchanged. Using the resolution coefficients after secondary correction and the processed deviation sequence obtained above, the comprehensive grey relational degree calculation process is completely repeated, and the new comprehensive grey relational degree obtained is the marine effectiveness of the evaluation unit. The higher the relational degree, the higher the effectiveness, representing a stronger degree of integration, coordination, or spatial homogeneity between the unit's overall condition and the surrounding environment; conversely, the lower the relational degree, the lower the effectiveness, indicating that the unit is an anomaly or mutation point significantly different from its neighborhood.
[0045] In an optional embodiment, the step of calculating the local Moran index of each evaluation unit based on the initial performance value and determining the spatial clustering type accordingly includes: When the p-value of the local Moran index salience test is less than 0.05 and the index value is positive, and the initial efficacy value is higher than the neighborhood mean, it is judged as a high-high clustering area; When the P-value is less than 0.05, the exponent is positive, and the initial efficiency value is lower than the neighborhood mean, it is identified as a low-low clustering region. When the P-value is less than 0.05, the exponent is negative, and the initial efficacy value is higher than the neighborhood mean, it is identified as a high-low anomaly zone. When the P-value is less than 0.05, the exponent is negative, and the initial efficiency value is lower than the neighborhood mean, it is identified as a low-high anomaly zone.
[0046] For each evaluation unit, a local Moran's index and corresponding p-value are calculated. This index is based on the unit's initial efficacy evaluation value, such as a preliminary composite score. This step aims to identify the spatial association pattern between each unit and its neighborhood in terms of efficacy values. The sign of the local Moran's index indicates whether the association is positive or negative, while the p-value tests whether the association is statistically prominent, typically using 0.05 as the prominence level. Each unit is classified according to a series of conditions. For example, evaluation unit A has an initial efficacy value of 0.9, and the average efficacy value of its neighboring units is 0.8. The calculated local Moran's index is 0.6, and the p-value is 0.01. Because the p-value is less than 0.05, the index is positive, and the unit value of 0.9 is higher than the neighborhood mean of 0.8, unit A is classified as a high-high clustering area, i.e., a hotspot. Similarly, evaluation unit B has an efficacy value of 0.8, a neighborhood mean of 0.3, a Moran's index of -0.7, and a p-value of 0.02. Since the P-value is less than 0.05, the exponent is negative, and the unit value of 0.8 is higher than the neighborhood mean of 0.3, unit B is determined to be a high-low anomaly zone, that is, a high-value island surrounded by low values. Similarly, all units are classified into high-high clustering, low-low clustering, high-low anomaly, low-high anomaly, or non-prominent categories according to the four determination rules.
[0047] To further optimize the resolution coefficient by utilizing the spatial clustering information of the cells, in an optional embodiment, the second weighted correction of the initially corrected resolution coefficient based on the spatial clustering type includes: A secondary correction weight is set for each spatial clustering type; For each index j within evaluation unit i, the resolution coefficients of the index after initial correction are... Multiplying the result by the second-correction weight corresponding to the unit yields the second-corrected resolution coefficient. .
[0048] Fixed weight values were assigned to different spatial clustering types. High-high clustering areas, as key focus areas, were assigned a weight of 1.2 to enhance their influence in the evaluation. Low-low clustering areas, as less prominent areas, were assigned a weight of 0.8 to moderately reduce their influence. High-low or low-high anomaly areas, with more complex spatial meanings, were assigned a neutral weight of 1.0 to maintain their original influence.
[0049] The weights are applied to each metric for each unit. Assuming unit A is identified as a high-high aggregation zone, all metrics will undergo a secondary correction using a weight of 1.2. If the resolution coefficient of unit A's water depth metric is after the initial correction... If the value is 0.95, then the resolution coefficient after the second correction is... If the assessment unit identifies a low-low aggregation zone, the initial correction resolution coefficient for the sea surface temperature index is... If the value is 0.5, then the resolution coefficient after the second correction is For units identified as anomalous or inconspicuous areas, the resolution coefficient is multiplied by 1.0 and remains unchanged. In this way, the resolution coefficient not only reflects the heterogeneity and local information entropy of the indicator itself, but also incorporates information about the unit within the macroscopic spatial pattern.
[0050] To generate the final evaluation result, in an optional embodiment, the recalculation of the comprehensive grey relational degree based on the second-corrected resolution coefficient and the processed deviation sequence includes: Using formula Recalculate the comprehensive grey relational degree of evaluation unit i across all k indicators, where The value of unit i in the processed deviation sequence on index j. The resolution coefficient is the result of the second correction.
[0051] It is necessary to determine the global parameters in the formula, that is, the processed bias sequence across all evaluation units and all indicators. The minimum and maximum absolute values. Assuming the calculated global minimum absolute deviation... The global maximum absolute deviation is 0.01. The value is 1.5. These two values remain unchanged in the calculation of all subsequent units.
[0052] For each evaluation unit i, calculate the grey relational coefficient of the unit on each indicator and average it. Taking evaluation unit i as an example, there are a total of k equal to 7 indicators. For the first indicator, water depth, assume the absolute value of the processed deviation. The resolution coefficient after secondary correction is 0.25. The correlation coefficient is 1.14. Therefore, the correlation coefficient of the water depth index is approximately 0.878. The correlation coefficients of the other six indicators are calculated using the same method, for example, 0.75, 0.92, 0.81, 0.88, 0.79, and 0.95 respectively. Adding these seven correlation coefficients and dividing by 7 yields the comprehensive grey relational degree of evaluation unit i. This yields a comprehensive evaluation value that reflects the overall effectiveness level of the unit.
[0053] In a second embodiment, the present invention also provides a marine area performance assessment system based on geospatial analysis technology, comprising: The calculation module is used to divide the target sea area into assessment units and obtain multi-source assessment indicators for each assessment unit; for any assessment unit, it extracts the indicator data of all units in the spatial neighborhood and calculates the first-order statistical moments of the neighborhood indicator data. The correction module is used to calculate the spatial Gini coefficient of each evaluation index in the whole domain and the local information entropy of each evaluation unit; when the spatial Gini coefficient of a certain evaluation index exceeds the preset threshold, the preset resolution coefficient corresponding to the index is initially corrected by using the local information entropy of the evaluation unit. The setting module is used to calculate the deviation sequence between the index value of each evaluation unit and the first-order statistical moment of the neighborhood index data, and to set an asymmetric processing function based on the local information entropy of the evaluation unit: for negative deviations, the absolute value is amplified by an exponential function; for positive deviations, the value is compressed by a logarithmic function. The evaluation module is used to calculate the comprehensive grey relational degree of each evaluation unit based on the processed deviation sequence and the initially corrected resolution coefficients to obtain an initial performance value; calculate the local Moran index of each evaluation unit based on the initial performance value, and determine the spatial clustering type accordingly; perform a second weighted correction on the initially corrected resolution coefficients based on the spatial clustering type, and recalculate the comprehensive grey relational degree based on the second corrected resolution coefficients and the processed deviation sequence to obtain the marine performance.
[0054] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, compact disc read-only memory (CD-ROM), optical storage, etc.) containing computer-usable program code.
[0055] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0056] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for evaluating marine effectiveness based on geospatial analysis technology, characterized in that, include: The target sea area is divided into assessment units, and multi-source assessment indicators for each assessment unit are obtained; For any evaluation unit, extract the index data of all units in the spatial neighborhood, and calculate the first-order statistical moments of the neighborhood index data; Calculate the spatial Gini coefficient of each evaluation index in the entire domain and the local information entropy of each evaluation unit; when the spatial Gini coefficient of a certain evaluation index exceeds a preset threshold, use the local information entropy of the evaluation unit to make an initial correction to the preset resolution coefficient corresponding to the index. Calculate the deviation sequence between the index value of each evaluation unit and the first-order statistical moments of the neighborhood index data, and set an asymmetric processing function based on the local information entropy of the evaluation unit: for negative deviations, amplify the absolute value through an exponential function; for positive deviations, compress the value through a logarithmic function. Based on the processed deviation sequence and the initial corrected resolution coefficient, the comprehensive grey relational degree of each evaluation unit is calculated to obtain the initial performance value. The local Moran index of each evaluation unit is calculated based on the initial performance value, and the spatial clustering type is determined accordingly. Based on the spatial clustering type, the resolution coefficients after the initial correction are subjected to a second weighted correction, and based on the resolution coefficients after the second correction and the processed deviation sequence, the comprehensive grey relational degree is recalculated to obtain the sea area effectiveness.
2. The method according to claim 1, characterized in that, The process of dividing the target sea area into assessment units and obtaining multi-source assessment indicators for each assessment unit includes: The target sea area was divided into N×M square evaluation units using the equidistant grid method in the projected coordinate system, with each unit having an actual size of 1km×1km. Data on water depth, bottom sediment, salinity, sea temperature, chlorophyll concentration, distance from major shipping routes, and distance from ports for each assessment unit were acquired and represented as multi-source assessment indicators.
3. The method according to claim 1, characterized in that, The step of extracting index data from all units within the spatial neighborhood and calculating the first-order statistical moments of the neighborhood index data includes: The eight evaluation units that share an edge or vertex with any evaluation unit are defined as a spatial neighborhood; The arithmetic mean of the various index data of all nine units in the spatial neighborhood is calculated and used as the first-order statistical moment.
4. The method according to any one of claims 1-3, characterized in that, When the spatial Gini coefficient of a certain evaluation indicator exceeds a preset threshold, the preset resolution coefficient corresponding to the indicator is initially corrected using the local information entropy of the evaluation unit, including: For any evaluation index j, when the spatial Gini coefficient is greater than a preset threshold, the initial correction resolution coefficient of the index in evaluation unit i is... Through formula Calculation, where To evaluate the local information entropy of unit i; otherwise, The value is taken as the global reference resolution coefficient r.
5. The method according to any one of claims 1-3, characterized in that, The step of setting the asymmetric processing function based on the local information entropy of the evaluation unit includes: For values The negative deviation is expressed using an exponential function. Magnification processing is performed; For values The positive deviation is expressed using a logarithmic function. Perform compression processing; Among them, parameters a and b are derived from local information entropy. Uniquely determined, the relationship is , .
6. The method according to any one of claims 1-3, characterized in that, The calculation of the local Moran index of each evaluation unit based on the initial performance value, and the determination of the spatial clustering type accordingly, includes: When the p-value of the local Moran index salience test is less than 0.05 and the index value is positive, and the initial efficacy value is higher than the neighborhood mean, it is judged as a high-high clustering area; When the P-value is less than 0.05, the exponent is positive, and the initial efficiency value is lower than the neighborhood mean, it is identified as a low-low clustering region. When the P-value is less than 0.05, the exponent is negative, and the initial efficacy value is higher than the neighborhood mean, it is identified as a high-low anomaly zone. When the P-value is less than 0.05, the exponent is negative, and the initial efficiency value is lower than the neighborhood mean, it is identified as a low-high anomaly zone.
7. The method according to any one of claims 1-3, characterized in that, The second weighted correction of the resolution coefficients after the initial correction based on the spatial clustering type includes: A secondary correction weight is set for each spatial clustering type; For each index j within evaluation unit i, the resolution coefficients of the index after initial correction are... Multiplying the result by the second-correction weight corresponding to the unit yields the second-corrected resolution coefficient. .
8. The method according to claim 7, characterized in that, The recalculation of the comprehensive grey relational degree based on the resolution coefficient after secondary correction and the processed deviation sequence includes: Using formula Recalculate the comprehensive grey relational degree of evaluation unit i on all k indicators, where Let be the value of unit i in index j in the processed deviation sequence.
9. A marine area performance evaluation system based on geospatial analysis technology, characterized in that, include: The calculation module is used to divide the target sea area into assessment units and obtain multi-source assessment indicators for each assessment unit. For any evaluation unit, extract the index data of all units in the spatial neighborhood, and calculate the first-order statistical moments of the neighborhood index data; The correction module is used to calculate the spatial Gini coefficient of each evaluation index in the whole domain and the local information entropy of each evaluation unit; when the spatial Gini coefficient of a certain evaluation index exceeds the preset threshold, the preset resolution coefficient corresponding to the index is initially corrected by using the local information entropy of the evaluation unit. The setting module is used to calculate the deviation sequence between the index value of each evaluation unit and the first-order statistical moment of the neighborhood index data, and to set an asymmetric processing function based on the local information entropy of the evaluation unit: for negative deviations, the absolute value is amplified by an exponential function; For positive deviations, the values are compressed using a logarithmic function; The evaluation module is used to calculate the comprehensive grey relational degree of each evaluation unit based on the processed deviation sequence and the initial corrected resolution coefficient, and obtain the initial performance value. The local Moran index of each evaluation unit is calculated based on the initial performance value, and the spatial clustering type is determined accordingly. Based on the spatial clustering type, the resolution coefficients after the initial correction are subjected to a second weighted correction, and based on the resolution coefficients after the second correction and the processed deviation sequence, the comprehensive grey relational degree is recalculated to obtain the sea area effectiveness.
10. The system according to claim 9, characterized in that, The process of dividing the target sea area into assessment units and obtaining multi-source assessment indicators for each assessment unit includes: The target sea area was divided into N×M square evaluation units using the equidistant grid method in the projected coordinate system, with each unit having an actual size of 1km×1km. Data on water depth, bottom sediment, salinity, sea temperature, chlorophyll concentration, distance from major shipping routes, and distance from ports for each assessment unit were acquired and represented as multi-source assessment indicators.