Village land remediation weight calculation method, device and medium based on timing correction

Abstract

The village land remediation weight calculation method, device and medium based on time sequence correction belong to the technical field of land resource investigation and planning evaluation, and solve the problem of how to reduce the interference of time sequence data mutation on village land remediation evaluation weight; including breakpoint detection on the time sequence of the combined weight of each evaluation index, calculation of the historical fluctuation standard deviation of the combined weight change amount between adjacent evaluation periods, when the absolute value of the combined weight change amount between any adjacent evaluation periods exceeds the preset multiple of the historical fluctuation standard deviation, it is determined that there is a breakpoint between the corresponding adjacent evaluation periods, and the evaluation period where the breakpoint is located is taken as a new starting evaluation period; the breakpoint detection mechanism based on the historical fluctuation standard deviation is introduced, the mutation in the weight sequence is identified, and after the breakpoint is detected, the whole time decay and exponential smoothing process is reset, so that the old data before the mutation does not interfere with the calculation result after the mutation.

Description

Technical Field

[0001] This invention belongs to the field of land and resources survey and planning evaluation technology, and relates to a method, equipment and medium for calculating the weight of village land consolidation based on time-series correction. Background Technology

[0002] Village land consolidation and ecological restoration is a dynamic process implemented over many years. With the annual updates of satellite remote sensing imagery and the gradual accumulation of socio-economic statistical data, the data foundation upon which planning evaluation is based is constantly being updated. Planning evaluation for village land consolidation and ecological restoration refers to the process of regularly assessing and monitoring the effectiveness of planning implementation by constructing and applying an evaluation indicator system. The evaluation indicator system includes target, criterion, and indicator layers. By calculating the weights of indicators at each level, this system quantifies the relative importance of different consolidation measures and restoration projects, providing a basis for dynamic adjustments to the plan.

[0003] However, existing weight calculation methods are generally static. Once weights are assigned to each indicator in a multi-indicator evaluation system, these weights remain fixed throughout the entire planning evaluation period, making it impossible to adjust and update the weights using data from new years. This static approach has several drawbacks: First, remote sensing data collected at the beginning of planning implementation may no longer accurately reflect the current land use situation, but under a static weight framework, their influence on the weights is exactly the same as the latest data, causing the evaluation results to lag behind reality. Second, the temporal evolution trend of indicator weights is completely ignored; as planning progresses, the relative importance of some evaluation indicators may change.

[0004] Existing technologies, such as the invention patent with publication number CN120106402A, disclose a method for weighted evaluation of low-carbon city planning intervals using a hybrid subjective and objective approach. This method includes real-time collection and preprocessing of indicator data; establishing a user behavior prediction model based on the indicator data to predict user indicator data; dynamically adjusting the weight allocation of subjective and objective evaluation indicators based on the predicted indicator data; constructing a comprehensive evaluation model to calculate the comprehensive score of each region in the city; and formulating optimization measures for each region based on the comprehensive score. However, existing weight calculation methods still suffer from a lack of adaptability to external policies: when planning objectives are revised or policies change, the weight sequence may experience abrupt changes, exceeding the normal time-series fluctuation range. If historical data accumulated before the change is continued to be used for smoothing and prediction after such a sudden change, data bias will occur.

[0005] Therefore, a dynamic weight correction method is needed that can respond to data updates, capture the temporal evolution of indicators, and identify changes, so that the weights of village land use consolidation evaluation can be adaptively adjusted as the planning is implemented. Summary of the Invention

[0006] The technical problem to be solved by this invention is how to reduce the interference of time-series data mutations on the evaluation weight of village land use consolidation.

[0007] The present invention solves the above-mentioned technical problems through the following technical solutions: The method for calculating village land consolidation weights based on time-series correction includes: S1, obtain the combined weights of village land consolidation for multiple consecutive evaluation periods and form a combined weight time series; wherein, the combined weights include subjective weights and objective weights, and the combined weight vector within a single evaluation period is obtained by assigning weights to the subjective weights and objective weights. S2. Configure decay rate parameters independently for each evaluation index, construct a multi-scale time decay function, perform time decay weighting on the objective weights of each evaluation period, recalculate the objective weights after weighting, and update the combined weight time series. S3. A two-parameter exponential smoothing model is used to perform horizontal and trend smoothing on the combined weight time series of each evaluation indicator, and the weight value of the next evaluation period is predicted. The predicted weight of the next evaluation period is weighted and normalized with the combined weight of the current evaluation period to obtain the dynamic correction weight of the current evaluation period. S4: Perform breakpoint detection on the combined weight time series of each evaluation indicator, calculate the historical standard deviation of the change in combined weight during adjacent evaluation weeks, and determine that there is a breakpoint in the corresponding adjacent evaluation week when the absolute value of the change in combined weight during any adjacent evaluation week exceeds the preset multiple of the historical standard deviation. Then, take the evaluation period where the breakpoint is located as the new starting evaluation period and re-execute S2 and S3 on the combined weight time series.

[0008] Furthermore, the combined weights described in S1 incorporate subjective weights by using the credibility coefficient u as a weighting factor. With objective weight Perform linear weighted synthesis and normalization, using the following logical representation:

[0009] in, This represents the combined weight.

[0010] Furthermore, S2 includes the following: S2.1, For the j-th indicator, calculate its mean value over all evaluation periods. with standard deviation And calculate the coefficient of variation. ; S2.2, Assign a decay rate parameter to the j-th index, the decay rate parameter Through the basic decay rate Coefficient of variation corresponding to the aforementioned index Joint solution, ; S2.3, Construct a multi-scale time decay function based on the decay rate parameter. Let T be the time corresponding to the current evaluation period, and t be the time corresponding to the evaluation period of the historical data, satisfying t≤T. Then, the decay coefficient of the j-th index at time t is... As shown in the following formula:

[0011] Where, when t=T, The closer the data is to the current time T, the higher the attenuation coefficient. The closer to 1, the older the data, the higher the attenuation coefficient. The closer it gets to 0; S2.4 Extract the original objective score data of each indicator in all consecutive evaluation periods, and perform time decay weighting on the objective score data corresponding to each evaluation period based on the multi-scale time decay function, and recalculate the objective weights after weighting.

[0012] Furthermore, S2.4 specifically includes: Let X(t) be the original objective score data matrix of the j-th indicator in the t-th evaluation period, containing n evaluation objects. Multiply each element in matrix X(t) by the attenuation coefficient of the corresponding evaluation period to obtain the attenuated weighted score data, as shown in the following formula:

[0013] in, This represents the score of the i-th evaluation object after attenuation weighting for the j-th indicator in the t-th evaluation period. This represents the original objective score of the i-th evaluation object for the j-th indicator in the t-th evaluation period. All data after attenuation weighting are standardized.

[0014] Furthermore, S3 includes the following: S3.1, for the j-th indicator, the combined weight time series over T consecutive evaluation periods is: ,in This represents the combined weight in the Tth evaluation period; S3.2, Define the initial level value. Initialize trend values The level value and trend value of the t-th evaluation period are calculated recursively, as shown in the following formula:

[0015]

[0016] in, The horizontal smoothing coefficient. This is the trend smoothing coefficient. , These are the smoothing level values ​​for the t-th and t-1-th evaluation periods, respectively; , These are the smoothed trend values ​​for the t-th and t-1-th evaluation periods, respectively; S3.3, based on the final level value Compared with the final trend value The sum of these values ​​yields the predicted weights for the next evaluation period. As shown in the following formula:

[0017] in, , These represent the level value and trend value for the Tth evaluation period, respectively. S3.4, the weighted sum of the predicted weights for the next evaluation period and the combined weights for the current evaluation period is weighted, synthesized, and normalized to obtain the final dynamically corrected weights, as shown in the following formula:

[0018] in, Let the final dynamic adjustment weight of the j-th indicator be . , Let be the predicted weights of the j-th and k-th indicators in the (T+1)-th evaluation period, and m be the number of evaluation indicators. , The combined weights of the j-th and k-th indicators actually calculated in the T-th evaluation period; This is the trend synthesis coefficient.

[0019] Furthermore, S4 includes the following: S4.1 For the j-th indicator, obtain its combined weight sequence across all evaluation periods, and calculate the weight change between adjacent evaluation periods. As shown in the following formula:

[0020] in, , They represent the first time. , The combined weights for each evaluation period; S4.2, Calculate the weight change during the adjacent evaluation periods. Standard deviation As a historical benchmark for the combined weight fluctuations of the corresponding indicators, the following formula is used:

[0021] in, Indicates the standard deviation of historical fluctuations. This represents the arithmetic mean of the weight changes of the current indicator across all adjacent evaluation periods. S4.3, Set the sensitivity coefficient Iterate through the weight changes of adjacent evaluation periods one by one. If the following formula is satisfied:

[0022] Then it is determined that there is a breakpoint at the (t+1)th evaluation period. After the breakpoint is detected, the following reset rule is executed: Reset the time corresponding to the current evaluation period in S2.3 to the time corresponding to the evaluation period where the breakpoint is located, and all historical data before the breakpoint will no longer participate in the attenuation weighting calculation; In S3.2, the initial level value is reset to the combined weight at the time corresponding to the evaluation period where the breakpoint is located, and the trend value is reset to the change in the combined weight of the first adjacent period after the breakpoint, and the recursion is restarted.

[0023] Furthermore, the method also includes: S5. Monitor the monotonic trend of the combined weight time series of each evaluation indicator. When the combined weight corresponding to the evaluation indicator changes unidirectionally for multiple consecutive evaluation periods and the cumulative change exceeds the preset threshold, it is determined that the evaluation indicator has a risk of deviation from the plan execution and an early warning signal is issued.

[0024] Furthermore, S5 specifically includes: For the j-th indicator, calculate the change in combined weights during adjacent evaluation periods, extract the positive and negative signs of each change, and if the signs of all combined weight changes are consistent over N consecutive evaluation periods, and if the cumulative change in weights between the first and last evaluation periods exceeds a preset threshold, then the indicator is determined to have a risk of deviation from the plan execution, and an early warning signal is issued.

[0025] An electronic device includes a memory and a processor, the memory being used to store a program that supports the processor in executing the aforementioned time-series-corrected village land consolidation weight calculation method, the processor being configured to execute the program stored in the memory.

[0026] A storage medium storing a computer program, which, when executed by a processor, performs the steps of the aforementioned method for calculating village land consolidation weights based on time-series correction.

[0027] The advantages of this invention are: (1) By independently configuring a decay rate parameter based on the coefficient of variation for each indicator, a multi-scale time decay function is constructed. This allows indicators with drastic interannual fluctuations and strong timeliness to decay faster, while indicators with moderate interannual fluctuations and strong regularity decay slower, thus achieving differentiated time processing for indicators with different data characteristics. The interannual fluctuation characteristics of different indicators vary significantly—indicators with drastic fluctuations have strong timeliness, and the reference value of their outdated data decreases rapidly; indicators with moderate fluctuations have strong regularity, and their historical data has a longer effective period. This invention uses the coefficient of variation to quantitatively characterize the fluctuation characteristics of each indicator. The larger the coefficient of variation, the greater the decay rate, and the smaller the coefficient of variation, the smaller the decay rate, thus achieving differentiated decay for each evaluation indicator.

[0028] This invention recalculates objective weights by applying attenuation weights to the original score data, thus reducing the contribution of older data to the weight calculation and giving greater influence to the most recent data. Compared to existing technologies that use a uniform attenuation factor for all indicators or do not apply attenuation at all, this invention effectively solves the problem of timeliness differences in multi-source time-series data, reduces the interference of sudden changes in time-series data on the weights of village land use improvement evaluation, and enables the evaluation results to respond promptly to data updates, significantly improving the timeliness and accuracy of village land use improvement evaluation.

[0029] (2) This invention employs a multi-scale, two-parameter exponential smoothing model to perform horizontal and trend smoothing on the combined weight time series of each evaluation indicator, giving higher weight to recent data fluctuations in trend prediction. The model eliminates random fluctuations in weights through a horizontal equation, extracting their central position; and captures the sustained direction of weight changes through a trend equation. Smoothing from two parameter perspectives not only smooths and filters historical weight sequences, but more importantly, it enables scientific prediction of weights for the next evaluation period based on the final horizontal and trend values.

[0030] This invention weights and synthesizes predicted values ​​with actual measured values ​​for the current period, ensuring that the final dynamically adjusted weights respect both current measured data and the long-term evolution trend of the weights. Compared to existing methods that rely solely on current data or perform simple linear extrapolation, this invention more accurately reflects the dynamic evolution of indicator weights during the planning and implementation process, avoids misjudgments of weights due to short-term random fluctuations, and improves the evaluation system's adaptability to time-series evolution patterns.

[0031] (3) Because the evaluation of village land consolidation planning has a strong policy orientation, external factors such as adjustments to planning objectives may cause structural changes in indicator weights. Existing dynamic weighting methods mostly rely on continuously differentiable functions for smoothing adjustments. When a change occurs, the old historical data accumulated before the change will continue to contaminate the evaluation results after the change, causing the evaluation results to lag behind reality. To address the above problems, this invention introduces a breakpoint detection mechanism based on the standard deviation of historical fluctuations. By identifying changes in the weight sequence, and resetting the entire time decay and exponential smoothing process after detecting a breakpoint, it prevents old data before the change from interfering with the calculation results after the change.

[0032] This invention calculates the historical standard deviation of the weight changes of each indicator over adjacent periods. When the absolute value of the change over a certain adjacent period exceeds a preset multiple of the standard deviation, a breakpoint is identified at that time. Upon detecting a breakpoint, the time decay baseline time is immediately reset and the exponential smoothing model is reset. All historical data before the breakpoint is excluded from subsequent calculations, and the decay weighting and smoothing recursion is restarted with the period containing the breakpoint as the new starting point. This mechanism enables the invention to automatically identify weight abrupt changes caused by external shocks such as policy changes and quickly eliminate interference from outdated data. It ensures that the evaluation results after the abrupt change are entirely based on valid data under the new policy context, significantly improving the evaluation system's responsiveness and resistance to interference from external policy changes.

[0033] (4) The present invention establishes a monotonic trend monitoring and early warning mechanism. When the weight of a certain indicator changes unidirectionally for multiple consecutive evaluation periods and the cumulative change exceeds the preset threshold, the system automatically issues a planning execution deviation early warning signal to provide forward-looking decision support for managers. Attached Figure Description

[0034] Figure 1 This is a flowchart of the village land consolidation weight calculation method based on time-series correction according to Embodiment 1 of the present invention. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments: Example 1 like Figure 1As shown, specifically, this embodiment discloses a method for calculating village land consolidation weights based on time-series correction, including the following steps: S1, Obtain combined weights: Obtain combined weights for village land consolidation across multiple consecutive evaluation periods and form a combined weight time series; wherein, the combined weights include subjective weights and objective weights, and the combined weight vector within a single evaluation period is obtained by assigning weights to the subjective and objective weights in combination.

[0037] In this embodiment, an annual evaluation cycle is used. For each evaluation indicator in the village land consolidation evaluation index system, the combined weight of village land consolidation is calculated continuously over several years, forming a corresponding weight time series for each year. Specifically, the process of obtaining the combined weights is as follows: First, the subjective weights W are obtained using the Analytic Hierarchy Process (AHP). sub The evaluation indexes are scored by multiple experts in planning, land and environment to form an expert scoring matrix. The expert scoring matrix is ​​compared pairwise to form a judgment matrix. The judgment matrix is ​​then calculated using the square root method. The product of each row of elements is calculated and then the root is taken. After normalization, the subjective weight vector is obtained. Next, the objective weight W is obtained using the CRITIC method. obj ; Finally, using the credibility coefficient u as the weighting coefficient, the subjective weights are... With objective weight Perform linear weighted synthesis and normalization, using the following logical representation:

[0038] in, The value represents the combined weight, and u is the weighting coefficient. In this embodiment, u is preferably 0.5.

[0039] In this embodiment, the village land consolidation evaluation index system includes a target layer, a criterion layer, and an indicator layer. The target layer includes improving village land consolidation and accelerating rural ecological restoration, corresponding to the problems the system aims to address in village land consolidation and ecological restoration. The criterion layer includes farmland consolidation, construction land consolidation, and rural ecological restoration, representing intermediate factors influencing decision-making. The indicator layer includes high-standard farmland construction, dry land to paddy field conversion, consolidation of scattered agricultural facilities construction land, garden land consolidation, rural homestead consolidation, revitalization of existing land, and ecological restoration of pits and ponds. The monitoring content of the corresponding evaluation indicators is shown in Table 1 below. Table 1. Hierarchical Structure of the Village Land Use Consolidation Evaluation Index System

[0040] S2, Multi-scale Time Decay and Objective Weight Recalculation: A decay rate parameter is independently configured for each evaluation indicator, a multi-scale time decay function is constructed, and the objective weights for each evaluation period are weighted by time decay. After weighting, the objective weights are recalculated, and the combined weight time series is updated. Specifically, S2 includes the following: S2.1, Suppose there are m indicators in the village land use consolidation evaluation index system, and the combined weights are obtained, spanning T evaluation periods. For the j-th indicator, calculate its mean over all evaluation periods. with standard deviation And calculate the coefficient of variation. .

[0041] S2.2, Assign a decay rate parameter to the j-th index, the decay rate parameter Through the basic decay rate Coefficient of variation corresponding to the aforementioned index The solution is obtained jointly, using the following logical representation:

[0042] Among them, the coefficient of variation The larger the coefficient of variation, the more drastic the interannual fluctuations and the stronger the timeliness of the indicator, and the greater the decay rate should be assigned; the smaller the coefficient of variation, the more gradual the interannual fluctuations and the stronger the regularity of the indicator, and the smaller the decay rate should be. The preferred value range is [0.1, 0.5].

[0043] S2.3, Construct a multi-scale time decay function based on the decay rate parameter. The multi-scale time decay function adopts a natural exponential decay form. Let T be the time corresponding to the current evaluation period, and t be the time corresponding to the evaluation period of the historical data, satisfying t≤T. Then, the decay coefficient of the j-th index at time t is... As shown in the following formula:

[0044] Where, when t=T, The closer the data is to the current time T, the higher the attenuation coefficient. The closer to 1, the older the data, the higher the attenuation coefficient. The closer a value is to 0, the smaller its contribution to subsequent calculations.

[0045] S2.4 Extract the original objective score data of each indicator in all consecutive evaluation periods, and perform time decay weighting on the objective score data corresponding to each evaluation period based on the multi-scale time decay function, and recalculate the objective weights after weighting.

[0046] Specifically, let X(t) be the original objective score data matrix of the j-th indicator in the t-th evaluation period, containing n evaluation objects. Multiply each element in matrix X(t) by the attenuation coefficient of the corresponding evaluation period to obtain the attenuated weighted score data, as shown in the following formula:

[0047] in, This represents the score of the i-th evaluation object after attenuation weighting for the j-th indicator in the t-th evaluation period. This represents the original objective score of the i-th evaluation object for the j-th indicator in the t-th evaluation period.

[0048] Furthermore, all data after attenuation weighting are standardized.

[0049] S3, Holt Two-Parameter Exponential Smoothing Prediction and Dynamic Correction: A two-parameter exponential smoothing model is used to perform horizontal and trend smoothing on the combined weight time series of each evaluation indicator, and the weight values ​​for the next evaluation period are predicted. The predicted weights for the next evaluation period are then weighted and normalized with the combined weights for the current evaluation period to obtain the dynamically corrected weights for the current evaluation period. Specifically, S3 includes the following: S3.1, for the j-th indicator, the combined weight time series over T consecutive evaluation periods is: ,in This represents the combined weight in the Tth evaluation period, which is also the combined weight at the end of the continuous evaluation period. In this embodiment, each combined weight in the combined weight time series is updated after being combined and weighted in step S1 and then decayed at multiple scales in step S2.

[0050] S3.2, Define the initial level value. Initialize trend values The level value and trend value of the t-th evaluation period are calculated recursively, as shown in the following formula:

[0051]

[0052] in, The horizontal smoothing coefficient. This is the trend smoothing coefficient. , These are the smoothing level values ​​for the t-th and t-1-th evaluation periods, respectively, reflecting the central position of the combined weights after eliminating random fluctuations; , These are the smoothed trend values ​​for the t-th and t-1-th evaluation periods, respectively, reflecting the direction of the continuity of weight changes.

[0053] S3.3, based on the final level value Compared with the final trend value The sum of these values ​​yields the predicted weights for the next evaluation period. As shown in the following formula:

[0054] in, , These represent the level value and trend value for the Tth evaluation period, respectively.

[0055] S3.4, the weighted sum of the predicted weights for the next evaluation period and the combined weights for the current evaluation period is weighted, synthesized, and normalized to obtain the final dynamically corrected weights, as shown in the following formula:

[0056] in, Let the final dynamic adjustment weight of the j-th indicator be . , Let be the predicted weights of the j-th and k-th indicators in the (T+1)-th evaluation period, and m be the number of evaluation indicators. , This represents the combined weight of the j-th and k-th indicators actually calculated in the T-th evaluation period. A larger trend synthesis coefficient η indicates greater confidence in the sustainability of the predicted trend; a smaller η indicates greater conservatism and reliance on current measured data. Preferably, in this embodiment, η is set to 0.5.

[0057] S4, Breakpoint Detection and Reset Based on Historical Fluctuations: Breakpoint detection is performed on the combined weight time series of each evaluation indicator, and the historical fluctuation standard deviation of the weight change during adjacent evaluation weeks is calculated. When the absolute value of the combined weight change during any adjacent evaluation week exceeds a preset multiple of the historical fluctuation standard deviation, it is determined that there is a breakpoint at that time position (corresponding moment). The evaluation period where the breakpoint is located is taken as the new starting evaluation period, and the multi-scale time decay process of S2 and the exponential smoothing process of S3 are re-executed on the combined weight.

[0058] In this embodiment, by introducing a breakpoint detection and weight reset mechanism, weight abrupt changes caused by policy adjustments or unforeseen events during planning implementation can be identified in a timely manner. Once a breakpoint is detected, the system uses that moment as a new benchmark, discards invalid historical data, and re-executes time decay weighting (step S2) and exponential smoothing prediction (step S3), thereby effectively eliminating historical data interference and avoiding trend lag in traditional smoothing models. Specifically, S4 includes the following: S4.1 For the j-th indicator, obtain its combined weight sequence across all evaluation periods, and calculate the weight change between adjacent evaluation periods. As shown in the following formula:

[0059] in, , They represent the first time. , The combined weights for each evaluation period.

[0060] S4.2, Calculate the weight change during the adjacent evaluation periods. Standard deviation As a historical benchmark for the combined weight fluctuations of the corresponding indicators, the following formula is used:

[0061] in, This represents the arithmetic mean of the weight changes of the current indicator over all adjacent evaluation periods. Let be the change in the weight of this indicator between adjacent evaluation periods in the t-th group. It represents the standard deviation of historical fluctuations.

[0062] S4.3, Set the sensitivity coefficient Iterate through the weight changes of adjacent evaluation periods one by one. If the following formula is satisfied:

[0063] Then it is determined that there is a breakpoint at the (t+1)th evaluation period. After the breakpoint is detected, the following reset rule is executed: (1) Time decay benchmark time reset: The time corresponding to the current evaluation period in S2.3 is reset to the time corresponding to the evaluation period where the breakpoint is located. All historical data before the breakpoint will no longer participate in the decay weighted calculation.

[0064] (2) Reset the exponential smoothing model: Reset the initial level value in S3.2 to the combined weight at the time corresponding to the evaluation period where the breakpoint is located, and reset the initial trend value to the combined weight change of the first adjacent period after the breakpoint, and start the recursion again.

[0065] In a preferred embodiment, the sensitivity coefficient Adjustments are made based on the actual application scenario: a larger value is used when the number of available evaluation periods is small to reduce false positives, and a smaller value is used when the number of available evaluation periods is large to improve detection sensitivity. Preferably, the sensitivity coefficient... The preferred value is 2.

[0066] S5, Monotonic Trend Monitoring and Early Warning: Monotonic trend monitoring is performed on the time series of combined weights for each evaluation indicator. When the combined weights corresponding to an indicator change unidirectionally for multiple consecutive evaluation periods and the cumulative change exceeds a preset threshold, it is determined that the indicator has a risk of deviation from the plan execution and an early warning signal is issued.

[0067] Specifically, for the j-th indicator, calculate the change in combined weights during adjacent evaluation periods and extract the sign of each change. If the signs of all combined weight changes are consistent across N+1 consecutive evaluation periods, the following logic applies:

[0068] in, Represents a sign function, when When >0, the value is 1; when When =0, the value is 0; when When <0, the value is -1; This represents the change in the combined weights during the Nth and N+1th evaluation periods.

[0069] Furthermore, if the cumulative change in weight between the first evaluation period and the final evaluation period exceeds a preset threshold, the following logic is satisfied:

[0070] in, Indicates the preset threshold. , The combined weights for the first and Tth evaluation periods are respectively indicated; if the indicator is found to have a risk of deviation from the plan, an early warning signal will be issued.

[0071] Furthermore, to verify the effectiveness of the village land consolidation weight calculation method based on time-series correction described in this embodiment, the land spatial planning data of six villages (defined as AF villages) in a certain region from 2023 to 2025 are used as an example to verify the technical solution provided in this embodiment.

[0072] (1) Data sources and preprocessing: In this embodiment, the objective score data comes from the annual land change survey database and the cultivated land quality grade database of the Natural Resources Bureau. After the area of ​​each map patch corresponding to each indicator is extracted by the GIS platform, it is standardized to the proportion of each map patch area to the total land area of ​​the village.

[0073] In this embodiment, the subjective weights are derived from independent scores given by five domain experts (including those specializing in planning, land resources, and environment) using the Analytic Hierarchy Process (AHP). These scores are then calculated using the square root method and subjected to a consistency check. Based on the method described in S1, a combined weight vector is obtained through combined weighting. Combined weight vectors for the continuous evaluation period from 2023 to 2025 are continuously collected to form a combined weight time series.

[0074] The village land use improvement evaluation index system involved in this embodiment includes seven index layers: high-standard farmland construction, dry land to paddy field conversion, land consolidation for scattered agricultural facilities, land consolidation for gardens, rural homestead consolidation, revitalization of existing land, and ecological restoration of pits and ponds.

[0075] (2) Calculation of multi-scale decay rate parameters: First, calculate the average value of each indicator over all three evaluation periods from 2023 to 2025. Standard deviation and coefficient of variation Based on the basic decay rate =0.3, calculate the decay rate of each index. The results are shown in Table 2 below.

[0076] Table 2. Mean, standard deviation, coefficient of variation, and decay rate of each indicator.

[0077] As shown in Table 2, the coefficient of variation for the index of revitalizing existing land is the largest. =0.224), indicating that this indicator has the most dramatic interannual fluctuations and the strongest timeliness, and its decay rate is correspondingly the largest. =0.367); the coefficient of variation for the dry-to-irrigated conversion index was the smallest ( =0.088), indicating that this indicator has the most gradual interannual fluctuations and the strongest regularity, and its decay rate is correspondingly the smallest. =0.326). Each indicator in Table 2 has achieved a differentiated attenuation configuration based on its own fluctuation characteristics.

[0078] (3) Calculation of time decay coefficient: Using 2025 as the current evaluation period (T=2025), the attenuation coefficients of data for each evaluation period (2023, 2024, 2025) are calculated. The results are shown in Table 3 below.

[0079] Table 3. Decay coefficients of each evaluation indicator in each evaluation period

[0080] As shown in Table 3 above, the attenuation coefficient for all indicators in 2025 (the current period) is 1, indicating that the latest data is completely retained. The attenuation coefficient for 2023 ranges from 0.480 to 0.521, indicating that the contribution of outdated data is significantly weakened. The attenuation coefficient for 2024 ranges from 0.693 to 0.722. Among them, the attenuation coefficient for the stock land revitalization indicator is the lowest in 2023 (0.480), because of its drastic fluctuations and strong timeliness, outdated data is attenuated to the greatest extent. The attenuation coefficient for the dry-to-irrigated conversion indicator is the highest in 2023 (0.521), because of its gentle fluctuations and strong regularity, outdated data retains a relatively high contribution.

[0081] (4) Combined weighted sequence and Holt two-parameter exponential smoothing prediction: For each evaluation indicator, obtain the combined weight sequence for the three evaluation periods of 2023, 2024, and 2025. Holt two-parameter exponential smoothing was performed, with the horizontal smoothing coefficient α set to 0.3 and the trend smoothing coefficient β... h Take 0.2, based on the final level value L. 2025 Compared with the final trend value b 2025 The sum of predicted portfolio weights for 2026 The trend synthesis coefficient η is defined as 0.5. The predicted value is weighted and synthesized with the current combination weight in 2025 and normalized to obtain the final dynamic correction weight in 2025. The results are shown in Table 4 below.

[0082] Table 4. Weights of each evaluation indicator combination, Holt predicted values, and dynamically adjusted weights.

[0083] As shown in Table 4 above, the two-parameter exponential smoothing model in this embodiment provides the predicted values ​​for 2026 by capturing the horizontal center and trend direction of the weight sequences of each evaluation indicator. For example, the weight of the garden land consolidation indicator showed a continuous downward trend from 2023 to 2025 (0.1422, 0.1343, and 0.1305 respectively). Therefore, the model predicts that its weight will further decrease to 0.1220 in 2026. After weighting and combining with the current value in 2025, the corrected weight is 0.1262. This corrected weight reflects both the current measured value and the continuous downward trend. On the other hand, the weight of the pond ecological restoration indicator showed a continuous upward trend (0.1287, 0.1420, and 0.1449 respectively). The model predicts that its weight will rise to 0.1600 in 2026, with a corrected weight of 0.1524, reflecting a reasonable extrapolation of the upward trend.

[0084] (4) Breakpoint detection: This embodiment performs breakpoint detection on the weight changes of each indicator in adjacent periods, and sets a sensitivity coefficient. The optimal value is 2. The weight change Δ of each evaluation indicator in adjacent periods, the standard deviation of historical fluctuation σ, and the judgment results are shown in Table 5 below.

[0085] Table 5. Results of Breakpoint Detection for Each Evaluation Index

[0086] As shown in Table 5 above, the absolute value of Δ(2024-2023) for the parkland consolidation index is 0.0079, exceeding its historical fluctuation threshold of 0.0057. Therefore, it was identified as a breakpoint in 2024. Upon verification, the occurrence of this breakpoint is related to the new parkland use control policy introduced in the region in 2024, which is a structural change in weight caused by external policy changes.

[0087] Upon detecting the breakpoint, the reset rules in step S4 are followed: the time decay baseline is reset to 2024, and historical data from 2023 and earlier are no longer included in subsequent decay weighting calculations; the exponential smoothing model is reset, specifically the initial value of the horizontal value is reset to the combined weight of 0.1343 for 2024, and the initial value of the trend value is reset to the change from 2024 to 2025 -0.0038, and the recursion restarts. The changes in adjacent periods for the remaining six indicators do not exceed their respective 2σ thresholds, indicating that their weight changes fall within the normal time-series fluctuation range and do not require a reset.

[0088] (6) Monotonic trend monitoring and early warning: Monotonic trend warning is implemented for each evaluation indicator: the number of consecutive unidirectional periods is set to N=2, and the direction of change and cumulative change of each evaluation indicator are shown in Table 6 below.

[0089] Table 6. Monitoring Results of Monotonic Trends for Each Evaluation Indicator

[0090] As shown in Table 6 and Table 5, none of the seven indicators triggered an alert.

[0091] In summary, this embodiment demonstrates the effectiveness of the technical solution described herein through a complete calculation process involving 6 villages, 7 indicators, and 3 evaluation periods. (1) In this embodiment, each evaluation index has a differentiated decay rate configuration based on its own coefficient of variation. The index of revitalizing existing land with drastic fluctuations decays the fastest (β=0.367), while the index of converting dry land to paddy fields with gentle fluctuations decays the slowest (β=0.326), effectively realizing the priority use of high-timeliness data.

[0092] (2) In this embodiment, the horizontal center and trend direction of the weight sequence of each indicator were successfully captured by using a two-parameter exponential smoothing model. For example, the continuous downward trend of the garden land consolidation indicator and the continuous upward trend of the pit and pond ecological restoration indicator were reasonably extrapolated.

[0093] (3) The breakpoint detection mechanism proposed in this embodiment successfully identified the structural change in the weight of the garden land consolidation index in 2024 caused by external policy changes (the change amount was 0.0079, which exceeded the threshold of 0.0057), and triggered a dual reset of the time decay benchmark and the exponential smoothing model, effectively avoiding the continuous interference of outdated data before the mutation on the subsequent evaluation results.

[0094] (4) This embodiment continuously tracks the evolution of the weights of each indicator based on the monotonic trend monitoring mechanism, without generating false alarms, thus verifying the rationality and robustness of the early warning rule.

[0095] Example 2 This embodiment further explains the extraction of the original objective score data of each indicator in all continuous evaluation periods in step S2.4 of Embodiment 1. The specific process is as follows: In this embodiment, an annual evaluation cycle is used to extract the actual surveying and statistical data of each village within the current evaluation cycle. The current land use area is calculated using GIS software (such as ArcGIS or QGIS) and compared with the planning data in the Natural Resources Bureau and the village and county planning database.

[0096] After completing the GIS extraction, in order to eliminate the impact of village size differences on the scoring, this embodiment first standardizes the original area data of each indicator, converting the area of ​​each indicator into a proportion of the total village land area, as shown in the following formula:

[0097] in, For the first The first village j The evaluation indicator represents the proportion of the village's total land area in year t. For the first The first village j The first indicator in the t Area per year For the first The total land area of ​​the villages.

[0098] Furthermore, this embodiment proposes a three-stage comprehensive scoring model (total score of 100 points) based on three dimensions: "completion progress," "maintenance status," and "change trend." This model comprehensively evaluates the performance of indicators from three dimensions. The completion progress score (30-60 points) measures the degree to which the evaluation indicator has achieved its target value. The completion progress is scored in segments based on the comparison between the current value, the base period value, and the target value, reflecting the degree of completion of the task objective. The maintenance status score (0-20 points) evaluates the level of maintenance of the current value compared to the base period value. The maintenance status score is calculated using the ratio of the current value to the base period value, ensuring that the evaluation indicator does not regress. The change trend score (0-20 points) measures the direction and speed of change between adjacent monitoring years. The change trend is dynamically scored based on the annual change rate, encouraging continuous improvement of the indicator. The total score for all three dimensions is 100 points. This comprehensive evaluation system focuses on the achievement of the final goal while also considering process maintenance and dynamic development trends, forming a holistic evaluation system for calculating the objective weights of the evaluation indicators in subsequent steps. Specifically, the scoring process of the model is as follows: S2.4.1, Calculate the completion progress score: The progress score is used to measure the degree to which the evaluation indicator has been achieved relative to the target value; for growth-type evaluation indicators, the progress score is calculated as follows:

[0099] For reduction-type evaluation indicators, the progress is calculated as follows:

[0100] in, This represents the completion progress score, with a value ranging from 30 to 60 points; in this embodiment, the current value is... , representing the first The percentage of the area of ​​the j-th evaluation indicator of a village in year t is the proportion of the village's total land area; the base period value is... , representing the first The target value is the proportion of the area of ​​the j-th evaluation indicator of a village to its total land area at the beginning of the planning year. , representing the first The proportion of the area of ​​the j-th evaluation indicator of a village to the total land area of ​​the village at the end of the planning year.

[0101] In this embodiment, the types of evaluation indicators are predetermined based on planning objectives and indicator attributes during the evaluation indicator system construction phase. The classification of evaluation indicator types is determined according to the nature of the indicators and planning objectives. Among them, growth indicators refer to indicators whose values ​​are better the higher they are, such as positive benefit indicators like greening rate and facility coverage rate; while reduction indicators refer to indicators whose values ​​are better the lower they are, such as negative impact indicators like pollution emissions and disaster losses.

[0102] S2.4.2, Calculate the retention case score: The maintenance score is used to measure the level of maintenance of the current value compared to the base period value, preventing a rebound in the effectiveness of the rectification; for growth-type evaluation indicators, the maintenance score is as follows:

[0103] For reduction-type evaluation indicators, the maintenance situation is as follows:

[0104] in, The score indicates the status of the situation, ranging from 0 to 20 points.

[0105] S2.4.3, Calculate the trend score: The aforementioned trend is used to measure the direction of change between adjacent monitoring years and to encourage continuous improvement; among them, for growth-type evaluation indicators, the annual rate of change Δ is calculated. r As shown in the following formula:

[0106] For decreasing evaluation indicators, calculate the annual rate of change Δ. r As shown in the following formula:

[0107] The annual rate of change Δ is expressed by the above formula. r Limited to The interval, based on the annual rate of change Δ r The trend of change is calculated using the following formula:

[0108] in, The score is based on the trend of change, ranging from 0 to 20 points; the previous year's score was... , representing the first The ratio of the area of ​​the j-th evaluation indicator of a village in year t-1 to the total land area of ​​the village is the actual value of the indicator in the previous monitoring year.

[0109] S2.4.4, add the scores of the above three dimensions to obtain the objective score of each indicator in the village. As shown in the following formula:

[0110] Furthermore, the objective score is calculated by iterating through each evaluation indicator to form an objective scoring matrix, and the original objective score data is presented in matrix form.

[0111] Example 3 This embodiment, based on embodiment two, uses the objective weighting method (CRITIC method) to obtain the objective weight W in step S1. obj To provide further explanation, the specific process is as follows: There are h villages to evaluate. m There are 1 evaluation index, and the original data matrix is: ;in, Indicates the first Let h represent the number of villages being evaluated and m represent the number of evaluation indicators.

[0112] Furthermore, regarding objective scores Normalization is performed; among them, the growth-type evaluation indicators (positive benefit indicators, the larger the better) are as follows:

[0113] For reduction-type evaluation indicators (negative impact indicators, the smaller the better), the formula is as follows:

[0114] in, Indicates the first The standardized value of the j-th indicator for evaluating a village. X j This represents a column vector containing the raw data of all evaluated villages on the j-th indicator.

[0115] Furthermore, the Pearson correlation coefficients among the indicators are calculated, and a correlation coefficient matrix R is constructed, containing elements... The calculation is shown in the following formula:

[0116] in, This represents the Pearson correlation coefficient between the j-th and k-th indicators. , Let j and k represent the standardized mean values ​​of the j-th and k-th indicators, respectively. Indicates the first The standardized value of the k-th indicator for evaluating a village.

[0117] For the j-th evaluation index, calculate the standard deviation of the j-th evaluation index. and conflict Standard deviation The larger the value, the stronger the indicator's ability to differentiate and evaluate villages; while the greater the conflict rate... The larger the value, the less information overlaps between the indicator and other indicators, and the greater the amount of independent information, as shown in the following formula:

[0118] Then calculate the information content of the j-th evaluation index. C j and objective weight As shown in the following formula:

[0119] In this embodiment, the CRITIC method not only considers the standard deviation (contrast strength) of the indicators, but also the conflict between the indicators, which can effectively eliminate redundant information and allocate more weight to indicators with a large amount of independent information.

[0120] Example 4 An apparatus includes a memory and a processor, the memory being used to store a program that supports the processor in executing the time-series-corrected village land consolidation weight calculation method of Embodiment 1, the processor being configured to execute the program stored in the memory.

[0121] Example 5 A storage medium storing a computer program, wherein the computer program is executed by a processor to perform the steps of the village land consolidation weight calculation method based on time-series correction in Embodiment 1.

[0122] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein can be considered as a ordered list of executable instructions for implementing logical functions, which can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-based system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0123] For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit a program for use in or in conjunction with an instruction execution system, apparatus, or device. More specific examples of computer-readable media (a non-exhaustive list) include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CD-ROM). Additionally, a computer-readable medium can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optical scanning of the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0124] In the description of this specification, references to terms such as "in a preferred embodiment," "preferred," "in this embodiment," "specific," or "furthermore," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0125] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for calculating village land consolidation weights based on time-series correction, characterized in that, include: S1, obtain the combined weights of village land consolidation for multiple consecutive evaluation periods and form a combined weight time series; wherein, the combined weights include subjective weights and objective weights, and the combined weight vector within a single evaluation period is obtained by assigning weights to the subjective weights and objective weights. S2. Configure decay rate parameters independently for each evaluation index, construct a multi-scale time decay function, perform time decay weighting on the objective weights of each evaluation period, recalculate the objective weights after weighting, and update the combined weight time series. S3. A two-parameter exponential smoothing model is used to perform horizontal and trend smoothing on the combined weight time series of each evaluation indicator, and the weight value of the next evaluation period is predicted. The predicted weight of the next evaluation period is weighted and normalized with the combined weight of the current evaluation period to obtain the dynamic correction weight of the current evaluation period. S4: Perform breakpoint detection on the combined weight time series of each evaluation indicator, calculate the historical standard deviation of the change in combined weight during adjacent evaluation weeks, and determine that there is a breakpoint in the corresponding adjacent evaluation week when the absolute value of the change in combined weight during any adjacent evaluation week exceeds the preset multiple of the historical standard deviation. Then, take the evaluation period where the breakpoint is located as the new starting evaluation period and re-execute S2 and S3 on the combined weight time series.

2. The method for calculating village land consolidation weights based on time-series correction according to claim 1, characterized in that, The combined weights described in S1 use the confidence coefficient u as a weighting factor to incorporate subjective weights. With objective weight Perform linear weighted synthesis and normalization, using the following logical representation: in, This represents the combined weights.

3. The method for calculating village land consolidation weights based on time-series correction according to claim 1, characterized in that, S2 includes the following: S2.1, For the j-th indicator, calculate its mean value over all evaluation periods. with standard deviation And calculate the coefficient of variation. ; S2.2, assign a decay rate parameter to the j-th index, the decay rate parameter Through the basic decay rate Coefficient of variation corresponding to the aforementioned index Joint solution, ; S2.3, Construct a multi-scale time decay function based on the decay rate parameter. Let T be the time corresponding to the current evaluation period, and t be the time corresponding to the evaluation period of the historical data, satisfying t≤T. Then, the decay coefficient of the j-th index at time t is... As shown in the following formula: Where, when t=T, The closer the data is to the current time T, the higher the attenuation coefficient. The closer to 1, the older the data, the higher the attenuation coefficient. The closer it gets to 0; S2.4 Extract the original objective score data of each indicator in all consecutive evaluation periods, and perform time decay weighting on the objective score data corresponding to each evaluation period based on the multi-scale time decay function, and recalculate the objective weights after weighting.

4. The method for calculating village land consolidation weights based on time-series correction according to claim 3, characterized in that, Specifically, S2.4 is as follows: Let X(t) be the original objective score data matrix of the j-th indicator in the t-th evaluation period, containing n evaluation objects. Multiply each element in matrix X(t) by the attenuation coefficient of the corresponding evaluation period to obtain the attenuated weighted score data, as shown in the following formula: in, This represents the score of the i-th evaluation object after attenuation weighting for the j-th indicator in the t-th evaluation period. This represents the original objective score of the i-th evaluation object for the j-th indicator in the t-th evaluation period. All data after attenuation weighting are standardized.

5. The method for calculating village land consolidation weights based on time-series correction according to claim 3, characterized in that, S3 includes the following: S3.1, for the j-th indicator, the combined weight time series over T consecutive evaluation periods is: ,in This represents the combined weight in the Tth evaluation period; S3.2, Define the initial level value. Initialize trend values The level value and trend value of the t-th evaluation period are calculated recursively, as shown in the following formula: in, The horizontal smoothing coefficient. This is the trend smoothing coefficient. , These are the smoothing level values ​​for the t-th and t-1-th evaluation periods, respectively; , These are the smoothed trend values ​​for the t-th and t-1-th evaluation periods, respectively; In the first The combined weights for each evaluation period; S3.3, based on the final level value Compared with the final trend value The sum of these values ​​yields the predicted weights for the next evaluation period. As shown in the following formula: in, , These represent the level value and trend value for the Tth evaluation period, respectively. S3.4, the weighted composite weight of the predicted weight for the next evaluation period and the combined weight for the current evaluation period is weighted and normalized to obtain the final dynamically corrected weight, as shown in the following formula: in, Let be the final dynamically adjusted weight for the j-th indicator. , Let be the predicted weights of the j-th and k-th indicators in the (T+1)-th evaluation period, and m be the number of evaluation indicators. , The combined weights of the j-th and k-th indicators actually calculated in the T-th evaluation period; This is the trend synthesis coefficient.

6. The method for calculating village land consolidation weights based on time-series correction according to claim 5, characterized in that, S4 includes the following: S4.1 For the j-th indicator, obtain its combined weight sequence across all evaluation periods, and calculate the weight change between adjacent evaluation periods. As shown in the following formula: in, , They represent the first time. , The combined weights for each evaluation period; S4.2, Calculate the weight change during the adjacent evaluation periods. Standard deviation As a historical benchmark for the combined weight fluctuations of the corresponding indicators, the following formula is used: in, Indicates the standard deviation of historical fluctuations. This represents the arithmetic mean of the weight changes of the current indicator across all adjacent evaluation periods. S4.3, Set the sensitivity coefficient Iterate through the weight changes of adjacent evaluation periods one by one. If the following formula is satisfied: Then it is determined that there is a breakpoint at the (t+1)th evaluation period. After the breakpoint is detected, the following reset rule is executed: Reset the time corresponding to the current evaluation period in S2.3 to the time corresponding to the evaluation period where the breakpoint is located, and all historical data before the breakpoint will no longer participate in the attenuation weighting calculation; In S3.2, the initial level value is reset to the combined weight at the time corresponding to the evaluation period where the breakpoint is located, and the trend value is reset to the change in the combined weight of the first adjacent period after the breakpoint, and the recursion is restarted.

7. The method for calculating village land consolidation weights based on time-series correction according to claim 1, characterized in that, The method further includes: S5. Monitor the monotonic trend of the combined weight time series of each evaluation indicator. When the combined weight corresponding to the evaluation indicator changes unidirectionally for multiple consecutive evaluation periods and the cumulative change exceeds the preset threshold, it is determined that the evaluation indicator has a risk of deviation from the plan execution and an early warning signal is issued.

8. The method for calculating village land consolidation weights based on time-series correction according to claim 7, characterized in that, Specifically, S5 is: For the j-th indicator, calculate the change in combined weights during adjacent evaluation periods, extract the positive and negative signs of each change, and if the signs of all combined weight changes are consistent over N consecutive evaluation periods, and if the cumulative change in weights between the first and last evaluation periods exceeds a preset threshold, then the indicator is determined to have a risk of deviation from the plan execution, and an early warning signal is issued.

9. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program that supports the processor in executing the village land consolidation weight calculation method based on time-series correction as described in any one of claims 1 to 8, and the processor is configured to execute the program stored in the memory.

10. A storage medium storing a computer program, characterized in that, When the computer program is run by the processor, it performs the steps of the village land consolidation weight calculation method based on time-series correction as described in any one of claims 1 to 8.

Images