A karst topography-constrained direct current landing point multi-index comprehensive evaluation method

Abstract

The application provides a kind of karst topography constrained DC landing point multi-index comprehensive evaluation method, comprising: obtaining the geological survey data of multiple candidate landing points of karst topography, power grid structure parameters and engineering economic parameters, generating contradictory focus weight vector;For each landing point, respectively adopt the first screening path based on spatial distance and the second screening path based on improved trend correlation, obtain the first and second candidate subsets;Coupling verification of two subsets, take the intersection as the determined recommended candidate set;Obtain the geographic information map containing power grid development planning information, combine the recommended candidate set, generate a landing point decision recommendation package containing an elastic adaptable space area risk early warning list.The method provided in the application adopts geological-electric unified quantitative mapping, dynamic contradiction focus identification and bidirectional independent screening coupling verification mechanism, which can realize adaptive decision support for DC landing site selection in complex geological environment, and improve the scientific nature, pertinence and engineering reliability of site selection evaluation.

Description

Technical Field

[0001] This application relates to the field of power system planning and decision support technology, and in particular to a multi-index comprehensive evaluation method for DC landing points constrained by karst topography. Background Technology

[0002] High-voltage direct current (HVDC) transmission is a key technology for achieving long-distance, high-capacity power transmission, and its converter stations, or HVDC landing points, are the core hubs of the entire system. The site selection of HVDC landing points is a complex systems engineering decision, requiring a trade-off between multiple objectives such as meeting grid technical requirements, engineering economics, and minimizing environmental impact. This is especially true in karst topographical areas with complex geological conditions, where geological hazard risks become a key factor restricting site selection. Therefore, establishing a scientific, objective, and comprehensive multi-index evaluation method is crucial for HVDC landing point site selection decisions.

[0003] In existing technologies, DC landing site selection evaluation often employs methods such as the Analytic Hierarchy Process (AHP) and fuzzy comprehensive evaluation combined with Geographic Information System (GIS) technology. These methods typically begin by having domain experts construct an evaluation index system encompassing multiple levels of technical, economic, and environmental factors. Then, the weights of each index are determined through expert scoring or questionnaires. Finally, each candidate landing site is weighted and scored to arrive at a comprehensive ranking. GIS is primarily used for spatial data management, visualization analysis, and preliminary regional screening, such as excluding prohibited construction areas like basic farmland and nature reserves.

[0004] However, the aforementioned existing technical solutions have several shortcomings in practical applications. First, the evaluation process relies too heavily on expert experience; the determination of indicator weights and the scoring of some indicators are highly subjective, potentially leading to significant differences in evaluation results due to changes in the expert team. Second, the evaluation methods for special geological conditions such as karst landforms are rather crude, making it difficult to accurately quantify the long-term, complex impacts of nonlinear, strongly coupled factors such as karst development intensity and groundwater stability on engineering safety and economy. Furthermore, traditional evaluation methods typically employ a fixed weighting system to assess all candidate solutions, failing to provide differentiated and focused evaluations based on the unique engineering challenges faced by different candidate scenarios. Summary of the Invention

[0005] This application provides a multi-index comprehensive evaluation method for DC landing points constrained by karst topography. To solve the above-mentioned technical problems, this application adopts the following technical methods: This application provides a multi-index comprehensive evaluation method for DC landing points constrained by karst topography, including: Obtain geological survey data, power grid structure parameters, and engineering economic parameters for multiple candidate landing sites located in karst landform areas; Based on the geological survey data, power grid structure parameters, and engineering economic parameters, a contradiction focus weight vector for the current candidate landing point is generated. Using the aforementioned contradiction focus weight vector, a first screening path based on spatial distance and a second screening path based on improved trend correlation are executed for each candidate landing point to obtain a first candidate subset and a second candidate subset. The first candidate subset and the second candidate subset are coupled and verified, and the intersection result is used as the recommended candidate set. Obtain a geographic information map containing power grid development planning information, and based on the determined recommended candidate set and the geographic information map, generate a landing point decision recommendation package containing flexible and adaptable spatial areas and a risk warning list.

[0006] Optionally, generating the conflict focus weight vector of the current candidate landing point based on the geological survey data, power grid structure parameters, and engineering economic parameters includes: Based on the geological survey data, the set of geological-electrical impact scores for each candidate landing point was calculated; Based on the power grid structure parameters and engineering economic parameters, the set of electrical performance indicators for each candidate landing point is calculated. Based on the geological-electrical impact score set and the electrical performance index set, the conflict degree between the indices is calculated to generate the conflict focus weight vector of the current candidate landing point.

[0007] Optionally, based on the geological survey data, a set of geological-electrical impact scores for each candidate landing point is calculated; including: Several key geological factors were extracted from the geological exploration data; Each key geological factor is input into the corresponding nonlinear influence function in the quantitative influence function library for calculation, and an independent risk quantitative score is generated for each key geological factor, including three dimensions: construction feasibility, long-term safety and full-cycle economics. By aggregating the independent risk quantification scores of all key geological factors for each candidate landing point on the same dimension, a set of geological-electrical impact scores for that candidate landing point is generated.

[0008] Optionally, the step of calculating a set of electrical performance indicators for each candidate landing point based on the power grid structure parameters and engineering economic parameters includes: Based on the power grid structure parameters, an equivalent model of the power system including the current candidate landing point access scheme is constructed; Based on the power system equivalent model and the engineering economic parameters, the electrical performance index set for each candidate landing point is obtained through power flow calculation and short-circuit calculation.

[0009] Optionally, the step of calculating the conflict degree between indicators based on the geological-electrical impact score set and the electrical performance index set, and generating the conflict focus weight vector of the current candidate landing point, includes: Based on the geological-electrical impact score set and the electrical performance index set, a comprehensive index set for the current candidate landing point is constructed; Calculate the degree of conflict between any two indicators in the comprehensive indicator set to form a conflict degree matrix; From the conflict degree matrix, identify one or more sets of indicators whose conflict degree values ​​exceed a preset conflict degree threshold and are related to each other. Construct an undirected graph based on the conflict degree matrix between the indicators, and use a community detection algorithm to define the corresponding indicators as core conflict clusters. The weighting of the indicators contained in the core contradiction cluster is adjusted to a first preset value, and the weighting of the indicators contained in the non-core contradiction cluster is adjusted to a second preset value, thereby generating a contradiction focus weight vector adapted to the current candidate landing point; wherein, the first preset value is greater than the second preset value.

[0010] Optionally, calculating the conflict degree between any two indicators in the comprehensive indicator set to form a conflict degree matrix includes: Obtain the sample value sequence from the historical engineering data or simulation scenario data of any two of the aforementioned indicators; The conditional probability of the second indicator being inferior when the first indicator sample value is excellent is calculated to obtain the degree of conflict from the first indicator to the second indicator. Calculate the conflict degree from the second indicator to the first indicator, and take the maximum value of the conflict degree in the two directions as the final conflict degree between the two indicators, and fill it into the conflict degree matrix.

[0011] Optionally, the step of using the contradiction focus weight vector to execute a first screening path based on spatial distance and a second screening path based on improved trend correlation for each candidate landing point to obtain a first candidate subset and a second candidate subset includes: In the first screening path based on spatial distance, the ideal solution method weighted by the contradiction focus weight vector is adopted to calculate the distance between each candidate landing point and the positive and negative ideal solutions, and the first number of schemes are selected according to the distance to form the first candidate subset; In the second screening path based on the improved trend correlation, the improved path from the current state to the ideal state of each candidate landing point is analyzed, the trend correlation between each improved path and a benchmark efficient improved path is calculated, and the second number of schemes are selected according to the magnitude of the trend correlation to form a second candidate subset. The first screening path and the second screening path are executed independently and in parallel. When executing the second screening path, the correlation grouping algorithm is used to cluster and classify the candidate landing points into core selection group, potential enhancement group and risk avoidance group.

[0012] Optionally, the step of performing coupled verification on the first candidate subset and the second candidate subset, and using the intersection result as the determined recommendation candidate set, includes: Determine whether there is any intersection between the first candidate subset and the second candidate subset; If so, the intersection will be directly used as the candidate set for recommendation; If not, an evaluation uncertainty alarm is triggered, and a review process is initiated; wherein, the review process includes re-verifying the input data or re-identifying the core contradiction clusters.

[0013] Optionally, the step of obtaining a geographic information map containing power grid development planning information, and generating a landing point decision recommendation package containing flexible and adaptable spatial areas and a risk warning list based on the determined recommendation candidate set and the geographic information map, includes: Taking the landing point location in the determined recommended candidate set as the center, spatial offset stability analysis is performed based on the geological survey data to determine the tolerance for geological condition fluctuations, and the flexibility of project layout is determined according to the site construction requirements in the engineering economic parameters. Taking into account the tolerance for geological condition fluctuations and the flexibility of engineering layout, a continuous area is delineated in the geographic information map as a flexible and adaptable spatial area. For each candidate landing point whose geographical coordinates fall within the range of the elastic adaptable space region, extract the core contradiction cluster; The core contradiction cluster is transformed into specific engineering risk descriptions and corresponding mitigation measures suggestions, generating a risk warning list; The geographical boundary information of the flexible and adaptable spatial area is bound to the risk warning list to form a landing point decision recommendation package.

[0014] Optionally, after the step of generating the landing point decision recommendation package containing a flexible and adaptable spatial area and a risk warning list, the method further includes post-evaluation feedback correction of the evaluation system; including: Collect actual operational performance data from the locations of operational DC power projects; The actual operational performance data is compared with the predicted data generated during the pre-commissioning evaluation phase of the project to calculate the prediction deviation pattern. Based on the predicted deviation pattern, the identification rules for the quantification influence function library or core contradiction cluster are adaptively corrected.

[0015] This application has the following beneficial effects: The method proposed in this application employs a unified geological-electrical quantitative mapping, dynamic contradiction focus identification, and bidirectional independent screening coupled verification mechanism, which can realize adaptive decision support for DC landing point selection in complex geological environments, thereby improving the scientific nature, pertinence, and engineering reliability of the site selection evaluation. Attached Figure Description

[0016] Figure 1 A flowchart illustrating a multi-index comprehensive evaluation method for DC landing points constrained by karst topography, provided in an embodiment of this application; Figure 2 A heatmap of the conflict degree matrix, a comprehensive evaluation index, is provided for embodiments of this application. Figure 3 A line graph comparing the dual-path evaluation scores of candidate landing points provided in the embodiments of this application. Detailed Implementation

[0017] To facilitate understanding by those skilled in the art, the present application will be further described below in conjunction with embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present application.

[0018] To solve the above technical problems, such as Figure 1 As shown, this application proposes a multi-index comprehensive evaluation method for DC landing points constrained by karst topography, including: Step S101: Obtain geological survey data, power grid structure parameters, and engineering economic parameters for multiple candidate landing sites located in karst landform areas; The geological survey data in this application include karst development rate, soil cover thickness, groundwater depth and topographic slope. The power grid structure parameters include candidate short-circuit capacity, DC rated power, existing DC feed-in rated power and multi-feed-in interaction factors. The engineering economic parameters include candidate short-circuit capacity, DC rated power, existing DC feed-in rated power and multi-feed-in interaction factors.

[0019] Step S102: Based on the geological survey data, power grid structure parameters, and engineering economic parameters, generate the contradiction focus weight vector of the current candidate landing point; Based on the aforementioned geological survey data, power grid structure parameters, and engineering economic parameters, calculations are performed to generate the conflict focus weight vector for the current candidate landing point. The calculation process is as follows: First, based on the geological survey data, the set of geological-electrical impact scores for each candidate landing point is calculated, as follows: Several key geological factors were extracted from geological exploration data and represented as follows: Each key geological factor is input into the corresponding nonlinear influence function in the quantitative influence function library for calculation, generating an independent risk quantitative score for each key geological factor, including three dimensions: construction feasibility, long-term safety, and full-cycle economics. As shown in the following formula: (1) in, Representing the The quantitative values ​​of key geological factors, such as the measured percentage of karst development rate at a certain landing point, are derived from the digital extraction of the original borehole data or geophysical profile data of the current candidate landing point. Representing the The factor in the th The risk threshold for each evaluation dimension is a feature inflection point value that leads to a surge in loss in that dimension in the historical sample set. It is derived from the model intercept parameter obtained by supervised learning fitting in the quantification influence function library. Representing the The risk sensitivity coefficient of a key geological factor, which characterizes the growth rate of geological disaster risk as the factor value increases, is derived from the model slope parameter determined by regression analysis in the quantitative influence function library. It is a natural constant.

[0020] It should be noted that the construction of the quantitative impact function library in this application is based on in-depth mining and physical meaning mapping of historical engineering samples. Firstly, in the pre-construction phase of the quantitative impact function library, a rigorous "geological feature-engineering consequence" alignment mechanism is established. Survey reports and operation and maintenance records of historical DC engineering projects in karst areas are collected. For each recorded engineering disturbance event, such as foundation settlement caused by karst cave collapse or slope displacement caused by soil-rock interface instability, a sample alignment operation is performed: extracting the geological feature vector under the precise coordinates of the event location. This includes measured values ​​of factors such as karst development rate, soil cover thickness, and groundwater depth, and simultaneously obtains the equivalent number of days of construction delay or the equivalent economic loss caused by the event. Through normalization, the loss data is transformed into risk quantification parameters. This forms a supervised learning sample set of "geological feature vector - risk target value". Next, nonlinear fitting and parameter solidification are performed. A nonlinear regression algorithm, such as gradient boosting regression or ridge regression, is used to train the sample set. By minimizing the deviation between predicted risk and actual loss, a preliminary mapping relationship from geological features to risk level is established. Based on this, in order to transform complex engineering impacts into standardized evaluation scores, this application solidifies the fitted model parameters into a logistic mapping function, forming a nonlinear impact function ultimately stored in a quantified impact function library.

[0021] Then, the independent risk quantification scores of all key geological factors for each candidate landing point on the same dimension are aggregated to generate the geological-electrical impact score of that candidate landing point, as shown in the following formula: (2) in, For the first The factor in the th The weights for each dimension are derived from historical sample contribution analysis or fixed weight coefficients set by expert experience, and satisfy the following conditions: .

[0022] By combining the different geological-electrical impact scores of the aforementioned candidate landing points, the set of geological-electrical impact scores for that candidate landing point can be obtained. Through the complete engineering path described above, from "sample alignment" to "parametric mapping," unstructured karst geological constraints are transformed into rigorous and reproducible quantitative evaluation indicators.

[0023] Based on this, an equivalent power system model incorporating current candidate connection schemes is constructed using grid structure parameters. Electrical performance indicators are obtained through power flow and short-circuit calculations. Specifically, for calculating the effective short-circuit ratio of multiple infeeders... ,have: (3) in, The short-circuit capacity at candidate landing point i is typically between 5000 and 30000 MVA, depending on the voltage level and grid structure of the grid. For this landing point, the rated DC power is typically 3000 to 12000 MW for conventional DC projects; The rated power of the j-th existing DC feed point is typically 3000 to 12000 MW. Let be the multi-feedback interaction factor between landing point j and landing point i. This is used to calculate the switching failure immunity index for candidate landing point i. ,have: (4) in, To determine the critical voltage drop amplitude that causes commutation failure at candidate landing point i, the single-phase or three-phase voltage drop depth applied at the converter bus of candidate landing point iii is gradually increased from small to large in the electromagnetic transient simulation platform, and the voltage drop value corresponding to the first trigger of commutation failure is recorded. This is the rated voltage.

[0024] Simultaneously, by combining engineering economic parameters and converting the total life-cycle cost into a comprehensive levelized cost of electricity (LCOE), the comprehensive LCOE cost for candidate landing point i is calculated. ,have: (5) in, The total construction investment includes additional geological costs. This is the capital recovery coefficient; The annual operation and maintenance cost can be calculated as 1.5% to 3.0% of the total construction investment. The annual power transmission is calculated based on the above-mentioned multi-infeed effective short-circuit ratio, commutation failure immunity index, and comprehensive cost per kilowatt-hour, which together constitute the set of electrical performance indicators. These indicators, along with the geological-electrical influence score set, are used for subsequent conflict degree calculations.

[0025] For example, the calculation process for the above-mentioned geological-electrical impact score set is as follows: Assuming a candidate location in a karst landform region, key geological factors are first extracted from geological survey data, and the extracted karst development rate factor is then quantified. The threshold is 15%, or 0.15. The nonlinear impact function matching this landform subclass is used from the quantification impact function library. When evaluating the feasibility of construction, a risk threshold derived from fitting historical data from the library is set. The risk sensitivity coefficient is 0.1. The value is 20, which is substituted into the Logistic mapping function. Perform the operation; the calculation process is as follows: The independent risk quantification score of this factor for the construction feasibility dimension is 0.73. Similarly, the score of the surface subsidence level factor for this dimension is calculated. The score for this dimension is 0.6, representing the topographic slope factor. The score is 0.4. Finally, a multi-dimensional aggregation calculation is performed to generate the construction feasibility impact score. Set the aggregation weight coefficients for each factor. The values ​​are 0.5, 0.3, and 0.2 respectively. Apply the formula... get This will enable a quantitative evaluation of the construction feasibility dimension within the geological-electrical impact score set for this location.

[0026] Based on the aforementioned set of geological-electrical impact scores and electrical performance indicators, the conflict degree between indicators is calculated to generate the conflict focus weight vector for the current candidate landing point. The calculation process is as follows: To achieve accurate identification of the core contradictions for each candidate landing point and scientific allocation of dynamic weights, a comprehensive index set for the current candidate landing point is first constructed. This set consists of three geological-electrical impact scores and electrical performance indicators obtained based on power grid simulation.

[0027] To quantify the interrelationships between indicators, calculate the relationship between any two indicators in the comprehensive indicator set. and The degree of conflict between During the calculation, sample value sequences of two indicators are obtained from historical engineering datasets or simulation scenario sequences. The conditional probability of the second indicator being inferior when the first indicator's sample value is superior is calculated to obtain the conflict degree from the first indicator to the second indicator. The conflict degree from the second indicator to the first indicator is then calculated, and the maximum value of the conflict degrees in both directions is taken as the final conflict degree between the two indicators and filled into the conflict degree matrix. The specific calculation formula is as follows: (6) in, and These represent different evaluation items in the comprehensive indicator set, and their values ​​are derived from the real-time calculated values ​​of the current landing point. This represents the conditional probability function in the historical sample space; and The state is determined by the upper and lower quartiles of the historical statistical distribution; that is, the optimal state refers to an index that is in a favorable position for the system. In a given range, a poorly performing state refers to an indicator being in an unfavorable position. The data for this interval is supported by actual operation records of already operational projects stored in the quantification influence function library.

[0028] To better understand the above process of calculating the conflict degree matrix, an example is provided here for more detailed explanation: First, based on the actual operational performance data and simulation scenario data of the already operational projects, a standardized sample sequence is established for each indicator in the comprehensive indicator set. For any two evaluation indicators... and The essence of its conflict degree is to quantify the risk probability of one side increasing while the other decreases. In the calculation, the anchoring indicator is first... The superior performance range, that is, the range within the statistical sample that is favorable for system operation. A subset of quantiles is defined as Status; synchronous observation within this subset at the same timestamp or in the same project sample, indicators In a position detrimental to system operation The frequency of the quantile interval, i.e. Status. This statistical frequency constitutes the indicator. Indicator The probability of one-way conflict. Because the constraints between indicators are often asymmetric in complex geological environments, the above process must be repeated in reverse to statistically determine the probability of conflict. In State In The conditional probability of the state. Finally, the maximum value is extracted from the conditional probability values ​​in these two directions as the final conflict degree scalar characterizing the intensity of the essential conflict, and this value is filled into the corresponding coordinates of the conflict degree matrix. middle.

[0029] Fill the conflict degree matrix with the calculated conflict degree, such as... Figure 2 As shown, different shades of grayscale parameters visually present the strong coupling conflict between karst geological constraints and power grid operation performance, providing a data foundation for identifying core contradiction clusters.

[0030] Subsequently, a group or multiple groups of indicators whose conflict degree values ​​exceed a preset conflict degree threshold and are interconnected are identified from the conflict degree matrix. An undirected graph is then constructed based on the conflict degree matrix between the indicators, and a community detection algorithm is used to define the corresponding indicators as core conflict clusters. This identification process involves determining the matrix elements... Implementation, in which To preset the conflict degree threshold, its value is usually set between 0.7 and 0.9, derived from the statistical mean of the conflict patterns of strongly coupled indicators in historical engineering failure cases; the so-called interrelated set of indicators refers to those that are related in the conflict degree matrix. Construct an undirected graph with edge weight thresholds and use the community detection algorithm to identify the largest connected subgraph or highly coupled index cluster.

[0031] Finally, the weighting of indicators contained in the core contradiction cluster is adjusted to a first preset value, and the weighting of indicators contained in non-core contradiction clusters is adjusted to a second preset value, generating a contradiction focus weight vector adapted to the current candidate landing point; wherein, the first preset value is greater than the second preset value. The original weights of the indicators contained in the core contradiction cluster are then corrected using a normalization correction strategy based on conflict degree weighting, resulting in corrected indicator weights. The calculation formula is: (7) in, Representing the The final weights of each indicator after dynamic adjustment; The original initial weight representing this indicator comes from the preset empirical values ​​of the power industry planning and design standards; This represents the set of indicators representing the core contradiction clusters identified; Representative indicators Other indicators within the contradictory cluster Measured values ​​of the degree of conflict between them; This represents the total number of indicators in the comprehensive indicator set. In this correction strategy, the numerator of the core contradiction cluster indicator is multiplied by an amplification factor. This allows for proactive improvement. The numerators of non-core contradiction cluster indicators remain unchanged, while their weights are compressed through denominator normalization, ultimately ensuring that both satisfy the constraint that the sum of their weights equals 1. This process amplifies the contribution of core contradiction indicators to decision-making, enabling the evaluation results to focus on the most severe constraints under karst topography, thus achieving personalized weighting adapted to specific geographical environments.

[0032] To facilitate understanding, a simple example is provided to illustrate the calculation process above, which involves calculating the conflict degree between indicators based on the geological-electrical influence score set and the electrical performance index set, and generating the conflict focus weight vector for the current candidate landing point: For a given candidate landing site, a comprehensive set of indicators is first constructed, including the construction feasibility impact score, the full-cycle economic impact score, and the effective short-circuit ratio of multiple infeeds. The construction feasibility indicator... Economic indicators For example, obtain its distribution in the historical engineering sample sequence to determine The optimal range, i.e., the top 25% and The worst-case scenario range, i.e., the bottom 25%. Statistical analysis shows that... In the best scenario The conditional probability of being inferior The reverse conditional probability is 0.85. The value is 0.70; the maximum value is used to determine the degree of conflict between the two indicators. The value is 0.85. This value exceeds the preset conflict threshold. It was identified as a core cluster of contradictions. Then the original weights of this indicator. To make corrections, assume that the measured values ​​of the degree of conflict between this indicator and other indicators within the core contradiction cluster are... The sum of the weighted conflict correction terms for all indicators is 0.85. The calculated value is 1.25. Substituting this value into the weight correction formula... Through this calculation, the weight of the core contradiction indicator was dynamically increased from 0.2 to 0.296, thus achieving an automatic focus of evaluation on the limiting factors.

[0033] Step S103: Using the contradiction focus weight vector, execute the first screening path based on spatial distance and the second screening path based on improved trend correlation for each candidate landing point to obtain the first candidate subset and the second candidate subset; The calculation process for this step is as follows: To achieve a multi-dimensional scientific evaluation of candidate landing points, an independent and parallel dual-path screening strategy is adopted to enhance the robustness of the decision. In the first screening path based on spatial distance, an ideal solution method weighted by a contradiction focus weight vector is used to calculate the distance between each candidate landing point and the positive and negative ideal solutions. The solutions are then sorted according to the distance to select the first number of options, forming the first candidate subset. The specific process is as follows: In the first selection path, a static superiority evaluation is performed using a ranking method that approximates the ideal solution by weighting the contradiction focus weight vector. A weighted normalized evaluation matrix is ​​constructed using the following formula: (8) in, Representing the The candidate solution is in the... The weighted normalized value of each indicator The total number of candidate landing points participating in the sorting. The first in the dynamically generated contradiction focus weight vector Each component. Then, the positive ideal solution vector is determined. With negative ideal solution vector Calculate the distance from each solution to the ideal solution: and (9) Then calculate the relative closeness. (10) in, The larger the value, the better the solution. Sort by distance and select the top number of solutions from the highest to the lowest relative proximity to form the first candidate subset. The first number is 20% to 30% of the total number of candidate landing points (rounded to the nearest integer), and no less than 3, to ensure that the subset has sufficient solution diversity and takes into account the effectiveness of subsequent coupling verification.

[0034] In the second screening path based on the improvement trend correlation, the improvement path from the current state to the ideal state for each candidate landing point is analyzed. The trend correlation between each improvement path and a baseline efficient improvement path is calculated. The second candidate subset is then selected by ranking the trend correlation based on the magnitude of the correlation. The specific calculation process is as follows: First, define the baseline efficient improvement path vector. It originates from the historical best performance index improvement trajectory before the commissioning of DC projects, stored in the quantification influence function library. Calculate the... The improvement path and the baseline path of the scheme are in the first... Correlation coefficients on each indicator : (11) in, Represents the absolute difference. and The resolution coefficient is the minimum and maximum absolute difference between all schemes across all indicators. The value is set to 0.5. Calculate the trend correlation: (12) in, To evaluate the total number of indicators, they are sorted according to their trend correlation. The top two numbers of solutions, ranked from highest to lowest trend correlation, are selected to form a second candidate subset. This second subset can be the same as the first, ensuring fair cross-validation of the two independent screening paths with an equal number of solutions. By executing the two paths in parallel, not only is the solution with the best overall performance identified, but the solution with the highest alignment between development trends and the ideal model is also considered. Figure 3 As shown, the comprehensive scores of each candidate landing point under the first path and the second path are compared. By comparing the fluctuation trends of the two curves, it is demonstrated that the present invention, while locking in the current optimal solution, also takes into account solutions with high improvement potential.

[0035] To facilitate understanding of the process of generating the first and second candidate subsets, the following example is provided: For example, for a set of candidate landing point schemes, the indicators are first weighted using a dynamically generated contradiction focus weight vector, with the first... Taking the first scheme as an example, in the... Normalized measured values ​​for each indicator The value is 0.8, corresponding to the weighted component. Given a value of 0.3, and assuming the root sum of squares for all possible solutions to this index is 2, substituting this into the weighted normalization formula yields... In the first selection path, the solution is calculated to the positive ideal solution. The distance, assumed to be 0.15, to the negative ideal solution The distance, assuming to be 0.05, if calculated... and The relative closeness Based on this value, the first candidate subset is selected and sorted. In the second selection path, a baseline value for the benchmark efficient improvement path on this indicator is defined. If it is 0.9, then the absolute difference Assuming the maximum difference among all options The minimum difference is 0.4. The resolution coefficient is 0.02. Taking 0.5 and substituting it into the correlation coefficient formula, we get... Assuming there are 5 evaluation indicators, and the sum of their correlation coefficients is 3.5, then the trend correlation is... The second candidate subset is selected by sorting them in this way, and the parallel evaluation of the two paths is completed.

[0036] It should be noted that the first screening path and the second screening path are executed independently and in parallel. When executing the second screening path, the correlation grouping algorithm is used to cluster and classify the candidate landing points into core selection group, potential enhancement group and risk avoidance group.

[0037] Specifically, to simplify and categorize the candidate solution set, this invention uses a correlation-based grouping algorithm to divide the initially selected candidate solutions into a core optimization group, a potential enhancement group, and a risk avoidance group. During algorithm execution, the algorithm first obtains the correlation coefficients of each solution within the intersection of two selection subsets. Based on the scores of each evaluation metric, a classification vector is constructed. For each solution, the relative offset feature value between it and the baseline efficient improvement path vector is calculated: (13) in, Representing the Overall path offset strength of each scheme; Representing the The first scheme is in the Normalized measured values ​​for each indicator; The representative is derived from the historical best engineering trajectory in the quantization influence function library. Benchmark values ​​for each indicator; The total number of evaluation metrics is given. Subsequently, density-based clustering logic is used to determine the scheme's affiliation, and the dynamic clustering radius determined by the hyperparameter adaptive mechanism is used as the neighborhood discrimination threshold. (14) in, The dynamic grouping radius in the current iteration; Represents the feature values ​​of all schemes in the current candidate set. Standard deviation; The preset aggregation coefficient is initially set to 0.5 based on the sample aggregation degree of historical successful site selection cases, and is adjusted during iteration according to the sample density. Perform adaptive adjustment: if the density of schemes in the current neighborhood... If it is below the historical average, then Automatically increase by 10% to relax aggregation conditions, and conversely decrease by 5%. Determine if the feature distance between schemes is less than [value missing]. This grouping algorithm clusters schemes with similar risk characteristics into different original clusters. During execution, the correlation grouping algorithm considers the density of schemes within a neighborhood. The iteration stops when the system stabilizes or reaches a preset maximum number of iterations, such as 100. Finally, the trend correlation of the proposed solutions is considered. From the mass center of each cluster to the positive ideal solution The proximity is graded using multiple criteria. The specific grading rule is defined as follows: cluster centroid proximity greater than a preset preferred threshold and average offset feature value are considered. The smallest cluster is defined as the core preferred group. In this embodiment, the preset preferred threshold value ranges from 0.8 to 0.9; the average trend correlation is... Clusters with scores greater than 0.8 but currently limited are defined as the potential enhancement group; the remaining schemes are classified as the risk aversion group. This process, through the structured hierarchical classification of discrete schemes, ensures that subsequent coupled verification can focus on the target group with the greatest engineering implementation value.

[0038] Step S104: Perform coupling verification on the first candidate subset and the second candidate subset, and use the intersection result as the recommended candidate set; The system determines whether there is an intersection between the first candidate subset and the second candidate subset. If so, the intersection is directly used as the recommended candidate set. If not, an evaluation uncertainty alarm is triggered, and a review process is initiated. This review process includes re-verifying the input data or re-identifying the core contradiction clusters. Specifically, to ensure the reliability of the recommended scheme under multiple criteria, this invention performs coupled verification by performing an intersection operation on the first and second candidate subsets. The coupled verification process involves comparing the candidate landing point identifier set in the first candidate subset with the candidate landing point identifier set in the second candidate subset one by one, extracting candidate landing points that exist simultaneously in both subsets, and forming the intersection result. Since the first screening path focuses on evaluating the superiority of the candidate landing point relative to the static spatial distance of the positive and negative ideal solutions, while the second screening path focuses on evaluating the dynamic trend fit of the candidate landing point from the current state to the ideal state, the two paths independently provide screening conclusions from different evaluation dimensions. Candidate landing points selected by both paths simultaneously possess both superior current comprehensive performance and good future improvement potential, thus having higher decision credibility. If the intersection result is not empty, the intersection is directly determined as the final recommendation candidate set; if the intersection result is empty, it indicates that there is a significant discrepancy in the evaluation conclusions of the two independent paths, triggering an evaluation uncertainty alarm and initiating a review process. The review process first assesses the dispersion of the evaluation results by calculating the comprehensive evaluation confidence coefficient: (15) in, The overall confidence level representing the current candidate set; For the first The relative similarity of each option in the first screening path; This represents the trend correlation obtained by the scheme in the second screening path; This represents the total number of candidate landing points participating in the coupling verification. If If the threshold of 0.3 is exceeded, the core contradiction cluster is re-identified, and the stability of the indicator input is verified by calculating the characteristic volatility of the current sample. (16) in, The standard deviation of characteristic fluctuations of the representative indicator sample; Normalized measured values ​​of representative indicators; Representing the The average value of each indicator in the current subset, with data sourced from real-time acquired power grid structure parameters and geological survey data; This represents the total number of indicators. If the characteristic volatility is abnormal, the process will revert to the data input stage for re-verification, or the conflict threshold will be adjusted. The weight vectors are regenerated until a non-empty intersection recommendation set is produced. This mechanism, through logical mutual exclusion checks and parameter backtracking, avoids decision bias caused by single algorithmic pitfalls or data noise, ensuring high reliability of the output results.

[0039] For example, to ensure the reliability of the recommendation scheme under multiple criteria, coupled verification is performed using subsets generated by two screening paths, assuming the total number of candidate landing points participating in the verification. There are 5. First, the dispersion of the results between the first and second candidate subsets is evaluated by calculating the comprehensive evaluation confidence coefficient. If the... The relative similarity of the solutions obtained in the first screening path Correlation with the trend obtained in the second path The absolute difference sequences are 0.35, 0.28, 0.40, 0.32, and 0.30, respectively. Substituting these values ​​into the formula... get Since the confidence coefficient exceeded the preset review threshold of 0.3, and the intersection of the two subsets was empty, the review process was automatically initiated by calculating the characteristic volatility of the current sample. To verify the stability of the indicator input. Taking a key indicator as an example, its normalized measured values ​​in 5 scenarios. The values ​​were 0.8, 0.6, 0.7, 0.5, and 0.9, respectively, and their average values ​​were calculated. The value is 0.7, assuming a total of The three core indicators have a sum of squared deviations of 0.1. Substituting these values ​​into the formula... get If the volatility is abnormal, the process should be reverted to the data input stage for re-verification, or the conflict threshold should be adjusted. The weight vectors are regenerated until a non-empty intersection recommendation set is produced, thereby effectively avoiding decision bias caused by data noise.

[0040] Step S105: Obtain a geographic information map containing power grid development planning information, and based on the determined recommended candidate set and the geographic information map, generate a landing point decision recommendation package containing flexible and adaptable spatial areas and a risk warning list.

[0041] To provide decision-makers with a final deliverable that is feasible for engineering implementation, this application uses spatial analysis technology to generate a landing site recommendation package that includes flexible and adaptable spatial areas and a risk warning list. The generation process is as follows: Centered on the location of the landing point in the recommended candidate set, spatial migration stability analysis is performed based on geological survey data to determine the tolerance for geological condition fluctuations, and the flexibility of the project layout is determined according to the site construction requirements in the engineering economic parameters. The use of geological survey data here differs from that in step S102. In step S102, geological survey data was extracted as key geological factors and input into a quantitative influence function library to generate influence scores reflecting the geological risk level of the candidate landing point itself. In this step, a new spatial analysis calculation is performed on the geological survey data: using data from multiple survey points around the recommended landing point, a spatially continuous distribution surface of geological factors is constructed through Kriging interpolation, and the spatial gradient value at the coordinates of the recommended landing point is calculated to quantify the rate of change of geological conditions when migrating from that landing point to the surrounding areas. This spatial gradient is a new parameter not calculated in the previous steps and is specifically used for delineating the boundary of the flexible adaptable spatial region in this step. Using the geographic coordinates of each landing point in the final recommended candidate set as the geometric center, a buffer zone based on geological homogeneity is delineated on the geographic information map to determine the flexible adaptable spatial region. The delineation rules for this area are based on the tolerance for fluctuations in geological conditions. With flexibility in project layout Jointly driven. Among them, the tolerance for fluctuations in geological conditions: (17) in, This represents the permissible geographical offset distance under the premise that geological factors do not change drastically. The preset threshold for geological factor fluctuations is typically set at 5% of the quantified value of key geological factors, derived from the maximum tolerance of historical engineering projects for fluctuations in foundation bearing capacity. The spatial gradient of geological factors around the current landing point is calculated as follows: First, the limited geological exploration point data is spatially continuous using the Kriging interpolation method to construct a spatial distribution surface of geological factors. Then, the spatial first-order partial derivative of this distribution surface at the coordinates of the current landing point is calculated using the ternary neighborhood difference method. The magnitude of the resulting vector is the spatial gradient value. The data comes from the spatial interpolation analysis of geological exploration data in the geographic information system.

[0042] Project layout flexibility: (18) This represents the minimum spatial radius required to meet the construction and future expansion needs of the converter station; The standard construction area of ​​the converter station (site construction requirements in the engineering economic parameters) is derived from the technical indicators in the power design code. This is the area redundancy coefficient, typically ranging from 0.1 to 0.2, used to compensate for construction errors and expansion needs. The minimum of these two values ​​is ultimately taken as the elastic buffer radius. (19) And define a radius of with the landing point as the center on the geographic information map. A circular or terrain-corrected continuous polygonal region is used as a flexible and adaptable spatial area.

[0043] Subsequently, for each candidate landing point obtained in the initial step whose geographical coordinates lie within the defined flexible and adaptable spatial area—including landing points in the final recommended candidate set serving as the regional center and other candidate landing points falling into the area—the contradictory relationships marked in the core contradiction cluster identification step are extracted. These relationships are then mapped into specific engineering risk language and corresponding mitigation measures, generating a risk warning list. For example, if the core contradiction cluster includes karst development rate and construction cost, it is transformed into "Karst Foundation Treatment Risk and Investment Overrun Warning" and matched with corresponding mitigation measures, such as "Recommend increasing geophysical exploration density and reserving special grouting costs."

[0044] By linking and binding the geographic geometric boundary information of the flexible and adaptable spatial area with a structured risk warning list, the evaluation results are not just isolated coordinate points, but a decision recommendation package with in-depth engineering insights and implementation flexibility.

[0045] To facilitate understanding, the process of generating the landing point decision recommendation package described above is illustrated as follows: Using a specific landing point location in the final recommendation candidate set as the center, a buffer zone based on geological homogeneity is delineated on the geographic information map to determine the flexible and adaptable spatial area. First, the tolerance for geological condition fluctuations is calculated. Set a preset threshold for geological factor fluctuations. This is 5% of the quantified value of the key geological factors at this landing point. If the quantified value of a certain geological factor is 0.8, then... The value is set to 0.04. A geological distribution surface is constructed using Kriging interpolation, and the spatial gradient at the current coordinates is calculated using the ternary neighborhood difference method. The value is 0.0002. Substitute it into the formula. get Meters. Next, calculate the flexibility of the project layout. The standard construction area of ​​the converter station is known. Set an area redundancy factor of 60,000 square meters. The value is 0.2, so we substitute it into the formula. get Meters. The minimum of the two values ​​is used to determine the elastic buffer radius. The area is defined as a flexible, adaptable space region with a radius of 1 meter. Then, the core contradiction clusters are extracted. If they include karst development rate and construction cost, they are transformed into karst foundation treatment risk and investment overrun warnings. Mitigation measures such as increasing geophysical exploration density and reserving funds for special grouting are then matched and suggested. A risk warning list is generated and bound to geographic boundary information to form a decision-making recommendation package.

[0046] After the landing point recommendation package is actually put into operation, the method also includes post-implementation evaluation feedback correction of the evaluation system: Specifically, to achieve continuous evolution and accuracy iteration of the evaluation system, this invention collects actual operational performance data from the completed DC power line projects, compares this data with the predicted data generated during the pre-commissioning evaluation phase (output of the earlier prediction model), and calculates the prediction deviation pattern to establish a closed-loop feedback correction mechanism. For a specific completed DC power line project, on the [date missing]... Calculate the prediction bias pattern based on the performance across each evaluation dimension: (20) in, Representing the Prediction bias rate in each dimension; This is the predicted composite score generated by multi-factor aggregation during the evaluation phase, and its value range is... ; This represents the normalized standard parameter representing the actual operational observations at the corresponding dimension after the landing point is put into operation. The calculation method for the normalized standard parameter is as follows: First, obtain the actual physical observations after the landing point is put into operation, such as the actual settlement displacement of the converter station foundation or the actual additional investment amount due to geological defects. Then, call the corresponding nonlinear influence function in the quantification influence function library, and substitute the actual physical observations as independent variables into the formula. Inverse mapping is performed to transform the actual losses of different dimensions into predicted scores. Having the same dimensions Interval standard parameters. This data comes from the project completion settlement report and the long-term on-site monitoring system; if If the deviation exceeds a preset tolerance threshold, adaptive correction of the system's underlying rules (the identification rules for the quantified influence function library or the core contradiction cluster) is triggered. When the deviation pattern reflects a systematic underestimation or overestimation of the quantified influence function library, the parameters in the nonlinear influence function are corrected. Taking the adjustment of the risk sensitivity coefficient as an example, the update formula is as follows: (twenty one) in, It is a revised risk sensitivity coefficient stored in the quantification impact function library; It is the original risk sensitivity coefficient; The preset learning rate factor is typically between 0.01 and 0.05, derived from experimental settings that require model stability. The sign function is used to determine the prediction direction. If the deviation pattern points to a deviation from the core contradiction cluster identification rule, then the conflict degree threshold is applied. The formula was optimized and adjusted as follows: (twenty two) in, It is the corrected conflict threshold; It is the original preset conflict threshold; To adjust the step size constant; The standard deviation of the classification residuals, representing the errors caused by identification in historical evaluation samples, is derived from the statistical analysis of the deviation distribution of operational samples. Through the aforementioned closed-loop correction process, practical operational experience can be continuously incorporated, enabling the self-evolution of evaluation parameters and enhancing the reliability of DC landing point evaluation in complex karst topography.

[0047] The aforementioned core steps—from data acquisition, geological risk quantification, conflict identification and weight generation, dual-path screening, coupled verification, to decision package generation—together constitute the main evaluation process, used to perform a one-time comprehensive evaluation and decision recommendation for the current candidate landing points. The adaptive correction mechanism described in this step is independent of the main process and is a closed-loop feedback process executed after the main process is completed and the recommended scheme has been put into actual operation. The correction targets are: first, the quantification influence function library relied upon by the geological risk quantification step in the main process, which eliminates systematic biases in geological risk prediction by adjusting its nonlinear influence function parameters; and second, the core conflict cluster identification rules relied upon by the conflict identification step in the main process, which correct erroneous identification of conflict clusters by optimizing the conflict degree threshold or grouping algorithm hyperparameters. The corrected parameters are then updated and overwritten to the corresponding modules for use in subsequent new evaluation tasks.

[0048] In summary, the method proposed in this application, which employs a unified geological-electrical quantitative mapping, dynamic contradiction focus identification, and bidirectional independent screening coupled verification mechanism, can achieve adaptive decision support for DC landing point selection in complex geological environments, thereby improving the scientific nature, pertinence, and engineering reliability of the site selection evaluation.

[0049] In some embodiments, this application also provides a computer system including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0050] This application also provides a computer-readable storage medium for storing a computer program. This computer-readable storage medium can be applied to a computer device, and the computer program causes the computer device to execute the corresponding processes in the methods described above in the embodiments of this application; for brevity, further details are omitted here.

[0051] The above embodiments are preferred implementations of this application. In addition, this application can be implemented in other ways. Any obvious substitutions without departing from the concept of this technical solution are within the protection scope of this application.

[0052] To facilitate understanding by those skilled in the art of the improvements made by this application compared to the prior art, some of the accompanying drawings and descriptions have been simplified, and for clarity, some other elements have been omitted from this application. Those skilled in the art should realize that these omitted elements may also constitute the content of this application.

Claims

1. A multi-index comprehensive evaluation method for DC landing points constrained by karst topography, characterized in that, include: Obtain geological survey data, power grid structure parameters, and engineering economic parameters for multiple candidate landing sites located in karst landform areas; Based on the geological survey data, power grid structure parameters, and engineering economic parameters, a contradiction focus weight vector for the current candidate landing point is generated. Using the aforementioned contradiction focus weight vector, a first screening path based on spatial distance and a second screening path based on improved trend correlation are executed for each candidate landing point to obtain a first candidate subset and a second candidate subset. The first candidate subset and the second candidate subset are coupled and verified, and the intersection result is used as the recommended candidate set. Obtain a geographic information map containing power grid development planning information, and based on the determined recommended candidate set and the geographic information map, generate a landing point decision recommendation package containing flexible and adaptable spatial areas and a risk warning list.

2. The method according to claim 1, characterized in that, The step of generating a conflict focus weight vector for the current candidate landing point based on the geological survey data, power grid structure parameters, and engineering economic parameters includes: Based on the geological survey data, the set of geological-electrical impact scores for each candidate landing point was calculated; Based on the power grid structure parameters and engineering economic parameters, the set of electrical performance indicators for each candidate landing point is calculated. Based on the geological-electrical impact score set and the electrical performance index set, the conflict degree between the indices is calculated to generate the conflict focus weight vector of the current candidate landing point.

3. The method according to claim 2, characterized in that, Based on the geological survey data, a set of geological-electrical impact scores for each candidate landing point was calculated; including: Several key geological factors were extracted from the geological exploration data; Each key geological factor is input into the corresponding nonlinear influence function in the quantitative influence function library for calculation, and an independent risk quantitative score is generated for each key geological factor, including three dimensions: construction feasibility, long-term safety and full-cycle economics. By aggregating the independent risk quantification scores of all key geological factors for each candidate landing point on the same dimension, a set of geological-electrical impact scores for that candidate landing point is generated.

4. The method according to claim 2, characterized in that, The set of electrical performance indicators for each candidate landing point is calculated based on the power grid structure parameters and engineering economic parameters; including: Based on the power grid structure parameters, an equivalent model of the power system including the current candidate landing point access scheme is constructed; Based on the power system equivalent model and the engineering economic parameters, the electrical performance index set for each candidate landing point is obtained through power flow calculation and short-circuit calculation.

5. The method according to claim 2, characterized in that, The step of calculating the conflict degree between indicators based on the geological-electrical impact score set and the electrical performance index set, and generating the conflict focus weight vector of the current candidate landing point, includes: Based on the geological-electrical impact score set and the electrical performance index set, a comprehensive index set for the current candidate landing point is constructed; Calculate the degree of conflict between any two indicators in the comprehensive indicator set to form a conflict degree matrix; From the conflict degree matrix, identify one or more sets of indicators whose conflict degree values ​​exceed a preset conflict degree threshold and are related to each other. Construct an undirected graph based on the conflict degree matrix between the indicators, and use a community detection algorithm to define the corresponding indicators as core conflict clusters. The weighting of the indicators contained in the core contradiction cluster is adjusted to a first preset value, and the weighting of the indicators contained in the non-core contradiction cluster is adjusted to a second preset value, thereby generating a contradiction focus weight vector adapted to the current candidate landing point; wherein, the first preset value is greater than the second preset value.

6. The method according to claim 5, characterized in that, The step of calculating the conflict degree between any two indicators in the comprehensive indicator set to form a conflict degree matrix includes: Obtain the sample value sequence from the historical engineering data or simulation scenario data of any two of the aforementioned indicators; The conditional probability of the second indicator being inferior when the first indicator sample value is excellent is calculated to obtain the degree of conflict from the first indicator to the second indicator. Calculate the conflict degree from the second indicator to the first indicator, and take the maximum value of the conflict degree in the two directions as the final conflict degree between the two indicators, and fill it into the conflict degree matrix.

7. The method according to claim 1, characterized in that, The first candidate subset and the second candidate subset are obtained by using the contradiction focus weight vector to perform a first screening path based on spatial distance and a second screening path based on improved trend correlation for each candidate landing point. include: In the first screening path based on spatial distance, the ideal solution method weighted by the contradiction focus weight vector is adopted to calculate the distance between each candidate landing point and the positive and negative ideal solutions, and the first number of schemes are selected according to the distance to form the first candidate subset; In the second screening path based on the improved trend correlation, the improved path from the current state to the ideal state of each candidate landing point is analyzed, the trend correlation between each improved path and a benchmark efficient improved path is calculated, and the second number of schemes are selected according to the magnitude of the trend correlation to form a second candidate subset. The first screening path and the second screening path are executed independently and in parallel. When executing the second screening path, the correlation grouping algorithm is used to cluster and classify the candidate landing points into core selection group, potential enhancement group and risk avoidance group.

8. The method according to claim 1, characterized in that, The step of performing coupled verification on the first candidate subset and the second candidate subset, and using the intersection result as the determined recommendation candidate set, includes: Determine whether there is any intersection between the first candidate subset and the second candidate subset; If so, the intersection will be directly used as the candidate set for recommendation; If not, an evaluation uncertainty alarm is triggered, and a review process is initiated; wherein, the review process includes re-verifying the input data or re-identifying the core contradiction clusters.

9. The method according to claim 1, characterized in that, The process of acquiring a geographic information map containing power grid development planning information, and generating a landing point decision recommendation package containing flexible and adaptable spatial areas and a risk warning list based on the determined recommendation candidate set and the geographic map, includes: Taking the landing point location in the determined recommended candidate set as the center, spatial offset stability analysis is performed based on the geological survey data to determine the tolerance for geological condition fluctuations, and the flexibility of project layout is determined according to the site construction requirements in the engineering economic parameters. Taking into account the tolerance for geological condition fluctuations and the flexibility of engineering layout, a continuous area is delineated in the geographic information map as a flexible and adaptable spatial area. For each candidate landing point whose geographical coordinates fall within the flexible and adaptable spatial region among multiple candidate landing points, the core contradiction cluster is extracted; The core contradiction cluster is transformed into specific engineering risk descriptions and corresponding mitigation measures suggestions, generating a risk warning list; The geographical boundary information of the flexible and adaptable spatial area is bound to the risk warning list to form a landing point decision recommendation package.

10. The method according to claim 1 or 9, characterized in that, The step of generating a landing point recommendation package that includes a flexible and adaptable spatial area and a risk warning list also includes post-evaluation feedback correction of the evaluation system; including: Collect actual operational performance data from the locations of operational DC power projects; The actual operational performance data is compared with the predicted data generated during the pre-commissioning evaluation phase of the project to calculate the prediction deviation pattern. Based on the predicted deviation pattern, the identification rules for the quantification influence function library or core contradiction cluster are adaptively corrected.

Images