Intelligent evaluation method and system for thickness of active layer of frozen soil

Abstract

The application relates to the technical field of data processing, and provides an intelligent evaluation method and system for the thickness of a frozen soil active layer, which comprises the following steps: obtaining basic information of frozen soil in a preset area; obtaining basic information of the environment of the frozen soil area; performing snowmelt prediction according to meteorological prediction information and snow cover information, obtaining soil water content prediction data, matching a heat conductivity coefficient in combination with frozen soil lithology information and frozen soil structure information; obtaining a freezing index or a thawing index; inputting the freezing index and the freezing period heat conductivity coefficient or the thawing index and the thawing period heat conductivity coefficient into an active layer depth change amount evaluation model to obtain an active layer depth change amount evaluation result, and determining frozen soil active layer thickness prediction data, so that the technical problem of low evaluation precision of the thickness of the frozen soil active layer is solved, the frozen soil thawing / freezing condition is comprehensively evaluated from the change of the surface temperature and the change of the snow / snowmelt, and the technical effect of improving the evaluation precision of the thickness of the frozen soil active layer is achieved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to an intelligent assessment method and system for the thickness of the active layer of permafrost. Background Technology

[0002] Permafrost refers to various rocks and soils below 0°C that contain ice, and is the surface layer of the earth with an average annual temperature of less than 0°C. The underlying layer is the perennial active layer. The active layer refers to the soil layer that covers permafrost and thaws in summer and freezes in winter. The seasonal changes in the thickness of the active layer have a serious impact on construction and survival.

[0003] Currently, the thickness of the active layer of permafrost is assessed in real time. Due to the complexity of analyzing changes in the active layer thickness, especially the complexity of snow cover environments, there is a lack of predictive assessment of changes in the active layer thickness. The key to construction and development depends on the characteristics of the permafrost foundation. Factors such as uneven permafrost foundations, large differences in loads, and complex shapes can cause foundation deformation. Before development in areas with permafrost distribution, the state of permafrost during the construction and use of buildings should be fully considered to provide a foundation for ensuring the durability of buildings.

[0004] In summary, the existing technology suffers from the technical problem of low accuracy in assessing the thickness of the active permafrost layer. Summary of the Invention

[0005] This application provides an intelligent assessment method and system for the thickness of the active permafrost layer, aiming to solve the technical problem of low assessment accuracy of the active permafrost layer in the prior art.

[0006] In view of the above problems, this application provides an intelligent assessment method and system for the thickness of the active layer of permafrost.

[0007] The first aspect of this application discloses an intelligent assessment method for the thickness of the active layer of permafrost, wherein the method includes: acquiring basic information about permafrost in a preset area, wherein the basic information about permafrost in the preset area includes permafrost lithology information and permafrost structure information; acquiring basic environmental information about the permafrost area, wherein the basic environmental information about the permafrost area includes snow cover information and meteorological forecast information; performing snowmelt prediction based on the meteorological forecast information and the snow cover information to obtain predicted soil moisture content data; matching thermal conductivity based on the permafrost lithology information, the permafrost structure information and the predicted soil moisture content data, wherein the thermal conductivity includes thermal conductivity during the thawing period or thermal conductivity during the freezing period; performing parameter statistics based on the meteorological forecast information to obtain a freezing index or a thawing index; inputting the freezing index and the thermal conductivity during the freezing period, or the thawing index and the thermal conductivity during the thawing period, into an assessment model for changes in the active layer depth to obtain an assessment result for changes in the active layer depth; and determining predicted data for the thickness of the active layer of permafrost based on the assessment result for changes in the active layer depth, wherein the predicted data for the thickness of the active layer of permafrost belongs to a first future time zone.

[0008] Another aspect of this application discloses an intelligent assessment system for the thickness of the active permafrost layer. The system includes: a permafrost information acquisition module for acquiring basic permafrost information of a preset area, wherein the basic permafrost information of the preset area includes permafrost lithology information and permafrost structure information; an environmental information acquisition module for acquiring basic environmental information of the permafrost area, wherein the basic environmental information of the permafrost area includes snow cover information and meteorological forecast information; a snowmelt prediction module for performing snowmelt prediction based on the meteorological forecast information and the snow cover information, and acquiring predicted soil moisture content data; and a thermal conductivity matching module for matching the permafrost lithology information and the permafrost structure information. The system is configured to: match the predicted soil moisture content data with thermal conductivity, wherein the thermal conductivity includes thermal conductivity during the thawing period or thermal conductivity during the freezing period; perform parameter statistics based on the meteorological forecast information to obtain a freezing index or a thawing index; input the freezing index and the thermal conductivity during the freezing period, or the thawing index and the thermal conductivity during the thawing period, into the active layer depth change assessment model to obtain the active layer depth change assessment result; and determine the predicted data based on the active layer depth change assessment result to determine the predicted data of the frozen soil active layer thickness, wherein the predicted data of the frozen soil active layer thickness belongs to the first future time zone.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] By employing the following techniques—acquiring basic information about permafrost in a pre-defined area; acquiring basic environmental information about the permafrost region; predicting snowmelt based on meteorological and snow cover information to obtain predicted soil moisture content data; matching thermal conductivity based on permafrost lithology, structure, and predicted soil moisture content data; statistically analyzing parameters based on meteorological forecasts to obtain a freezing index or thawing index; and inputting the freezing index and thermal conductivity during freezing, or the thawing index and thermal conductivity during thawing, into an assessment model for changes in active layer depth to obtain assessment results for changes in active layer depth and determine predicted data for the thickness of the active permafrost layer—this approach achieves the technical effect of comprehensively assessing permafrost thawing / freezing conditions from the perspectives of surface temperature changes and snow / snowmelt changes, using soil moisture content changes as a starting point, thereby improving the accuracy of permafrost active layer thickness assessment.

[0011] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0012] Figure 1 This application provides a possible flowchart of an intelligent assessment method for the thickness of the active layer of permafrost.

[0013] Figure 2 This application provides a schematic diagram illustrating a possible process for generating and acquiring predicted soil moisture content data in an intelligent assessment method for the thickness of the active permafrost layer.

[0014] Figure 3 This application provides a schematic diagram of a possible process for matching thermal conductivity in an intelligent assessment method for the thickness of the active layer of permafrost.

[0015] Figure 4 This application provides a possible structural schematic diagram of an intelligent assessment system for the thickness of the active layer of permafrost.

[0016] Figure labeling: 100 Frozen soil information acquisition module, 200 Environmental information acquisition module, 300 Snow melt prediction module, 400 Thermal conductivity matching module, 500 Parameter statistics module, 600 Depth change assessment module, 700 Prediction data determination module. Detailed Implementation

[0017] This application provides an intelligent assessment method and system for the thickness of the active permafrost layer, which solves the technical problem of low assessment accuracy of the active permafrost layer thickness. It achieves the technical effect of comprehensively assessing the thawing / freezing of permafrost by taking soil moisture content changes as the starting point and considering changes in surface temperature and snow accumulation / melting, thereby improving the assessment accuracy of the active permafrost layer thickness.

[0018] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0019] Example 1

[0020] like Figure 1 As shown in the figure, this application provides an intelligent assessment method for the thickness of the active layer of permafrost, wherein the method includes:

[0021] S10: Obtain basic information on frozen soil in a preset area, wherein the basic information on frozen soil in the preset area includes frozen soil lithology information and frozen soil structure information;

[0022] S20: Obtain basic environmental information of the permafrost region, wherein the basic environmental information of the permafrost region includes snow cover information and meteorological forecast information;

[0023] Specifically, the thickness of the active layer of permafrost will vary with seasonal temperature changes. The basic information of permafrost in the preset area includes permafrost lithology information (commonly, it can be silt, silt, clay) and permafrost structure information (commonly, it can be square, strip-shaped). Permafrost can be randomly sampled in the preset area to obtain the basic information of permafrost in the preset area.

[0024] The basic environmental information of the permafrost region includes snow cover information (including snow thickness and snowfall) and meteorological forecast information (which can be meteorological forecast information for a preset area over a future natural week, including predicted snowfall, predicted daily average temperature, and other related data). The meteorological forecast information in the basic environmental information of the permafrost region can be obtained by connecting to the network using the longitude and latitude of the preset area. The snow cover information in the basic environmental information of the permafrost region can be collected using a rain gauge to provide data support for subsequent analysis.

[0025] S30: Based on the meteorological forecast information and the snow cover information, perform snow melting prediction and obtain soil moisture content prediction data;

[0026] like Figure 2 As shown, step S30 includes the following steps:

[0027] S31: Based on the meteorological forecast information, obtain a temperature forecast sequence, wherein the temperature forecast sequence belongs to the first future time zone, and the time unit is days;

[0028] S32: Input the temperature prediction sequence into the snow melting rate matching table to obtain the snow melting rate sequence;

[0029] S33: Based on the snow cover information, perform snow melting prediction according to the snow melting rate sequence to obtain a snow melting amount prediction sequence and a snow melting thickness prediction sequence;

[0030] S34: Perform frequent term analysis based on the snowmelt amount prediction sequence to obtain the soil moisture content prediction data.

[0031] Specifically, snowmelt prediction is performed based on the meteorological forecast information and the snow cover information to obtain soil moisture content prediction data. This includes: arranging the meteorological forecast information according to the time unit of days to obtain a temperature prediction sequence, wherein the temperature prediction sequence belongs to the first future time zone (the first future time zone is in the form of: 0:00: temperature 2℃, 1:00: temperature -1℃, 2:00: temperature -3℃, 3:00: temperature -2℃); inputting the temperature prediction sequence into a snowmelt rate matching table (snow melts when it is above zero degrees Celsius, and the snowmelt rate matching table is obtained by statistical analysis under the condition of no rainfall and no snowfall, in the form of: temperature 19℃: snowmelt rate 0.0274cm / min, temperature 24℃: snowmelt rate 0.0374cm / min), and obtaining a corresponding snowmelt rate sequence according to the temperature prediction sequence;

[0032] Using the snow cover information as a starting point, snowmelt amount (e.g., the average temperature from 11:00 to 12:00 is 19℃, and the snowmelt thickness from 11:00 to 12:00 = 0.0274cm / min × 60min; if the snow thickness in the snow cover information is greater than the snowmelt thickness from 11:00 to 12:00, then the snowmelt amount from 11:00 to 12:00 = snowmelt thickness from 11:00 to 12:00 × snowmelt area = snowmelt amount at 11:00 - snowmelt amount at 12:00) is used as variable values ​​for snowmelt prediction, obtaining snowmelt amount prediction sequences and snowmelt thickness prediction sequences. The sequence elements in the snowmelt amount prediction sequence and the snowmelt thickness prediction sequence correspond one-to-one. Frequent term analysis is performed based on the snowmelt amount prediction sequence to obtain the soil moisture content prediction data (fundamentally, after the melting of the surface snow covering the permafrost region, it will seep into the underlying active permafrost layer. If the surface of the permafrost region is not covered by snow, and there is no rainfall or snowfall, the soil moisture content will not change if the evaporation of soil moisture and the water use of permafrost vegetation are not considered). This provides data support for predicting the thickness of the active permafrost layer.

[0033] Step S34 includes the following steps:

[0034] S341: The basic information of frozen soil in the preset area also includes frozen soil vegetation cover information, wherein the frozen soil vegetation cover information includes vegetation type information and vegetation quantity information;

[0035] S342: Obtain the predicted snow melt amount for day i based on the predicted snow melt amount sequence;

[0036] S343: Using the frozen soil lithology information, frozen soil structure information, vegetation type information, vegetation quantity information and the predicted snow melt amount on the i-th day as constraints, and using soil water absorption as the target data, a retrieval is performed to obtain multiple soil water absorption record results, wherein the multiple soil water absorption record results include multiple record frequency parameters.

[0037] S344: Filter the soil water absorption recording results according to the multiple recording frequency parameters to obtain the predicted soil water absorption data for day i;

[0038] S345: Based on the predicted soil water absorption data for day i, the soil moisture content calibration results for day i-1 are calibrated to generate the soil moisture content calibration results for day i.

[0039] S346: When i belongs to a preset time node of the first future time zone, the soil moisture content calibration result of the i-th day is set as the soil moisture content prediction data.

[0040] Specifically, frequent term analysis is performed based on the snowmelt amount prediction sequence to obtain the soil moisture content prediction data. This includes: the basic information of the frozen soil in the preset area also includes frozen soil vegetation cover information, which includes vegetation type information (common plants such as alpine grassland, photinia, Arctic orchid berries, and golden jasmine) and vegetation quantity information (the vegetation quantity information can be the area of ​​alpine grassland and the number of Arctic orchid berry plants); the i-th sequence element in the snowmelt amount prediction sequence is determined, and the i-th sequence element in the snowmelt amount prediction sequence is the snowmelt amount prediction data for day i.

[0041] Using the permafrost lithology information, permafrost structure information, vegetation type information, vegetation quantity information, and the predicted snowmelt amount on day i as constraints, and soil water absorption as the target data, multiple soil water absorption records are retrieved from the data storage unit of the intelligent assessment system for the thickness of the active permafrost layer. These multiple soil water absorption records include multiple recording frequency parameters (multiple recording frequency parameters: recording frequency of historical permafrost lithology information, recording frequency of historical permafrost structure information, recording frequency of historical vegetation type information, recording frequency of historical vegetation quantity information, recording frequency of historical snowmelt data, and recording frequency of historical soil water absorption). The multiple soil water absorption records also include historical permafrost lithology information, historical permafrost structure information, historical vegetation type information, historical vegetation quantity information, historical snowmelt data, and historical soil water absorption.

[0042] Based on the multiple recording frequency parameters (generally, the recording frequency of historical information on permafrost lithology, historical information on permafrost structure, historical information on vegetation type, historical information on vegetation quantity, historical snowmelt data, and historical soil water absorption are all greater than per hour / time; if the recording frequency of historical snowmelt data is 30 minutes / time, even / odd-numbered historical snowmelt data can be filtered), the multiple recording frequency parameters are limited to per hour / time, and the soil water absorption recording results are filtered to obtain a soil water absorption record table. (The soil water absorption record table is in the following format: 0 o'clock: historical information on permafrost lithology, historical information on permafrost structure, historical information on vegetation type, historical information on vegetation quantity, historical snowmelt data, and historical soil water absorption at 0 o'clock; 1 o'clock: historical information on permafrost lithology, historical information on permafrost structure, historical information on vegetation type, historical information on vegetation quantity, historical snowmelt data, and historical soil water absorption at 1 o'clock). Based on the changing patterns of the recorded data in the soil water absorption record table, the predicted soil water absorption data for day i is obtained.

[0043] Based on the predicted soil water absorption data for day i, the soil moisture content calibration results for day i-1 are calibrated by time period (due to the limited reliability of meteorological forecasts and other related information beyond the 24-hour period, the current period is limited to day i-1; simply put, to ensure the accuracy of the soil moisture content calibration results, only the 24-hour period is calibrated), thus completing the time period calibration of the soil moisture content calibration results for day i-1. The result of the time period calibration is defined as the soil moisture content calibration result for day i. When i belongs to a preset time node of the first future time zone (preset time node: the last period of the first future time zone, i.e., 23:00 to 24:00 / 0:00), the soil moisture content calibration result for day i is set as the predicted soil moisture content data to ensure the accuracy of the predicted soil moisture content data.

[0044] Step S344 includes the following steps:

[0045] S344-1: Set the soil water absorption deviation threshold;

[0046] S344-2: Perform hierarchical cluster analysis on the soil water absorption record results according to the soil water absorption deviation threshold to obtain the cluster results of the soil water absorption record results, wherein the cluster results of the soil water absorption record results include the intra-class feature values ​​of the soil water absorption record results;

[0047] S344-3: Based on the clustering results of the soil water absorption records, classify and sum the multiple record frequency parameters to generate the intra-class frequency of the soil water absorption records;

[0048] S344-4: Perform mean analysis on the intra-class feature values ​​of the soil water absorption record results whose intra-class frequencies meet the intra-class frequency threshold to generate the predicted soil water absorption data for the i-th day.

[0049] Specifically, the soil water absorption record results are filtered according to the multiple recording frequency parameters to obtain the predicted soil water absorption data for day i. This includes: setting a soil water absorption deviation threshold (preset parameter index); performing hierarchical cluster analysis on the soil water absorption record results according to the soil water absorption deviation threshold (hierarchical cluster analysis: for example, the first record data 1 of historical soil water absorption has a frequency of 3, and the second record data 2 of historical soil water absorption has a frequency of 2, then the intra-cluster feature value after clustering is equal to (3 / (3+2)*1)+(2 / (3+2)*2), and then dividing the whole by 2), to obtain the clustering result of the soil water absorption record results. The clustering result of the soil water absorption record results includes intra-cluster feature values ​​of soil water absorption record results, intra-cluster feature values ​​of historical frozen soil lithology information, intra-cluster feature values ​​of historical frozen soil structure information, intra-cluster feature values ​​of historical vegetation type information, intra-cluster feature values ​​of historical vegetation quantity information, and intra-cluster feature values ​​of historical snowmelt data.

[0050] Based on the clustering results of the soil water absorption records, the multiple recording frequency parameters are classified and summed to generate the intra-class frequency of the soil water absorption records (if the sum of the multiple recording frequency parameters is 48 for the first category, the intra-class frequency of the soil water absorption records in the first category is 30 min / time); based on the coefficient of variation method, the information contained in the intra-class frequency of the soil water absorption records that meets the intra-class frequency threshold (a user-defined intra-class frequency threshold) is directly used as weights, and the weighted mean of the intra-class feature values ​​of the soil water absorption records is calculated using the coefficient of variation method (the coefficient of variation method is an objective weighting method), generating the predicted soil water absorption data for the i-th day, providing data support for the assessment of the active layer thickness of permafrost.

[0051] S40: Based on the frozen soil lithology information, the frozen soil structure information and the predicted soil moisture content data, match the thermal conductivity, wherein the thermal conductivity includes the thermal conductivity during the thawing period or the thermal conductivity during the freezing period.

[0052] like Figure 3 As shown, step S40 includes the following steps:

[0053] S41: Obtain multiple expert decision-making modules, wherein information is not shared between any two expert decision-making modules, and the multiple expert decision-making modules have multiple decision credibility levels;

[0054] S42: Input the frozen soil lithology information, the frozen soil structure information and the soil moisture content prediction data into the multiple expert decision modules to obtain multiple thermal conductivity decision results;

[0055] S43: Perform a weighted average analysis based on the multiple decision credibility and the multiple thermal conductivity decision results to obtain the thermal conductivity during the melting period or the thermal conductivity during the freezing period, and add it to the thermal conductivity.

[0056] Specifically, based on the permafrost lithology information, the permafrost structure information, and the predicted soil moisture content data, the thermal conductivity is matched. This includes: acquiring multiple expert decision-making modules (initially human decision-making; after a period of time, when the accumulated data from human decision-making is large, an expert decision-making module is constructed for automated processing). Information is not shared between any two expert decision-making modules, and the multiple expert decision-making modules have multiple decision credibility levels. The thermal conductivity includes the thermal conductivity during the thawing period (when the daily average temperature is >0℃, the thermal conductivity during the thawing period is directly matched) or the thermal conductivity during the freezing period (when the daily average temperature is <0℃, the thermal conductivity during the freezing period is directly matched).

[0057] The data accumulated during the human decision-making stage (including permafrost lithology information, permafrost structure information, soil moisture content, and calibration results, all stored in association, with each data segment stored independently on multiple independent address fragments) serves as a knowledge base. Multiple expert decision-making modules are constructed (information is not shared between any two modules). Each module corresponds one-to-one with a specific address fragment. The permafrost lithology information, permafrost structure information, and predicted soil moisture content are used as input data. The data is fed into multiple expert decision-making modules, and intelligent processing is performed according to these modules to output multiple thermal conductivity decision results. A weighted mean analysis is then performed based on the reliability of the multiple decisions and the multiple thermal conductivity decision results (weighted mean analysis is equivalent to weighted mean calculation; the specific operation steps have already been explained and will not be repeated here) to obtain the thermal conductivity during the thawing period (if the daily average temperature is >0℃, i.e., the thermal conductivity during the thawing period) or the thermal conductivity during the freezing period (if the daily average temperature is <0℃, i.e., the thermal conductivity during the thawing period). This thermal conductivity is then added to the thermal conductivity data to provide data support for the analysis of frozen soil conditions (generally, frozen soil conditions mean thawing in summer and freezing in winter).

[0058] S50: Perform parameter statistics based on the meteorological forecast information to obtain the freezing index or thawing index;

[0059] Step S50 includes the following steps:

[0060] S51: Based on the meteorological forecast information, obtain the surface temperature forecast sequence, wherein the surface temperature forecast information belongs to the first future time zone;

[0061] S52: Using 0℃ as the critical value, the surface temperature prediction sequence is grouped to obtain positive temperature prediction sequences and negative temperature prediction sequences.

[0062] S53: Calculate the freezing index based on the positive temperature prediction sequence to obtain the freezing index;

[0063] S54: Calculate the freezing index based on the negative temperature prediction sequence to obtain the melting index.

[0064] Specifically, based on the meteorological forecast information, parameter statistics are performed to obtain a freezing index or a thawing index. This includes: the surface temperature forecast information belongs to the first future time zone; based on the meteorological forecast information and the distribution pattern of the first future time zone, a surface temperature forecast sequence is obtained; using 0℃ as a critical value, the surface temperature forecast sequence is grouped, with those above 0℃ classified as positive temperature forecast sequences and those below 0℃ classified as negative temperature forecast sequences, thus obtaining positive and negative temperature forecast sequences; a freezing index is calculated based on the positive temperature forecast sequence (freezing index calculation: freezing index = the sum of the products of the duration of the daily average temperature below 0℃ and its value), thus obtaining the freezing index; a freezing index is calculated based on the negative temperature forecast sequence (freezing index calculation: thawing index = the sum of the products of the duration of the daily average temperature above 0℃ and its value), thus obtaining the thawing index, providing data support for assessing changes in permafrost conditions.

[0065] S60: Input the freezing index and the thermal conductivity during freezing period, or the melting index and the thermal conductivity during melting period, into the active layer depth change assessment model to obtain the active layer depth change assessment result;

[0066] S70: Based on the assessment results of the change in the depth of the active layer, determine the predicted data of the active layer thickness of the permafrost, wherein the predicted data of the active layer thickness of the permafrost belongs to the first future time zone.

[0067] Specifically, an assessment model for the change in active layer depth is constructed. The freezing index and the thermal conductivity during freezing, or the thawing index and the thermal conductivity during thawing, are used as input data to the assessment model to obtain the assessment results. The permafrost structure information is used as the initial quantity, and the assessment results for the change in active layer depth are used as the change quantity (which can be positive or negative). These are accumulated to determine the predicted data for the active layer thickness of the permafrost. The predicted data for the active layer thickness belongs to the first future time zone, allowing for advance estimation of the active layer thickness and providing technical support for the construction and development of permafrost regions.

[0068] Step S60 includes the following steps:

[0069] S61: Based on the frozen soil lithology information and the frozen soil structure information, collect frozen soil data to obtain active layer freezing record data and active layer thawing record data;

[0070] S62: The active layer freezing record data includes freezing index record data, thermal conductivity record data during freezing period, and first active layer depth change record data;

[0071] S63: The active layer melting record data includes melting index record data, thermal conductivity record data during the melting period, and second active layer depth change record data;

[0072] S64: Based on the frozen index record data, the frozen period thermal conductivity record data, and the first active layer depth change record data, train the first evaluation unit for the active layer depth change based on random forest.

[0073] S65: Based on the melting index recorded data, the thermal conductivity recorded data during the melting period, and the second active layer depth change recorded data, train the second evaluation unit for the active layer depth change based on random forest;

[0074] S66: Merge the first evaluation unit of the change in active layer depth and the second evaluation unit of the change in active layer depth to generate the evaluation model of the change in active layer depth.

[0075] Specifically, constructing an assessment model for the depth change of the active layer includes: based on the permafrost lithology information and the permafrost structure information, collecting permafrost data in a preset area to obtain active layer freezing record data (data obtained from records of daily average surface temperature greater than 0℃) and active layer thawing record data (data obtained from records of daily average surface temperature less than 0℃); the active layer freezing record data includes freezing index record data, thermal conductivity record data during the freezing period, and first active layer depth change record data; the active layer thawing record data includes thawing index record data, thermal conductivity record data during the thawing period, and second active layer depth change record data;

[0076] Based on random forest, the frozen index record data, the frozen period thermal conductivity record data, and the first active layer depth change record data are set as root point features: the frozen index record data root point features are the index features corresponding to multiple sets of frozen indices; the frozen period thermal conductivity record data root point features are the index features corresponding to multiple sets of frozen period thermal conductivity; the first active layer depth change record data root point features are the index features corresponding to multiple sets of first active layer depth change. The space is divided using all values ​​of a certain feature, and training continues until the loss function is minimized to determine the optimal split point. After determining the optimal split point (the optimal split point defines the region, determining the data processing logic of the first evaluation unit, and the optimal split point defines the corresponding output value of the first evaluation unit), the first evaluation unit for the active layer depth change is constructed using the root point features defined by the optimal split point and the multiple sets of frozen indices, multiple sets of frozen period thermal conductivity, and multiple sets of first active layer depth changes corresponding to the defined root point features.

[0077] Based on random forest, the melting index record data, the thermal conductivity record data during the melting period, and the record data of the second active layer depth change are set as root point features to train the second evaluation unit of the active layer depth change (the specific operation steps have been described and will not be repeated here); the first evaluation unit and the second evaluation unit of the active layer depth change are combined in parallel by multiple threads (the first evaluation unit and the second evaluation unit of the active layer depth change are independent processing units) to generate the evaluation model of the active layer depth change, providing model support for evaluating the change of active layer depth.

[0078] In summary, the intelligent assessment method and system for the thickness of the active layer of permafrost provided in this application have the following technical effects:

[0079] 1. By employing methods such as acquiring basic information about permafrost in a preset area; acquiring basic environmental information about the permafrost area; predicting snow melt based on meteorological forecasts and snow cover information; acquiring predicted soil moisture content data; combining permafrost lithology and structure information; matching thermal conductivity; statistically acquiring freezing or thawing indices; and inputting the freezing index and thermal conductivity during freezing or thawing indices and thermal conductivity during thawing into an active layer depth change assessment model to obtain the active layer depth change assessment results and determine the predicted data for the thickness of the active permafrost layer, this application provides an intelligent assessment method and system for the thickness of the active permafrost layer. This achieves the technical effect of comprehensively assessing the thawing / freezing of permafrost based on changes in soil moisture content, surface temperature changes, and snow / snow melt changes, thereby improving the accuracy of the assessment of the thickness of the active permafrost layer.

[0080] 2. By setting a soil water absorption deviation threshold, hierarchical cluster analysis was performed on the soil water absorption records to obtain the clustering results. Multiple recording frequency parameters were classified and summed to generate the intra-class frequency of the soil water absorption records. The mean value analysis was performed on the intra-class feature values ​​of the soil water absorption records whose intra-class frequencies met the intra-class frequency threshold to generate the predicted soil water absorption data for day i, providing data support for the assessment of the active layer thickness of permafrost.

[0081] Example 2

[0082] Based on the same inventive concept as the intelligent assessment method for the active layer thickness of permafrost in the foregoing embodiments, such as Figure 4 As shown in the figure, this application provides an intelligent assessment system for the thickness of the active layer of permafrost, wherein the system includes:

[0083] The permafrost information acquisition module 100 is used to acquire basic information about permafrost in a preset area, wherein the basic information about permafrost in the preset area includes permafrost lithology information and permafrost structure information.

[0084] The environmental information acquisition module 200 is used to acquire basic environmental information of the permafrost region, wherein the basic environmental information of the permafrost region includes snow cover information and meteorological forecast information;

[0085] The snowmelt prediction module 300 is used to predict snowmelt based on the meteorological forecast information and the snow cover information, and to obtain soil moisture content prediction data.

[0086] The thermal conductivity matching module 400 is used to match the thermal conductivity based on the frozen soil lithology information, the frozen soil structure information and the predicted soil moisture content data, wherein the thermal conductivity includes the thermal conductivity during the thawing period or the thermal conductivity during the freezing period.

[0087] The parameter statistics module 500 is used to perform parameter statistics based on the meteorological forecast information to obtain the freezing index or thawing index.

[0088] The depth change assessment module 600 is used to input the freezing index and the thermal conductivity during the freezing period, or the melting index and the thermal conductivity during the melting period, into the active layer depth change assessment model to obtain the active layer depth change assessment result.

[0089] The prediction data determination module 700 is used to determine the predicted data of the active layer thickness of the permafrost based on the evaluation result of the change in the depth of the active layer, wherein the predicted data of the active layer thickness of the permafrost belongs to the first future time zone.

[0090] Furthermore, the system includes:

[0091] A temperature prediction sequence acquisition module is used to acquire a temperature prediction sequence based on the meteorological prediction information, wherein the temperature prediction sequence belongs to the first future time zone, and the time unit is days;

[0092] The snow melting rate sequence acquisition module is used to input the temperature prediction sequence into the snow melting rate matching table to obtain the snow melting rate sequence;

[0093] The snowmelt prediction sequence acquisition module is used to predict snowmelt based on the snowmelt rate sequence and the snow cover information, and to acquire a snowmelt amount prediction sequence and a snowmelt thickness prediction sequence.

[0094] The frequent term analysis module is used to perform frequent term analysis based on the snowmelt amount prediction sequence to obtain the soil moisture content prediction data.

[0095] Furthermore, the system includes:

[0096] The frozen soil vegetation cover data determination module is used to determine the basic information of frozen soil in the preset area, which also includes frozen soil vegetation cover information, wherein the frozen soil vegetation cover information includes vegetation type information and vegetation quantity information.

[0097] The snowmelt amount prediction data acquisition module is used to acquire the snowmelt amount prediction data for day i based on the snowmelt amount prediction sequence.

[0098] The water absorption record result acquisition module is used to retrieve multiple soil water absorption record results by using the frozen soil lithology information, the frozen soil structure information, the vegetation type information, the vegetation quantity information and the predicted snow melt amount on the i-th day as constraints and soil water absorption as the target data. The multiple soil water absorption record results include multiple recording frequency parameters.

[0099] The water absorption record result filtering module is used to filter the soil water absorption record results according to the multiple recording frequency parameters to obtain the predicted soil water absorption data for day i.

[0100] The moisture content calibration module is used to calibrate the soil moisture content calibration results for day i-1 based on the predicted soil water absorption data for day i, and generate the soil moisture content calibration results for day i.

[0101] The soil moisture content prediction data confirmation module is used to set the soil moisture content calibration result of day i as the soil moisture content prediction data when i belongs to a preset time node of the first future time zone.

[0102] Furthermore, the system includes:

[0103] The water absorption deviation threshold setting module is used to set the soil water absorption deviation threshold.

[0104] The hierarchical clustering analysis module is used to perform hierarchical clustering analysis on the soil water absorption record results according to the soil water absorption deviation threshold, and obtain the clustering results of the soil water absorption record results, wherein the clustering results of the soil water absorption record results include the intra-class feature values ​​of the soil water absorption record results.

[0105] The classification and summation module is used to classify and sum the multiple recording frequency parameters according to the clustering results of the soil water absorption records, and generate the intra-class frequency of the soil water absorption records.

[0106] The mean analysis module is used to perform mean analysis on the intra-class feature values ​​of the soil water absorption record results whose intra-class frequency meets the intra-class frequency threshold, and generate the predicted soil water absorption data for the i-th day.

[0107] Furthermore, the system includes:

[0108] A surface temperature prediction sequence acquisition module is used to acquire a surface temperature prediction sequence based on the meteorological prediction information, wherein the surface temperature prediction information belongs to the first future time zone.

[0109] The sequence grouping module is used to group the surface temperature prediction sequence with 0℃ as the critical value to obtain positive temperature prediction sequences and negative temperature prediction sequences.

[0110] The first freezing index calculation module is used to calculate the freezing index based on the positive temperature prediction sequence and obtain the freezing index.

[0111] The second freezing index calculation module is used to calculate the freezing index based on the negative temperature prediction sequence and obtain the melting index.

[0112] Furthermore, the system includes:

[0113] The frozen soil data acquisition module is used to acquire frozen soil data based on the frozen soil lithology information and the frozen soil structure information, and to obtain active layer freezing record data and active layer thawing record data.

[0114] The first data acquisition module is used to record the frozen data of the active layer, including the frozen index data, the thermal conductivity data during the freezing period data, and the depth change data of the first active layer.

[0115] The second data acquisition module is used to record the melting data of the active layer, including melting index data, thermal conductivity data during the melting period data, and the change in depth of the second active layer data.

[0116] The first training module is used to train the first evaluation unit of the change in the depth of the active layer based on random forest, according to the frozen index record data, the frozen period thermal conductivity record data and the first active layer depth change record data.

[0117] The second training module is used to train the second evaluation unit of the change in the depth of the active layer based on random forest, according to the melting index recorded data, the thermal conductivity recorded data during the melting period, and the change in the depth of the second active layer recorded data.

[0118] The unit merging module is used to merge the first evaluation unit of the change in active layer depth and the second evaluation unit of the change in active layer depth to generate the evaluation model of the change in active layer depth.

[0119] Furthermore, the system includes:

[0120] The decision module acquisition module is used to acquire multiple expert decision modules, wherein information is not shared between any two expert decision modules, and the multiple expert decision modules have multiple decision credibility levels;

[0121] The prediction data input module is used to input the frozen soil lithology information, the frozen soil structure information and the predicted soil moisture content data into the multiple expert decision modules to obtain multiple thermal conductivity decision results;

[0122] The thermal conductivity addition module is used to perform a weighted average analysis based on the multiple decision credibility and the multiple thermal conductivity decision results to obtain the thermal conductivity during the melting period or the thermal conductivity during the freezing period, and add it to the thermal conductivity.

[0123] In summary, any step of the method described above can be stored as a computer instruction or program in an unrestricted computer memory, and can be called and identified by an unrestricted computer processor to implement any method in the embodiments of this application, without any additional restrictions.

[0124] Furthermore, the "first" or "second" mentioned above may not only represent a sequential relationship, but may also represent a specific concept, and / or refer to the individual or collective selection of multiple elements. Clearly, those skilled in the art can make various modifications and variations to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A smart assessment method for the thickness of the active layer of permafrost, characterized in that, include: Obtain basic information about permafrost in a preset area, wherein the basic information about permafrost in the preset area includes permafrost lithology information and permafrost structure information; Obtain basic environmental information of the permafrost region, including snow cover information and meteorological forecast information; Based on the meteorological forecast information and the snow cover information, snow melting prediction is performed to obtain soil moisture content prediction data; Based on the frozen soil lithology information, the frozen soil structure information, and the predicted soil moisture content data, the thermal conductivity is matched, wherein the thermal conductivity includes the thermal conductivity during the thawing period or the thermal conductivity during the freezing period. Based on the meteorological forecast information, parameter statistics are performed to obtain the freezing index or thawing index; Input the freezing index and the thermal conductivity during freezing period, or the melting index and the thermal conductivity during melting period, into the active layer depth change assessment model to obtain the active layer depth change assessment result; Based on the assessment results of the change in the active layer depth, the predicted data of the active layer thickness of the permafrost is determined, wherein the predicted data of the active layer thickness of the permafrost belongs to the first future time zone.

2. The method as described in claim 1, characterized in that, The step of predicting snow melt based on the meteorological forecast information and the snow cover information, and obtaining soil moisture content prediction data, includes: Based on the meteorological forecast information, a temperature forecast sequence is obtained, wherein the temperature forecast sequence belongs to the first future time zone, and the time unit is days; Input the temperature prediction sequence into the snowmelt rate matching table to obtain the snowmelt rate sequence; Based on the snow melt rate sequence and the snow cover information, snow melt prediction is performed to obtain a snow melt amount prediction sequence and a snow melt thickness prediction sequence. Frequent term analysis is performed on the snowmelt amount prediction sequence to obtain the soil moisture content prediction data.

3. The method as described in claim 2, characterized in that, The step of performing frequent term analysis based on the snowmelt amount prediction sequence to obtain the soil moisture content prediction data includes: The basic information of the frozen soil in the preset area also includes frozen soil vegetation cover information, wherein the frozen soil vegetation cover information includes vegetation type information and vegetation quantity information. Based on the snowmelt prediction sequence, obtain the snowmelt prediction data for day i. Using the frozen soil lithology information, frozen soil structure information, vegetation type information, vegetation quantity information, and the predicted snow melt amount on day i as constraints, and soil water absorption as the target data, multiple soil water absorption records are retrieved, wherein the multiple soil water absorption records include multiple recording frequency parameters. The soil water absorption record results are filtered according to the multiple recording frequency parameters to obtain the predicted soil water absorption data for day i. Based on the predicted soil water absorption data for day i, the soil moisture content calibration results for day i-1 are calibrated to generate the soil moisture content calibration results for day i. When i belongs to a preset time node of the first future time zone, the soil moisture content calibration result of the i-th day is set as the soil moisture content prediction data.

4. The method as described in claim 3, characterized in that, The step of filtering the soil water absorption record results based on the multiple recording frequency parameters to obtain the predicted soil water absorption data for day i includes: Set a threshold for soil water absorption deviation; Hierarchical cluster analysis is performed on the soil water absorption record results based on the soil water absorption deviation threshold to obtain the cluster results of the soil water absorption record results, wherein the cluster results of the soil water absorption record results include the intra-class feature values ​​of the soil water absorption record results; Based on the clustering results of the soil water absorption records, the multiple record frequency parameters are classified and summed to generate the intra-class frequency of the soil water absorption records. Mean analysis is performed on the intra-class feature values ​​of the soil water absorption record results whose intra-class frequency meets the intra-class frequency threshold to generate the predicted soil water absorption data for the i-th day.

5. The method as described in claim 1, characterized in that, The step of performing parameter statistics based on the meteorological forecast information to obtain the freezing index or thawing index includes: Based on the meteorological forecast information, a surface temperature prediction sequence is obtained, wherein the surface temperature prediction sequence belongs to the first future time zone; Using 0℃ as the critical value, the surface temperature prediction sequence is grouped to obtain positive and negative temperature prediction sequences. The freezing index is calculated based on the positive temperature prediction sequence to obtain the freezing index; The freezing index is calculated based on the negative temperature prediction sequence to obtain the melting index.

6. The method as described in claim 1, characterized in that, The step of inputting the freezing index and the thermal conductivity during the freezing period, or the melting index and the thermal conductivity during the melting period, into the active layer depth change assessment model to obtain the active layer depth change assessment result includes: Based on the frozen soil lithology information and the frozen soil structure information, frozen soil data is collected to obtain active layer freezing record data and active layer thawing record data. The active layer freezing record data includes freezing index record data, thermal conductivity record data during the freezing period, and first active layer depth change record data. The active layer melting record data includes melting index record data, thermal conductivity record data during the melting period, and second active layer depth change record data. Based on the frozen index record data, the frozen period thermal conductivity record data, and the first active layer depth change record data, a first evaluation unit for the active layer depth change is trained using random forest. Based on the melting index recorded data, the thermal conductivity recorded data during the melting period, and the second active layer depth change recorded data, a second evaluation unit for the active layer depth change is trained using random forest. The first evaluation unit and the second evaluation unit of the change in active layer depth are merged to generate the evaluation model of the change in active layer depth.

7. The method as described in claim 1, characterized in that, The step involves matching thermal conductivity based on the frozen soil lithology information, the frozen soil structure information, and the predicted soil moisture content data. The thermal conductivity includes either the thermal conductivity during the thawing period or the thermal conductivity during the freezing period, and includes: Multiple expert decision-making modules are obtained, wherein information is not shared between any two expert decision-making modules, and the multiple expert decision-making modules have multiple decision credibility levels; The frozen soil lithology information, the frozen soil structure information, and the predicted soil moisture content data are input into the multiple expert decision modules to obtain multiple thermal conductivity decision results. A weighted mean analysis is performed based on the multiple decision credibility and the multiple thermal conductivity decision results to obtain the thermal conductivity during the melting period or the thermal conductivity during the freezing period, and then added to the thermal conductivity.

8. An intelligent assessment system for the thickness of the active layer of permafrost, characterized in that, A smart assessment method for the thickness of the active layer of permafrost as described in any one of claims 1-7 includes: The permafrost information acquisition module is used to acquire basic information about permafrost in a preset area, wherein the basic information about permafrost in the preset area includes permafrost lithology information and permafrost structure information. An environmental information acquisition module is used to acquire basic environmental information of the permafrost region, wherein the basic environmental information of the permafrost region includes snow cover information and meteorological forecast information; The snowmelt prediction module is used to predict snowmelt based on the meteorological forecast information and the snow cover information, and to obtain soil moisture content prediction data. A thermal conductivity matching module is used to match thermal conductivity based on the frozen soil lithology information, the frozen soil structure information and the predicted soil moisture content data, wherein the thermal conductivity includes the thermal conductivity during the thawing period or the thermal conductivity during the freezing period. The parameter statistics module is used to perform parameter statistics based on the meteorological forecast information to obtain the freezing index or thawing index. The depth change assessment module is used to input the freezing index and the thermal conductivity during the freezing period, or the melting index and the thermal conductivity during the melting period, into the active layer depth change assessment model to obtain the active layer depth change assessment result. The prediction data determination module is used to determine the predicted data of the active layer thickness of the permafrost based on the evaluation results of the change in the depth of the active layer, wherein the predicted data of the active layer thickness of the permafrost belongs to the first future time zone.

Images