Fully automatic identification method, system and device for foundation coefficient

Abstract

The present invention discloses a fully automatic identification method, system and device for foundation coefficient, and relates to the technical field of data identification. After collecting foundation-related data in a construction area, a foundation coefficient identification model is trained and obtained. The foundation coefficient of each detection point in the construction area is identified using the foundation coefficient identification model. If there is an abnormality in the foundation coefficient, a multi-level alarm instruction is sent to the outside through multiple channels. After adjusting the warning threshold using the trained threshold adjustment model, a hidden danger area is screened out in the construction area according to the position of the detection point and the value of the foundation coefficient. The real-time risk data in the hidden danger area is used as input, and the trained construction risk identification model is used for scoring, and the corresponding risk treatment measures are matched in the hidden danger area according to the obtained risk score. After identifying and screening the hidden danger area in the construction area through the value of the foundation coefficient and its location information, targeted treatment measures are taken.

Description

Technical Field

[0001] The present invention relates to the technical field of data identification, and in particular to a method, system and device for fully automatic identification of foundation coefficients. Background Art

[0002] The foundation coefficient is an important parameter to measure the ability of soil to resist compression deformation when subjected to uniformly distributed plane pressure. It is obtained through precise static pressure plate load test. The foundation coefficient not only reflects the physical and mechanical properties of the soil, but is also directly related to the safety and stability of the engineering structure. In the design and construction of railways, highways, large buildings and infrastructure, the accurate determination of the foundation coefficient is particularly important.

[0003] The foundation coefficient helps engineers gain a deeper understanding of the bearing capacity of the foundation and evaluate the deformation of the foundation under the expected load, so as to formulate a scientific and reasonable foundation treatment plan to ensure the safety and reliability of the engineering structure. At the same time, the change of the foundation coefficient also reflects the effect of the foundation treatment, providing an important basis for the control and acceptance of engineering quality.

[0004] In the Chinese invention patent with the authorization announcement number CN103422484B, a method for estimating the K30 value of coarse-grained soil filler for high-speed railway roadbed is proposed. The main steps are: A. Determination of the maximum dry density of the filler sample, B. Determination of the maximum bite force Cmax (in kPa / M) of the filler sample under the compaction coefficient control value K, and C. Calculation of the foundation coefficient K30 (in MPa / m). This method has a clear principle and is easy to operate. It can use conventional indoor direct shear test data to accurately and reliably estimate the foundation coefficient K30 of coarse-grained soil filler under the specified compaction coefficient, thereby greatly reducing the number of rolling tests in the on-site test section, reducing the investment in manpower and material resources, and shortening the construction period.

[0005] Combined with the above application and the contents of the prior art:

[0006] When carrying out construction in areas prone to landslides and mud-rock flows, especially when carrying out construction of bridges, tunnels, and reservoir dams, the geological conditions in the construction area may be abnormal. For example, when the geological water level is too high or the soil moisture content is too high, the soil bearing capacity in the construction area will be affected to a certain extent. Therefore, before construction, it is necessary to detect and identify the foundation coefficient in the construction area to verify whether the foundation coefficient meets the construction requirements and whether there are any safety hazards.

[0007] The existing foundation coefficient detection method usually collects various data in the construction area, and then obtains it through data analysis and calculation. This identification method is usually simple, but inefficient and requires regular measurements. Not only is the real-time feedback effect poor, but the detection and identification efficiency is actually low. When there are abnormal environmental conditions in the construction area, especially when there is more precipitation and soil moisture content in the construction area and adjacent areas, the foundation coefficient in the construction area may fluctuate to a certain extent, resulting in foundation construction risks in the construction area, which are difficult to detect and deal with in time.

[0008] To this end, the present invention provides a fully automatic identification method, system and device for foundation coefficients. Summary of the invention

[0009] 1. Technical issues to be resolved

[0010] In view of the deficiencies of the prior art, the present invention provides a fully automatic identification method, system and device for foundation coefficients. The method collects foundation-related data in the construction area and then trains to obtain a foundation coefficient identification model. The foundation coefficient identification model is used to identify the foundation coefficients of each detection point in the construction area. If there is an abnormality in the foundation coefficient, a multi-level alarm instruction is issued to the outside through multiple channels. After the warning threshold is adjusted using the trained threshold adjustment model, the hidden danger area is screened out in the construction area according to the location of the detection point and the value of the foundation coefficient. The real-time risk data in the hidden danger area is used as input, and the trained construction risk identification model is used for scoring. According to the acquired risk score, the corresponding risk treatment measures are matched in the hidden danger area. After the hidden danger area is identified and screened out in the construction area by the value of the foundation coefficient and its location information, targeted treatment measures are taken. Thus, the technical problems raised in the background technology are solved.

[0011] (II) Technical solution

[0012] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0013] The fully automatic identification method of foundation coefficient includes generating an environmental anomaly degree from environmental condition data in a construction area, and if the obtained environmental anomaly degree exceeds an anomaly threshold, collecting foundation-related data in the construction area;

[0014] After the foundation coefficient identification model is trained and obtained, it is tested and a performance value is generated by the obtained performance index. If the performance value is lower than the performance threshold, the foundation coefficient identification model is optimized;

[0015] Use the optimized foundation coefficient identification model to identify the foundation coefficient of each detection point in the construction area. If there is an abnormality in the foundation coefficient, send a multi-level alarm command to the outside through multiple channels;

[0016] After adjusting the warning threshold using the trained threshold adjustment model, the potential danger areas are screened out in the construction area according to the location of the detection point and the value of the foundation coefficient;

[0017] The real-time risk data in the hidden danger area is used as input, and the trained construction risk identification model is used to score it. According to the obtained risk score, the corresponding risk treatment measures are matched in the hidden danger area.

[0018] Further, environmental condition data is periodically monitored and collected in the construction area, and environmental condition data in the area is generated after being aggregated;

[0019] The environmental abnormality is generated from the environmental condition data in the environmental condition data set. If the acquired environmental abnormality exceeds the abnormality threshold, a data collection instruction is issued to the outside.

[0020] Further, after receiving the data collection instruction, various test data and test data are collected in the construction area;

[0021] A sensor network is set up in the construction area, which collects various data in real time in the construction area, and aggregates the detection data, test data and real-time data to generate a foundation-related data set.

[0022] Furthermore, some data are extracted from the foundation-related data set to train and obtain a foundation coefficient identification model;

[0023] The foundation coefficient identification model is tested and its performance is monitored in real time to obtain corresponding performance indicators, which are summarized to generate a model indicator data set, and the performance value of the foundation coefficient identification model is generated from the model indicator data set.

[0024] Furthermore, a number of evenly distributed detection points are selected in the construction area, and after the corresponding foundation coefficients are detected and obtained at the detection points, the optimized foundation coefficient identification model outputs corresponding identification data;

[0025] A similarity analysis is performed between the two, and the corresponding data similarity is obtained. If the similarity is lower than a preset similarity threshold, the optimized foundation coefficient identification model is optimized.

[0026] Furthermore, the optimized foundation coefficient identification model is used to identify the foundation coefficients of each inspection point in the construction area using the real-time foundation-related data as input;

[0027] The identified foundation coefficient is taken as input and identified using the trained abnormal data recognition model. If the foundation coefficient at the detection point is abnormal, an early warning instruction is issued to the outside.

[0028] Furthermore, if the number of warning instructions received continuously within the preset observation period exceeds the expected number, the time node when the warning instruction is received is used as the warning node, and the warning value is constructed based on the warning node and the warning location information;

[0029] If the warning value exceeds the preset warning threshold, an alarm command will be issued to the outside.

[0030] Furthermore, a correlation analysis is performed between the environmental anomaly degree and the foundation coefficient. If the value of the correlation coefficient exceeds the expectation, a trained threshold adjustment model is obtained after training the machine learning algorithm;

[0031] After receiving the alarm instruction, the trained threshold adjustment model is used to adjust the current warning threshold.

[0032] Furthermore, each detection point is marked with the foundation coefficient, and the construction area is divided into several sub-areas using the trained classifier according to the location of the detection point and the value of the foundation coefficient, and the potential danger areas are screened out in the several sub-areas.

[0033] Furthermore, environmental condition data, foundation coefficient and construction data in the potential hazard area are obtained as real-time risk data;

[0034] Real-time risk data is used as input and scored using the trained construction risk identification model to obtain risk scores for potential hazard areas.

[0035] Furthermore, based on the relationship between the risk score and the risk threshold, the foundation construction risks in the potential danger area are divided into different risk levels, and corresponding risk treatment measures are matched in the sub-areas according to the risk level. After the risk treatment measures are executed, a monitoring and treatment report is automatically generated.

[0036] The fully automatic identification system of foundation coefficient includes: an environmental analysis unit, which generates an environmental abnormality degree from environmental condition data in the construction area, and collects foundation-related data in the construction area if the acquired environmental abnormality degree exceeds an abnormality threshold;

[0037] The model performance analysis unit, after training and acquiring the foundation coefficient identification model, tests it and generates a performance value based on the acquired performance index. If the performance value is lower than the performance threshold, the foundation coefficient identification model is optimized;

[0038] The abnormal alarm unit uses the optimized foundation coefficient identification model to identify the foundation coefficient of each detection point in the construction area. If there is an abnormality in the foundation coefficient, a multi-level alarm instruction is issued to the outside through multiple channels;

[0039] The area recognition unit uses the trained threshold adjustment model to adjust the warning threshold, and then selects the potential danger area in the construction area according to the location of the detection point and the value of the foundation coefficient;

[0040] The emergency response unit takes the real-time risk data in the hidden danger area as input, uses the trained construction risk identification model to score it, and matches the corresponding risk treatment measures in the hidden danger area according to the obtained risk score.

[0041] The fully automatic identification device of foundation coefficient includes:

[0042] At least one processor; a memory for storing executable instructions, wherein the instructions are executed by at least one of the processors to enable at least one of the processors to perform the steps of the method for fully automatic identification of foundation coefficients.

[0043] (III) Beneficial effects

[0044] The present invention provides a fully automatic identification method, system and device for foundation coefficient, which have the following beneficial effects:

[0045] 1. The abnormality of the construction environment can be evaluated based on the degree of environmental abnormality. If there are abnormal environmental conditions in the construction area, adjustments can be made to the construction safety hazards in the construction area before construction.

[0046] 2. The performance of the foundation coefficient identification model can be comprehensively judged based on the performance value, and the foundation coefficient identification model can be optimized in a targeted manner based on the obtained performance test data, thereby improving the reliability of the identification results when it is used to identify the foundation coefficient.

[0047] 3. Use data similarity as a judgment indicator to verify whether the reliability of the foundation coefficient identification model can meet expectations. If it still cannot meet expectations, it can be re-optimized to fully guarantee the reliability of the foundation coefficient identification model.

[0048] 4. The foundation coefficient identification model is used to identify the foundation coefficient at various locations in the construction area, completing the real-time monitoring and identification process. When the foundation coefficient is abnormal, corresponding processing can be made in time; the foundation coefficient in the construction area can be quickly identified with higher efficiency, accuracy and reliability.

[0049] 5. When the foundation coefficient in the construction area is abnormal, the degree of abnormality is judged. When the degree of abnormality exceeds expectations, an alarm instruction is issued to the outside to realize multi-level alarm for abnormal foundation coefficient and ensure the safety of subsequent construction in the construction area.

[0050] 6. The trained threshold adjustment model is used to adjust the current warning threshold so that the warning threshold is in a dynamic adjustment state and adapted to the environmental conditions in the construction area. Through the value of the foundation coefficient and its location information, after identifying and screening the hidden danger areas in the construction area, targeted treatment measures are taken to reduce construction risks.

[0051] 7. By conducting risk assessment on the potential hazard area, the degree of potential hazard in the potential hazard area can be evaluated based on the obtained risk score, so as to facilitate the judgment of the degree of potential hazard and the adoption of targeted treatment measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a schematic diagram of the flow of the fully automatic identification method of foundation coefficient of the present invention;

[0053] Figure 2 This is a schematic diagram of the structure of the fully automatic identification system for foundation coefficients of the present invention;

[0054] Figure 3 It is a schematic diagram of the structure of the fully automatic identification device of foundation coefficient of the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] See also Figure 1 The present invention provides a fully automatic identification method for foundation coefficients, comprising:

[0057] Step 1: Generate an environmental anomaly degree from the environmental condition data in the construction area. If the acquired environmental anomaly degree exceeds the anomaly threshold, collect foundation-related data in the construction area and aggregate them to generate a foundation-related data set;

[0058] The step 1 includes the following contents:

[0059] Step 101: If the construction area is a landslide and debris flow prone area, periodically monitor and collect environmental condition data in the construction area, including temperature data, precipitation data, air humidity and soil humidity, groundwater level and hydrological data, etc.; summarize the collected environmental condition data to generate environmental condition data in the area;

[0060] Step 102: Under dimensionless conditions, the environmental anomaly degree is generated from the environmental condition data in the environmental condition data set, and the permeability and stability of the foundation are evaluated by the environmental anomaly degree, in the following manner:

[0061] Where: is the environmental factor, , and They are The values ​​of precipitation, soil moisture and groundwater level at the time, , and are the corresponding weight coefficients, obtained by referring to the analytic hierarchy process; is the mean value of the environmental coefficient, is the historical standard deviation of the environmental coefficient;

[0062] Based on historical data and management expectations for the safety of the foundation in the construction area, an abnormal threshold is set in advance; if the acquired environmental abnormality exceeds the abnormal threshold, it means that the environmental conditions in the construction area will have a certain impact on the construction process. At this time, a data collection instruction is issued to the outside;

[0063] When in use, when construction needs to be carried out in the construction area, by collecting environmental condition data and generating environmental anomaly degrees, the abnormality of the construction environment can be evaluated based on the environmental anomaly degrees. If there are abnormalities in the environmental conditions in the construction area, adjustments can be made to the construction safety hazards in the construction area before construction;

[0064] Step 103: After receiving the data collection instruction, various test data and test data are collected in the construction area, such as using drilling and geological radar to draw a detailed geological profile, including soil thickness, soil composition and structure; soil sampling at different depths and locations to obtain representative soil samples: obtaining geological information at different depths underground through drilling, including core sample records; geotechnical tests: geotechnical tests are carried out, including: Compression test: Determine the compressibility and deformation characteristics of the soil. Shear test: Determine the shear strength of the soil and evaluate its stability. Permeability test: Determine the permeability of the soil to evaluate the impact of groundwater on the stability of the foundation. Physical property measurement: such as the measurement of parameters such as density, water content, and particle grading.

[0065] Collect existing engineering project data in the area, including foundation treatment methods, construction records and monitoring results. Use IoT devices, such as settlement sensors, inclinometers and groundwater level monitoring equipment, to collect foundation data in real time; set up a sensor network in the construction area, including settlement sensors, inclinometers, groundwater level monitoring equipment, seismic sensors, soil moisture sensors and temperature sensors; the sensor network collects various data in real time in the construction area, including settlement, settlement rate, cumulative settlement, inclination angle, groundwater level, water quality parameters and soil moisture; pre-process the foundation-related data collected above, and generate a collection of foundation-related data in the construction area after aggregation;

[0066] When using, combine the contents in steps 101 to 103:

[0067] By collecting, monitoring and testing various data in the construction area, comprehensive initial data acquisition helps to obtain more accurate foundation coefficients when predicting and identifying foundation coefficients;

[0068] The existing foundation coefficient detection method usually collects various data in the construction area, and then obtains it through data analysis and calculation. This identification method is usually simple, but inefficient and requires regular measurements. Not only is the real-time feedback effect poor, but the detection and identification efficiency is actually low. When there are abnormal environmental conditions in the construction area, especially when there is more precipitation and soil moisture content in the construction area and adjacent areas, the foundation coefficient in the construction area may fluctuate to a certain extent, resulting in foundation construction risks in the construction area, which are difficult to detect and deal with in time.

[0069] Step 2: After the foundation coefficient identification model is trained and obtained, it is tested and a performance value is generated by the obtained performance index. If the performance value is lower than the performance threshold, the foundation coefficient identification model is optimized;

[0070] The step 2 includes the following contents:

[0071] Step 201: extract some data from the foundation-related data set as sample data, annotate the sample data, train a convolutional neural network with the annotated sample data, and obtain a trained foundation coefficient recognition model;

[0072] Test the foundation coefficient identification model and monitor its performance in real time to obtain corresponding performance indicators, such as accuracy, recall rate, mean square error, silhouette coefficient and Brill score, and summarize the obtained performance indicators to generate a model indicator data set;

[0073] When in use, after extracting sample data for training and obtaining the foundation coefficient identification model, the corresponding performance index can be obtained through performance testing, and the model performance can be evaluated based on the performance index;

[0074] Step 202: Generate performance values ​​of the foundation coefficient identification model from the model index data set, where the mean square error , silhouette coefficient Brill Rating Perform linear normalization and map the corresponding data values ​​to the interval According to the following formula: in, , and is the weight coefficient, and its value falls within Within, the weight coefficient is obtained by referring to the hierarchical analysis method;

[0075] Combined with historical data and management expectations for model performance, a performance threshold is set in advance; if the performance value is lower than the performance threshold, it means that the recognition and prediction functions of the trained foundation coefficient recognition model cannot meet expectations and cannot achieve the effect of recognition and prediction of foundation coefficients. At this time, the recognition model needs to be optimized, such as data set optimization, model structure optimization, hyperparameter tuning, deep learning and transfer learning, etc.

[0076] When using, combine the contents in steps 201 and 202:

[0077] Corresponding performance values ​​are obtained on the basis of obtaining several performance indicators. The performance of the foundation coefficient identification model can be comprehensively judged according to the performance values. If the performance of the foundation coefficient identification model is poor, the foundation coefficient identification model can be optimized according to the obtained performance test data, thereby improving the reliability of the identification result when it is used to identify the foundation coefficient.

[0078] Step 3: Use the optimized foundation coefficient identification model to identify the foundation coefficients of each detection point in the construction area. If there is an abnormality in the foundation coefficient, send a multi-level alarm instruction to the outside through multiple channels;

[0079] The step three includes the following contents:

[0080] Step 301, select a number of evenly distributed detection points in the construction area, detect and obtain corresponding foundation coefficients at the detection points, use them as detection data, use real-time foundation-related data at the detection points as input, and output corresponding identification data from the optimized foundation coefficient identification model;

[0081] Perform similarity analysis between the detection data and the recognition data, and obtain the corresponding data similarity;

[0082] If the similarity is lower than the preset similarity threshold, it means that the reliability of the optimized foundation coefficient identification model is still insufficient. At this time, the optimized foundation coefficient identification model is subjected to secondary optimization, such as data set optimization, model structure optimization, hyperparameter tuning, deep learning and transfer learning, to obtain the optimized foundation coefficient identification model again;

[0083] When using, on the basis of one optimization, when the trained foundation coefficient identification model is used in actual application, the data obtained by detection and the identification data are compared, and the data similarity is used as a judgment index to verify whether the reliability of the foundation coefficient identification model can meet expectations. If it still cannot meet expectations, it can be optimized again to fully guarantee the reliability of the foundation coefficient identification model;

[0084] Step 302: using the real-time foundation-related data as input, using the optimized foundation coefficient identification model to identify the foundation coefficient of each detection point in the construction area; training the machine learning algorithm with the labeled sample data to obtain the trained abnormal data identification model; using the identified foundation coefficient as input, using the trained abnormal data identification model to identify it, and if the foundation coefficient at the detection point is abnormal, issuing an early warning instruction to the outside;

[0085] When in use, after optimizing the foundation coefficient identification model, on the basis of monitoring various foundations and related data in the construction area, the foundation coefficient at various locations in the construction area is identified through the foundation coefficient identification model, completing the real-time monitoring and identification process, and when the foundation coefficient is abnormal, corresponding processing can be made in time; at the same time, compared with manual detection and analysis, it can also realize the rapid identification of the foundation coefficient in the construction area, with higher efficiency, accuracy and reliability.

[0086] Step 303: If the number of warning instructions received continuously within the preset observation period exceeds the expectation, the time node of receiving the warning instruction is used as the warning node. Under dimensionless conditions, a warning value is constructed according to the warning node and the warning position information in the following manner: Where: and is the weight coefficient, obtained by referring to the hierarchical analysis method, For the observation period The number of warning nodes in For time The number of warning instructions received at For time Warning location information of the place; is the modulus of the location information vector, which represents the measure of the warning location information; Used to filter out those that exceed the expected number of time point, ;

[0087] Pre-set warning thresholds based on safety management expectations for foundation coefficients in the construction area;

[0088] If the warning value exceeds the preset warning threshold, it means that the foundation coefficient in the construction area is highly abnormal, and the risk of potential safety hazards is also high. Safety measures need to be taken in the construction area. At this time, an alarm instruction is issued to the outside through multiple channels such as SMS and applications.

[0089] When using, combine the contents in steps 301 to 303:

[0090] On the basis of detecting and identifying the foundation coefficient in the construction area, when the foundation coefficient in the construction area is abnormal, the degree of abnormality can be judged. When the degree of abnormality exceeds expectations, an alarm instruction can be issued to the outside to realize multi-level alarm for foundation coefficient abnormality and ensure the safety of subsequent construction in the construction area.

[0091] Step 4: After adjusting the warning threshold using the trained threshold adjustment model, the potential danger area is screened out in the construction area according to the location of the detection point and the value of the foundation coefficient;

[0092] The step 4 includes the following contents:

[0093] Step 401: perform correlation analysis on the environmental anomaly degree and the foundation coefficient at each detection point to obtain a corresponding correlation coefficient. If the value of the correlation coefficient exceeds a preset correlation threshold, a machine learning algorithm is trained based on historical monitoring data and actual engineering experience as sample data to obtain a trained threshold adjustment model.

[0094] After receiving the alarm command, the trained threshold adjustment model is used to adjust the current warning threshold; the threshold is adjusted dynamically to adapt to different environments and conditions;

[0095] When in use, based on the identification of the foundation coefficient in the construction area, the trained threshold adjustment model adjusts the current warning threshold so that the warning threshold is in a dynamic adjustment state and adapts to the environmental conditions in the construction area;

[0096] Step 402: Mark each detection point with the foundation coefficient, divide the construction area into several sub-areas using the trained classifier according to the position of the detection point and the value of the foundation coefficient, and screen out potential danger areas in the several sub-areas according to the expectation of foundation safety;

[0097] When using, combine the contents in steps 401 and 402:

[0098] The foundation coefficient of each detection point is obtained. After identifying and screening the potential danger areas in the construction area through the value of the foundation coefficient and its location information, targeted treatment measures can be taken for the potential danger areas to reduce the construction risk when the foundation coefficient does not meet expectations.

[0099] Step 5: Using the real-time risk data in the potential hazard area as input, the trained construction risk identification model is used to score the data, and the corresponding risk treatment measures are matched in the potential hazard area according to the obtained risk score;

[0100] The step five includes the following contents:

[0101] Step 501: Acquire environmental condition data, foundation coefficient, construction data, etc. in the hidden danger area as real-time risk data; train a machine learning algorithm with the labeled sample data to obtain a trained construction risk identification model, use the real-time risk data as input, use the trained construction risk identification model to score, obtain the risk score of each sub-area or hidden danger area, and label the hidden danger area or each sub-area according to the risk score;

[0102] When in use, after screening out the potential danger areas, risk assessment is performed on the potential danger areas. Based on the obtained risk scores, the potential danger levels in the potential danger areas can be evaluated, so as to facilitate the judgment of the potential danger levels.

[0103] Step 502: pre-set a risk threshold based on historical data and management expectations for construction risks; divide the foundation construction risks in the sub-area or potential hazard area into different risk levels based on the relationship between the risk score and the risk threshold, and match the corresponding risk treatment measures in the sub-area based on the risk level, including: for low-risk sub-areas, continuously monitor the status and do not take additional measures; for medium-risk sub-areas, periodically inspect each detection point in the construction area and perform reinforcement; for high-risk sub-areas, take emergency reinforcement measures and suspend construction;

[0104] After the risk treatment measures are executed, a monitoring and treatment report is automatically generated. The report content includes: basic information: project name, monitoring date, monitoring location, etc.; monitoring data: real-time monitoring data including foundation coefficient, settlement, inclination angle, groundwater level, etc.; analysis results: multi-model fusion prediction results, anomaly detection results, trend analysis results, comprehensive risk score, etc.; early warning information: multi-level early warning and alarm triggering conditions, detailed response measures and suggestions; change trend: comparative analysis with historical monitoring data to show the long-term trend of foundation changes.

[0105] When using, combine the contents in steps 501 and 502:

[0106] When the foundation coefficient at some locations in the construction area is different from the expected one, which may lead to potential safety hazards, targeted treatment measures can be taken for the potential danger area based on the risk level of the potential danger area (actually not limited to the potential danger area, all sub-areas within the construction area are acceptable) to ensure the safety of the construction area.

[0107] The Analytic Hierarchy Process (AHP) is a decision-making method that decomposes decision-related elements into levels such as goals, criteria, and plans, and conducts qualitative and quantitative analysis on this basis. It is particularly suitable for dealing with target systems with hierarchical and staggered evaluation indicators, and when the target value is difficult to describe quantitatively, the AHP is an effective decision-making tool.

[0108] The core of the hierarchical analysis method is to decompose the decision problem into multiple levels to form a hierarchical structure, which usually includes the target level, the criterion level, the sub-criterion level and the solution level. By solving the eigenvector of the judgment matrix, the priority weight of each element of each level to an element of the previous level is obtained, and the final weight of each alternative solution to the total goal is recursively merged by the weighted sum method, so as to find the optimal solution.

[0109] See also Figure 2 The present invention provides a fully automatic identification system for foundation coefficients, comprising:

[0110] An environmental analysis unit generates an environmental anomaly degree from environmental condition data in the construction area, and if the acquired environmental anomaly degree exceeds an anomaly threshold, collects foundation-related data in the construction area;

[0111] The model performance analysis unit, after training and acquiring the foundation coefficient identification model, tests it and generates a performance value based on the acquired performance index. If the performance value is lower than the performance threshold, the foundation coefficient identification model is optimized;

[0112] The abnormal alarm unit uses the optimized foundation coefficient identification model to identify the foundation coefficient of each detection point in the construction area. If there is an abnormality in the foundation coefficient, a multi-level alarm instruction is issued to the outside through multiple channels;

[0113] The area recognition unit uses the trained threshold adjustment model to adjust the warning threshold, and then selects the potential danger area in the construction area according to the location of the detection point and the value of the foundation coefficient;

[0114] The emergency response unit takes the real-time risk data in the hidden danger area as input, uses the trained construction risk identification model to score it, and matches the corresponding risk treatment measures in the hidden danger area according to the obtained risk score.

[0115] See also Figure 3, fully automatic identification device for foundation coefficient, including,

[0116] At least one processor; a memory for storing executable instructions, wherein the instructions are executed by at least one of the processors to enable at least one of the processors to perform the steps of the method for fully automatic identification of foundation coefficients.

[0117] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0118] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0119] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0120] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0121] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0122] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0123] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0124] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A fully automatic identification method for foundation coefficients, characterized in that: include, The environmental anomaly degree Ytp is generated from the environmental condition data in the construction area. If the obtained environmental anomaly degree Ytp exceeds the anomaly threshold, the foundation-related data is collected in the construction area and summarized to generate a foundation-related data set; wherein, the environmental anomaly degree Ytp is generated from the environmental condition data in the environmental condition data set in the following manner: Where: H(t) is the environmental coefficient, Rp(t), Hs(t) and Gm(t) are the precipitation, soil moisture and groundwater level at time t, respectively, and F p 、F h and F G are the corresponding weight coefficients respectively; μ E is the mean value of the environmental coefficient, σ E is the historical standard deviation of the environmental coefficient; After the foundation coefficient identification model is trained, it is tested and the performance value E(t) generated by the obtained performance index is obtained. If the performance value is lower than the performance threshold, the foundation coefficient identification model is optimized; wherein, the mean square error Me, the silhouette coefficient Lk and the Brill score Bm are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and the performance value E(t) is generated according to the following formula: E(t)=S1*Me+S2*Lk+S3*Bm Among them, S1, S2 and S3 are weight coefficients, and their values ​​all fall within [0, 1]; Step 3: Use the optimized foundation coefficient identification model to identify the foundation coefficient of each detection point in the construction area. If the foundation coefficient is abnormal, issue a multi-level alarm command to the outside through multiple channels; if the number of warning commands received continuously within the preset observation period exceeds the expectation, the time node at which the warning command is received is used as the warning node, and the warning value Fxt is constructed based on the warning node and the warning location information. If the warning value Fxt exceeds the preset warning threshold, an alarm command is issued to the outside; the warning value Fxt is constructed in the following way: Where: α and β are weight coefficients, obtained by referring to the hierarchical analysis method, N is the number of warning nodes in the observation period T, n(t) is the number of warning instructions received at time t, P(t) is the warning location information at time t; ||P(t)|| is the modulus length of the location information vector, which represents the measurement of the warning location information; H(x) is used to screen those time points that exceed the expected number N, After adjusting the warning threshold using the trained threshold adjustment model, the potential danger areas are screened out in the construction area according to the location of the detection point and the value of the foundation coefficient; The real-time risk data in the hidden danger area is used as input, and the trained construction risk identification model is used to score it. According to the obtained risk score, the corresponding risk treatment measures are matched in the hidden danger area.

2. The fully automatic identification method of foundation coefficient according to claim 1 is characterized in that: Environmental condition data is periodically monitored and collected in the construction area, and the environmental condition data in the area is generated after being summarized. The environmental abnormality degree Ytp is generated from the environmental condition data in the environmental condition data set; If the obtained environmental abnormality Ytp exceeds the abnormal threshold, various test data and experimental data are collected in the construction area; A sensor network is set up in the construction area, and the sensor network collects various data in real time in the construction area, including settlement amount, settlement rate, cumulative settlement, inclination angle, groundwater level, water quality parameters and soil moisture, and summarizes them to generate a set of foundation-related data in the construction area.

3. The fully automatic identification method of foundation coefficient according to claim 2 is characterized in that: Extract part of the data from the foundation-related data set to obtain the trained foundation coefficient identification model; test the foundation coefficient identification model and monitor its performance in real time to obtain corresponding performance indicators, which are summarized to generate a model indicator data set; and generate a performance value E(t) of the foundation coefficient identification model from the model indicator data set.

4. The fully automatic identification method of foundation coefficient according to claim 3 is characterized in that: Select several evenly distributed detection points in the construction area, detect and obtain the corresponding foundation coefficients at the detection points, and use them as detection data. Use the real-time foundation-related data at the detection points as input, and the output of the optimized foundation coefficient identification model is the corresponding identification data; Perform similarity analysis between the detection data and the recognition data, and obtain the corresponding data similarity; If the similarity is lower than a preset similarity threshold, the optimized foundation coefficient identification model is optimized.

5. The fully automatic identification method of foundation coefficient according to claim 4 is characterized in that: Taking real-time foundation-related data as input, the optimized foundation coefficient identification model is used to identify the foundation coefficient of each detection point in the construction area; taking the identified foundation coefficient as input, the trained abnormal data identification model is used to identify it. If the foundation coefficient at the detection point is abnormal, an early warning instruction is issued to the outside.

6. The fully automatic identification method of foundation coefficient according to claim 5 is characterized in that: Perform correlation analysis on the environmental anomaly degree and the foundation coefficient. If the value of the correlation coefficient exceeds the pre-correlation threshold, obtain the trained threshold adjustment model after training the machine learning algorithm. After receiving the alarm instruction, use the trained threshold adjustment model to adjust the current warning threshold. Each detection point is marked with the foundation coefficient. According to the location of the detection point and the value of the foundation coefficient, the construction area is divided into several sub-areas using the trained classifier, and the potential danger areas are screened out in several sub-areas.

7. The fully automatic identification method of foundation coefficient according to claim 6 is characterized in that: Obtain environmental condition data, foundation coefficient and construction data in the potential hazard area as real-time risk data; Using real-time risk data as input, the trained construction risk identification model is used to score and obtain the risk score of the potential hazard area; Based on the relationship between risk score and risk threshold, the foundation construction risks in the potential danger area are divided into different risk levels, and corresponding risk treatment measures are matched in the sub-areas according to the risk level; after the risk treatment measures are implemented, a monitoring and treatment report is automatically generated.

8. A fully automatic identification system for foundation coefficients, using the fully automatic identification method for foundation coefficients according to any one of claims 1 to 7, characterized in that: include, An environmental analysis unit generates an environmental anomaly degree from environmental condition data in the construction area, and if the acquired environmental anomaly degree exceeds an anomaly threshold, collects foundation-related data in the construction area; The model performance analysis unit, after training and acquiring the foundation coefficient identification model, tests it and generates a performance value based on the acquired performance index. If the performance value is lower than the performance threshold, the foundation coefficient identification model is optimized; The abnormal alarm unit uses the optimized foundation coefficient identification model to identify the foundation coefficient of each detection point in the construction area. If there is an abnormality in the foundation coefficient, a multi-level alarm instruction is issued to the outside through multiple channels; The area recognition unit uses the trained threshold adjustment model to adjust the warning threshold, and then selects the potential danger area in the construction area according to the location of the detection point and the value of the foundation coefficient; The emergency response unit takes the real-time risk data in the hidden danger area as input, uses the trained construction risk identification model to score it, and matches the corresponding risk treatment measures in the hidden danger area according to the obtained risk score.

9. The fully automatic identification device of foundation coefficient is characterized by: include, at least one processor; A memory for storing executable instructions, wherein the instructions are executed by at least one of the processors to enable at least one of the processors to perform the steps of the method according to any one of claims 1 to 7.

Images