Method for intelligently and rapidly judging corrosivity grade of gypsum-salt-containing stratum soil water

Abstract

The invention belongs to the technical field of soil and water corrosion grade detection, discloses an intelligent rapid distinguishing method for a gypsum salt-containing stratum soil and water corrosion grade, and aims to solve the problems of long time period, high cost and low detection efficiency of the soil and water corrosion grade through multiple groups of indoor tests. The method comprises the following steps: (1) collecting information of existing sampling points in an acquisition area; (2) taking the longitude A1, the latitude A2 and the elevation A3 of the sampling point as input parameters, and taking the corrosion grade B1 of soil to concrete and the corrosion grade B2 of water to concrete at the sampling point as output parameters; (3) taking A1, A2, A3, B1 and B2 as training data to carry out training and learning; and (4) utilizing a BP neural network model to carry out discrimination. According to the invention, the BP neural network model is utilized to rapidly obtain the water and soil corrosivity grade of a certain undetected site.

Description

Technical Field

[0001] The invention belongs to the technical field of soil and water corrosivity detection, and specifically relates to an intelligent and rapid discrimination method for soil and water corrosivity levels in gypsum-salt-containing strata. Background Art

[0002] In the process of urban infrastructure construction, it is an important task to confirm the soil and water corrosiveness level of the proposed construction site. Water and soil corrosiveness test refers to the determination of the chemical, physicochemical or biochemical erosion ability of water and soil in the engineering geological survey site on concrete and steel materials under certain environmental conditions. When the concrete structure is above the groundwater level, soil samples should be taken for soil corrosiveness level test; when it is in groundwater or surface water, water samples should be taken for water corrosiveness level test; when part of the structure is above the groundwater level and part is below the groundwater level, soil samples and water samples should be taken for corrosion level test respectively; only when there is sufficient experience and data to believe that the corrosion of water and soil on the construction site to the building foundation can be ignored, water and soil corrosion test can be omitted.

[0003] Especially for the highly corrosive strata containing gypsum and calcium sulfate, it is particularly important to determine the soil and water corrosiveness level of the site. At present, the soil and water corrosiveness levels are generally determined through multiple sets of indoor tests. For example, the test items for water corrosion on concrete structures include: pH value, Ca 2+ Mg 2+ , Cl － 、CO3 2- 、HCO3 - OH － , free CO2, corrosive CO2, SO4 2－ NH + 4. Total mineralization; soil corrosion to concrete structure test items include pH value, Ca 2+ Mg 2+ , Cl － 、SO4 2－ 、CO3 2- 、HCO3 - Analysis of soluble salts (soil-water ratio 1:5). The corrosiveness of water and soil to building structures can be divided into four levels: slight, weak, medium and strong.

[0004] Therefore, indoor tests on site have problems such as long time period and high cost, and are particularly inconvenient for rapid prediction during on-site investigations. Summary of the invention

[0005] In order to solve the problem that the soil and water corrosiveness levels of construction sites are measured through multiple groups of indoor tests, which has the disadvantages of long time period, high cost and low detection efficiency, an intelligent and rapid identification method for the soil and water corrosiveness levels of gypsum-salt strata is provided. The existing data in the existing area is used as the data source, and a soil and water corrosiveness level determination model suitable for the area is established. When used, the location information of the site to be tested is input into the model, so that the soil and water corrosiveness level information at the location can be quickly obtained, and indoor test detection is not required, thereby improving the efficiency of corrosion level detection.

[0006] In order to solve the technical problem, the technical solution adopted by the present invention is:

[0007] An intelligent and rapid method for distinguishing the corrosiveness level of soil and water in gypsum-salt formations, characterized by comprising:

[0008] (1) Collect and obtain the coordinate information, stratum information, soil corrosivity grade information, and water corrosivity grade information of the existing sampling points in the area; wherein the coordinate information of the sampling points includes the longitude, latitude, and elevation information of the sampling points;

[0009] (2) The collected data information is sorted and identified, and the longitude A1, latitude A2 and elevation A3 of the sampling point are used as input parameters, and the corrosiveness level B1 of the soil on concrete and the corrosiveness level B2 of the water on concrete at the sampling point are used as output parameters;

[0010] (3) Using the BP neural network algorithm to construct a BP neural network model, and using the sorted groups of input parameters A1, A2, A3 and output parameters B1, B2 as training data to train and learn the component BP neural network model;

[0011] (4) The trained BP neural network model is used to judge the soil corrosivity level and water corrosivity level at a certain location in the area.

[0012] In some embodiments, in step (4), the longitude, latitude and elevation of the certain location are used as inputs of the BP neural network model, and the BP neural network model obtains the soil corrosivity level and water corrosivity level of the certain location through calculation.

[0013] In some embodiments, the engineering geological conditions of the certain place in step (4) should be the same as those of the area in step (1); wherein the engineering geological conditions include topography, stratum lithology, geological structure, hydrogeology and surface geology.

[0014] In some embodiments, the number of sampling points in step (1) shall not be less than 200.

[0015] In some embodiments, the sampling points in step (1) should cover the area.

[0016] In some embodiments, the sorting and screening of the collected data information described in step (2) at least includes eliminating data from sampling points with obvious errors.

[0017] In some embodiments, after completing step (4), sampling verification is performed at a certain position in the area; if the verified corrosive level information is consistent with the judgment of the BP neural network model, the data is directly included in the database of the BP neural network model; if the verified corrosive level information is inconsistent with the judgment of the BP neural network model, the sampling verification shall prevail, and the sampling verification data shall be included in the database of the BP neural network model; in this way, the BP neural network model is continuously trained and improved, and the accuracy of the judgment results of the BP neural network model is gradually improved.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] The intelligent and rapid identification method for soil and water corrosivity level of gypsum-salt-containing strata of the present invention has the same engineering geological conditions for multiple sites belonging to the same area. When the water and soil corrosivity level of a certain site is identified, data from other sites that have been constructed and sampled can be used to construct and train a BP neural network model, and the BP neural network model can be used to identify the corrosivity level of a certain undetected (unconstructed) site, thereby quickly obtaining the corrosivity level of water and soil of the undetected (unconstructed) site to concrete structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a schematic diagram of a flow chart of the present invention;

[0021] Figure 2 This is the principle block diagram of the BP neural network. DETAILED DESCRIPTION

[0022] The present invention is further described below in conjunction with embodiments, which are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, other embodiments obtained by ordinary technicians in the field without creative work are all within the scope of protection of the present invention.

[0023] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention; the terms "first", "second", and "third" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance; in addition, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] Combined with Figure 1 and attached Figure 2 The intelligent and rapid identification method of soil and water corrosivity level in gypsum-salt-containing strata of the present invention comprises:

[0025] (1) Collect and obtain the coordinate information, stratum information, soil corrosivity grade information, and water corrosivity grade information of the existing sampling points in the area; wherein the coordinate information of the sampling points includes the longitude, latitude, and elevation information of the sampling points;

[0026] (2) The collected data information is sorted and identified, and the longitude A1, latitude A2 and elevation A3 of the sampling point are used as input parameters, and the corrosiveness level B1 of the soil on concrete and the corrosiveness level B2 of the water on concrete at the sampling point are used as output parameters;

[0027] (3) Using the BP neural network algorithm to construct a BP neural network model, and using the sorted groups of input parameters A1, A2, A3 and output parameters B1, B2 as training data to train and learn the component BP neural network model;

[0028] (4) The trained BP neural network model is used to judge the soil corrosivity level and water corrosivity level at a certain location in the area.

[0029] Combined with Figure 2, BP neural network model is a commonly used model in machine models. BP neural network is a multi-layer feedforward network trained by error back propagation (abbreviated as error back propagation). Its algorithm is called BP algorithm. Its basic idea is gradient descent method, which uses gradient search technology to minimize the mean square error between the actual output value and the expected output value of the network. The basic BP algorithm includes two processes: forward propagation of signals and back propagation of errors. That is, when calculating the error output, it is done from input to output, while adjusting the weights and thresholds from output to input. During forward propagation, the input signal acts on the output node through the hidden layer, and after nonlinear transformation, the output signal is generated. If the actual output does not match the expected output, it enters the error back propagation process. Error back propagation is to propagate the output error through the hidden layer to the input layer layer by layer, and distribute the error to all units in each layer, and use the error signal obtained from each layer as the basis for adjusting the weights of each unit. By adjusting the connection strength between the input node and the hidden layer node, the connection strength between the hidden layer node and the output node, and the threshold, the error is reduced along the gradient direction. After repeated learning and training, the network parameters (weights and thresholds) corresponding to the minimum error are determined, and the training is stopped. Therefore, the BP neural network model is very suitable for the judgment of the water and soil corrosivity level of the present invention.

[0030] In some embodiments, in step (4), the longitude, latitude and elevation of the certain location are used as inputs of the BP neural network model, and the BP neural network model obtains the soil corrosivity level and water corrosivity level of the certain location through calculation.

[0031] In some embodiments, the engineering geological conditions of the certain place in step (4) should be the same as those of the area in step (1); wherein the engineering geological conditions include topography, stratum lithology, geological structure, hydrogeology and surface geology.

[0032] In some embodiments, the number of sampling points in step (1) shall not be less than 200. In actual application, as much data of sampling points as possible should be collected to better train the BP neural network model.

[0033] In some embodiments, the sampling points in step (1) should cover the area. That is, the sampling points cannot come from a certain slice of the area. For example, if the area is divided into 10 slices, then the data of the sampling points should include all 10 slices in the area, and the sampling points cannot be selected from only one or two slices; if the area is divided into 5 slices, then the data of the sampling points should include all 5 slices in the area.

[0034] In some embodiments, the sorting and screening of the collected data information described in step (2) at least includes eliminating data from sampling points with obvious errors.

[0035] In some embodiments, after completing step (4), sampling verification is performed at a certain position in the area; if the verified corrosive level information is consistent with the judgment of the BP neural network model, the data is directly included in the database of the BP neural network model; if the verified corrosive level information is inconsistent with the judgment of the BP neural network model, the sampling verification is used as the standard, and the sampling verification data is included in the database of the BP neural network model; in this way, the BP neural network model is continuously trained and improved, and the accuracy of the judgment results of the BP neural network model is gradually improved. In other words, the BP neural network model is continuously trained.

[0036] The intelligent and rapid identification method for soil-water corrosivity level of gypsum-salt strata of the present invention has the same engineering geological conditions for multiple sites in the same area. When identifying the water and soil corrosivity level of a certain site, data from other sites that have been constructed and sampled can be used to construct and train a BP neural network model, and the BP neural network model can be used to identify the corrosivity level of a certain site that has not been constructed, thereby quickly obtaining the corrosivity level of water and soil on concrete structures in the certain site that has not been constructed.

[0037] Before construction of a site, an engineering survey will be conducted. In order to meet the needs of planning, design, construction, operation and comprehensive management of engineering construction, the terrain, geology and hydrology are surveyed, explored and tested, and the corresponding results are provided. However, the engineering surveys of various sites are currently conducted independently, resulting in a large waste of data resources. Therefore, the present invention collects data and constructs a BP neural network model, and uses the trained BP neural network model to reduce the subsequent workload, which has great economic significance.

Claims

1. An intelligent and rapid method for distinguishing the corrosiveness level of soil and water in gypsum-salt formations, characterized in that: include: (1) Collect and obtain the coordinate information, stratum information, soil corrosivity grade information, and water corrosivity grade information of the existing sampling points in the area; wherein the coordinate information of the sampling points includes the longitude, latitude, and elevation information of the sampling points; (2) The collected data information is sorted and identified, and the longitude A1, latitude A2 and elevation A3 of the sampling point are used as input parameters, and the corrosiveness level B1 of the soil on concrete and the corrosiveness level B2 of the water on concrete at the sampling point are used as output parameters; (3) Using the BP neural network algorithm to construct a BP neural network model, and using the sorted groups of input parameters A1, A2, A3 and output parameters B1, B2 as training data to train and learn the component BP neural network model; (4) The trained BP neural network model is used to judge the soil corrosivity level and water corrosivity level at a certain location in the area.

2. The intelligent and rapid identification method for soil and water corrosivity level in gypsum-salt formations according to claim 1 is characterized in that: In step (4), the longitude, latitude and elevation of the certain location are used as inputs of the BP neural network model, and the BP neural network model obtains the soil corrosivity level and water corrosivity level of the certain location through calculation.

3. The intelligent and rapid identification method for soil and water corrosivity level in gypsum-salt formations according to claim 1 is characterized in that: The engineering geological conditions of the certain place in step (4) should be the same as those of the area in step (1); wherein the engineering geological conditions include topography, stratum lithology, geological structure, hydrogeology and surface geology.

4. The intelligent and rapid identification method for soil water corrosivity level in gypsum-salt strata according to claim 1 is characterized in that: The number of sampling points in step (1) shall not be less than 200.

5. The intelligent and rapid identification method for soil and water corrosivity level in gypsum-salt formations according to claim 4 is characterized in that: The sampling points in step (1) should cover this area.

6. The intelligent and rapid identification method for soil and water corrosivity level in gypsum-salt formations according to claim 1 is characterized in that: The sorting and screening of the collected data information described in step (2) at least includes eliminating the data of the sampling points with obvious errors.

7. The intelligent and rapid identification method for soil water corrosivity level in gypsum-salt formations according to any one of claims 1 to 6, characterized in that: After completing step (4), sampling verification is performed at a certain position in the area; if the verified corrosion level information is consistent with the judgment of the BP neural network model, the data is directly included in the database of the BP neural network model; if the verified corrosion level information is inconsistent with the judgment of the BP neural network model, the sampling verification shall prevail, and the sampling verification data shall be included in the database of the BP neural network model; in this way, the BP neural network model is continuously trained and improved, and the accuracy of the judgment results of the BP neural network model is gradually improved.

Images