Soil remediation technology evaluation method and system

Abstract

The invention discloses a soil remediation technology assessment method and system, and the method comprises the steps: extracting key parameters of target soil, obtaining a series of remediation schemes described in a parameterization manner from a pre-established soil remediation scheme database, and processing the data through employing an embedded coding technology and a vectorization coding technology based on deep learning, and mining semantic information of the parameterized description of the target soil and the parameterized description of the soil remediation scheme. And further, semantic query is performed on the target soil and each remediation scheme, so that deep matching between the target soil characteristics and the remediation schemes is realized, and the remediation scheme of the target soil is intelligently recommended based on a matching result. Thus, the soil and the remediation scheme can be accurately matched, the screening accuracy of the soil remediation scheme is improved, the recommendation result is made to better meet the actual soil requirement, then a direct and scientific action guide is provided for the soil remediation project, and efficient implementation of remediation work is guaranteed.

Description

Technical Field

[0001] The present application relates to the field of intelligent assessment, and more specifically, to a soil remediation technology assessment method and system. Background Art

[0002] In the field of soil remediation, with the acceleration of industrialization and urbanization, soil pollution problems are becoming increasingly serious. In order to effectively treat contaminated soil and ensure the sustainable use of land resources, soil remediation technology has made great progress. There are many different types of soil remediation methods, including physical, chemical, biological and other types of remediation technologies, each of which has its scope of application and limitations. However, a major challenge facing current soil remediation practices is the lack of an efficient, accurate and universal method to match the specific conditions of a specific contaminated site with the most appropriate remediation solution.

[0003] The traditional soil remediation process usually involves a series of discrete steps, such as field investigation, laboratory analysis, risk assessment, and the selection of the final remediation plan. These methods often rely on the experience and judgment of experts and reference to technical literature, which can easily lead to problems such as strong subjectivity, low efficiency, high cost, and inaccurate results in the formulation of remediation plans. In addition, due to the lack of unified standards and database support, the comparability and repeatability of soil remediation projects in different regions are also poor, further limiting the progress of soil remediation technology evaluation.

[0004] In recent years, with the development of information technology, digital management and intelligent decision support systems have begun to be introduced into the field of soil remediation. Chinese patent CN114819880A proposes a method and system for soil remediation and digital management. The invention constructs a database that characterizes and parameters soil remediation technology, and combines the contaminated soil database and the control soil database to provide users with a one-stop service platform from soil sample detection, remediation plan determination to effect evaluation. Although this method has optimized the supply of soil remediation services to a certain extent, it still faces challenges in practical applications. For example, how to more intelligently match the target soil with the most suitable remediation plan, and how to use advanced algorithms to achieve accurate recommendations for remediation paths, these are problems that existing technologies have not been able to fully solve.

[0005] Therefore, a soil remediation technology evaluation method and system are expected. Summary of the invention

[0006] In order to solve the above-mentioned technical problems, the present application is proposed. The embodiment of the present application provides a soil remediation technology evaluation method and system, which extracts the key parameters of the target soil, obtains a series of parametrically described remediation schemes from a pre-established soil remediation scheme database, and uses deep learning-based embedded coding technology and vector coding technology to process these data to mine the semantic information of the parametric description of the target soil and the parametric description of the soil remediation scheme. Furthermore, by performing semantic queries on the target soil and each remediation scheme, a deep-level match between the target soil characteristics and the remediation scheme is achieved, and the remediation scheme for the target soil is intelligently recommended based on the matching results. In this way, the soil and the remediation scheme can be accurately matched, the accuracy of the screening of soil remediation schemes can be improved, and the recommended results can be more in line with the actual soil needs, thereby providing a direct and scientific action guide for soil remediation projects and ensuring the efficient implementation of remediation work.

[0007] According to one aspect of the present application, a soil remediation technology evaluation method is provided, which includes: Extract the parameters of the target soil; extracting a set of parameterized descriptions of soil remediation schemes from a soil remediation scheme database; Embedding the set of the target soil parameters and the soil remediation scheme parameterized description to obtain a set of embedded coding features of the target soil parameterized description and embedded coding features of the soil remediation scheme parameterized description; Performing feature dynamic semantic query encoding based on selection range ratio anchoring on the set of the target soil parameterized description embedded coding features and the soil remediation solution parameterized description embedded coding features to obtain a target soil-remediation solution query response coding feature; Based on the target soil-remediation solution query response coding features, a remediation solution for the target soil is recommended.

[0008] According to another aspect of the present application, a soil remediation technology evaluation system is provided, which includes: A target soil parameter extraction module is used to extract the parameters of the target soil; A soil remediation scheme parameterized description extraction module is used to extract a set of soil remediation scheme parameterized descriptions from a soil remediation scheme database; An embedded coding module, used for embedding the set of the target soil parameters and the soil remediation scheme parameterized description to obtain a set of embedded coding features of the target soil parameterized description and embedded coding features of the soil remediation scheme parameterized description; A feature dynamic semantic query encoding module is used to perform feature dynamic semantic query encoding based on selection range ratio anchoring on the set of the target soil parameterized description embedded encoding features and the soil remediation solution parameterized description embedded encoding features to obtain a target soil-remediation solution query response encoding feature; The restoration scheme recommendation module is used to recommend a restoration scheme for the target soil based on the target soil-restoration scheme query response coding feature.

[0009] Compared with the prior art, the present application provides a soil remediation technology evaluation method and system, which extracts the key parameters of the target soil and obtains a series of parametrically described remediation solutions from a pre-established soil remediation solution database, and uses deep learning-based embedded coding technology and vector coding technology to process these data to mine the semantic information of the parametric description of the target soil and the parametric description of the soil remediation solution. Furthermore, by performing semantic queries on the target soil and each remediation solution, a deep-level match between the target soil characteristics and the remediation solution is achieved, and the remediation solution for the target soil is intelligently recommended based on the matching results. In this way, the soil and remediation solution can be accurately matched, the accuracy of soil remediation solution screening can be improved, and the recommendation results can be more in line with actual soil needs, thereby providing a direct and scientific action guide for soil remediation projects and ensuring the efficient implementation of remediation work. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0011] Figure 1 is a flow chart of a soil remediation technology evaluation method according to an embodiment of the present application; Figure 2 A data flow diagram of a soil remediation technology evaluation method according to an embodiment of the present application; Figure 3 This is a flowchart of sub-step S3 of the soil remediation technology evaluation method according to an embodiment of the present application; Figure 4 is a flowchart of sub-step S4 of the soil remediation technology evaluation method according to an embodiment of the present application; Figure 5 It is a block diagram of a soil remediation technology evaluation system according to an embodiment of the present application. DETAILED DESCRIPTION

[0012] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0013] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0014] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0015] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0016] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0017] In the technical solution of the present application, a soil remediation technology evaluation method and system are proposed. Figure 1 The present invention is a flow chart of a soil remediation technology evaluation method according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of the soil remediation technology evaluation method according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the embodiment of the present application, the soil remediation technology assessment method and system include the following steps: S1, extracting the parameters of the target soil; S2, extracting a set of parameterized descriptions of the soil remediation scheme from a soil remediation scheme database; S3, embedding and encoding the parameters of the target soil and the set of parameterized descriptions of the soil remediation scheme to obtain a set of embedded coding features of the target soil parameterized description and a set of embedded coding features of the soil remediation scheme parameterized description; S4, performing feature dynamic semantic query encoding based on selection range ratio anchoring on the set of embedded coding features of the target soil parameterized description and the set of embedded coding features of the soil remediation scheme parameterized description to obtain a target soil-remediation scheme query response coding feature; S5, recommending a remediation scheme for the target soil based on the target soil-remediation scheme query response coding feature.

[0018] In particular, the S1 extracts the parameters of the target soil. It should be understood that the complexity and diversity of soil pollution problems determine that each piece of contaminated soil has its own uniqueness, and these characteristics directly affect the difficulty of remediation work and the specific measures to be taken. Therefore, before formulating a remediation plan, it is essential to understand and master the actual conditions of the target soil. This not only helps to identify the types of pollutants and their concentration distribution, but also provides a basis for selecting the most appropriate remediation technology. In addition, accurate target soil parameters can significantly improve the pertinence of the remediation plan and avoid the problem of waste of resources and poor governance effects due to insufficient or incorrect information. Specifically, to achieve effective soil parameter extraction, detailed data collection is first required. This includes but is not limited to field investigation, sampling, and laboratory analysis. Among them, the field investigation aims to obtain first-hand information on soil types, geological structures, hydrological conditions, etc.; representative samples are collected from different depths and locations through sampling methods (grid method, random sampling method); laboratory analysis is used to determine the physical and chemical properties of the samples, such as pH value, redox potential (Eh), conductivity, organic matter content, heavy metal concentration, etc. These parameters can fully reflect the physical, chemical and biological characteristics of the soil and are of great value in predicting the restoration effect. For example, pH value and redox potential can indicate the acidity and redox state of the soil environment, which has a direct impact on certain microbial activities and chemical reactions; the levels of nutrients such as total nitrogen, phosphorus and potassium are related to the possibility of vegetation growth; and the presence and concentration of heavy metals and other toxic and harmful substances are one of the core factors that determine the urgency and difficulty of restoration.

[0019] In one example, first, to achieve effective soil parameter extraction, detailed data collection is required. This includes but is not limited to a comprehensive field survey of the target soil area. The field survey aims to obtain first-hand information on soil type, geological structure, hydrological conditions, etc., which are essential for understanding the basic properties of the soil. For example, understanding the soil texture (sand, loam, clay) can help predict the migration behavior of pollutants in the soil; and understanding the changes in groundwater levels can help determine whether it is necessary to consider the impact of groundwater flow on the spread of pollution. Subsequently, based on the field survey, representative soil samples need to be obtained through scientific and reasonable sampling methods. Grid method or random sampling method is usually used to collect samples from different depths and locations to ensure that the samples can truly reflect the average characteristics of the entire contaminated area. The depth and density of sampling depend on the complexity of soil contamination and the expected remediation goals. For shallow pollution, surface sampling can be taken; for deep pollution, drilling equipment may be required to assist in stratified sampling. In addition, considering the heterogeneity of pollutant distribution in the soil, appropriately increasing the number of sampling points can also improve the reliability of the data. Finally, the collected soil samples are sent to a professional laboratory for detailed physical and chemical property determination. Laboratory analysis is mainly used to accurately measure the content of various components in the sample, such as pH, redox potential (Eh), conductivity, organic matter content, heavy metal concentration, etc. These parameters can comprehensively reflect the physical, chemical and biological characteristics of the soil and are of great value in predicting the effect of restoration. For example, pH and redox potential can indicate the acidity and redox state of the soil environment, which has a direct impact on the activities of certain microorganisms and chemical reactions; the levels of nutrient elements such as total nitrogen, phosphorus, and potassium are related to the possibility of vegetation growth; and the presence or absence of heavy metals and other toxic and hazardous substances and their concentration are one of the core factors that determine the urgency and difficulty of restoration. In addition to the above-mentioned routine testing items, more in-depth research such as identification of specific pollutant types and analysis of microbial community structure may also be involved depending on the specific situation. All of these data will be integrated as part of the target soil parameters.

[0020] In particular, S2 extracts a set of parameterized descriptions of soil remediation schemes from a soil remediation scheme database. The soil remediation schemes in the soil remediation scheme database fully consider the differences between different types of pollutants (such as heavy metals, organic matter, radioactive substances, etc.), different geographical environmental conditions (such as climate zones, geological structures, etc.), and different land use types (such as agricultural land, industrial land, residential areas, etc.). In particular, to ensure the authority and timeliness of the data, the soil remediation scheme database usually integrates information resources on soil remediation from multiple channels such as government environmental protection departments, scientific research institutions, and university laboratories. Each remediation scheme contains a series of quantitative expressions of key attributes and conditions, including but not limited to soil properties (pH value, texture (sand, loam, clay), organic matter content, moisture content, etc.), pollution type (such as heavy metals, organic pollutants (such as petroleum hydrocarbons, pesticides), radioactive substances, etc.), pollution degree (such as pollutant concentration level), environmental conditions (temperature, humidity, etc.), remediation targets (target pollutant concentrations) and remediation methods. In one example, physical remediation methods may involve parameters such as excavation depth, soil replacement volume, and filter media; chemical remediation may focus on agent type, dosage, reaction time, etc.; biological remediation focuses more on microbial species, inoculation amount, nutritional supplementation, etc. By standardizing the definition of these parameters, errors caused by inconsistent terminology or misunderstandings can be effectively reduced. In the technical solution of the present application, by extracting a set of parameterized descriptions of soil remediation schemes, the standardized understanding of each soil remediation scheme can be strengthened, avoiding the ambiguity and uncertainty brought about by relying on qualitative descriptions in the past, and providing a solid foundation for improving the accuracy of soil remediation scheme screening.

[0021] In particular, the S3 embeds the set of the target soil parameters and the soil remediation scheme parameterized description to obtain a set of target soil parameterized description embedded coding features and soil remediation scheme parameterized description embedded coding features. In particular, in a specific example of the present application, Figure 3 As shown, the S3 includes: S31, embedding the parameters of the target soil based on the Word2Vec model to obtain a target soil parameterized description embedding coding vector as the target soil parameterized description embedding coding feature; S32, vectorizing and encoding each soil remediation scheme parameterized description in the set of soil remediation scheme parameterized descriptions to obtain a set of soil remediation scheme parameterized description embedding coding vectors as the set of soil remediation scheme parameterized description embedding coding features.

[0022] Specifically, in S31, the parameters of the target soil are embedded coded based on the Word2Vec model to obtain the embedded coding vector of the target soil parameterized description as the embedded coding feature of the target soil parameterized description. Since soil characteristics involve a variety of physical, chemical and biological parameters, these parameters are usually high-dimensional, and it is difficult to directly use the original data for analysis and comparison. The process of embedded coding is to map high-dimensional data to low-dimensional space, simplifying the data structure while retaining key information. Therefore, in the technical solution of the present application, the parameters of the target soil are embedded coded based on the Word2Vec model to obtain the embedded coding feature of the target soil parameterized description. The Word2Vec model can learn the associated combination characteristics between multiple parameters in soil characteristics, adapt to diverse soil parameter combinations, enable computers to process and analyze these data more efficiently, and improve the operating efficiency of the entire system.

[0023] In one example, each soil parameter, such as pH, redox potential, heavy metal concentration, etc., is regarded as a "vocabulary", and the entire soil sample constitutes a "document". By training on a "corpus" consisting of multiple soil samples, the Word2Vec model can learn the distribution patterns and intrinsic connections of these parameters under different soil environments. For a new target soil sample, its various parameter values ​​are regarded as points in a multidimensional vector space defined by the Word2Vec model. In this way, the originally high-dimensional and complex soil parameters are converted into a relatively low-dimensional but information-rich vector, namely the target soil parameterized description embedded coding vector. The target soil parameterized description embedded coding vector not only retains the specific values ​​of the original parameters, but also implies the possible interactions and synergistic effects between the parameters. For example, certain soil characteristics (such as heavy metal concentration) have a significant impact on the effect of a specific remediation technology, but this relationship is often difficult to capture by a simple distance metric; the Word2Vec model can more comprehensively reflect these complex interactions.

[0024] Specifically, the S32 vectorizes each soil remediation scheme parameterized description in the set of the soil remediation scheme parameterized description to obtain a set of soil remediation scheme parameterized description embedded coding vectors as a set of soil remediation scheme parameterized description embedded coding features. That is, each soil remediation scheme parameterized description in the set of the soil remediation scheme parameterized description is vectorized to obtain a set of soil remediation scheme parameterized description embedded coding vectors as a set of soil remediation scheme parameterized description embedded coding features. Among them, vectorized coding refers to a technology that maps high-dimensional data to low-dimensional space, which can simplify the data structure on the basis of retaining the original information, which is convenient for subsequent algorithm processing. By vectorizing the description of each soil remediation scheme parameterized description in the set of the soil remediation scheme parameterized description, not only the description of the remediation scheme is simplified, but also the expressiveness of the model is enhanced. Since each soil remediation scheme parameterized description is given a multidimensional vector representation, it is possible to capture the intrinsic connection between the parameters of the soil remediation scheme while retaining the original information to a greater extent, enhance the expressiveness of the model, and thus provide a solid foundation for subsequent intelligent matching. In the specific implementation, the category information (such as pollution type, remediation method, etc.) in the parametric description of the soil remediation scheme can be vectorized using One-Hot Encoding, and the parameter attribute text and numerical parameters can be vectorized using word embedding technology. Finally, the vectorized multi-dimensional vectors are combined as the embedded coding features of the parametric description of the soil remediation scheme, thereby unifying the data dimensions and providing strong data support for subsequent intelligent matching and recommendation.

[0025] In particular, the S4, the set of embedded coding features of the target soil parameterized description and the set of embedded coding features of the soil remediation scheme parameterized description are subjected to feature dynamic semantic query encoding based on the selection range ratio anchoring to obtain the target soil-remediation scheme query response coding features. It should be understood that since the set of embedded coding features of the target soil parameterized description and the set of embedded coding features of the soil remediation scheme parameterized description respectively contain the embedded semantic features of the target soil and the embedded semantic context features of each soil remediation scheme, in the matching process of the target soil-remediation scheme, the requirements of different application scenarios may be different. Direct semantic matching of the two is difficult to meet the diverse actual needs and lacks adaptive adjustment capabilities. In order to overcome the limitations in the matching process and realize the intelligent and precise matching of the target soil-remediation scheme, in the technical solution of the present application, advanced algorithms are further introduced to perform intelligent matching. Specifically, the set of embedded coding features of the target soil parameterized description and the set of embedded coding features of the soil remediation scheme parameterized description are subjected to feature dynamic semantic query encoding based on the selection range ratio anchoring to obtain the target soil-remediation scheme query response coding features. That is, by introducing a dynamic selection mechanism and semantic encoding technology, it is possible to automatically identify the feature representation that best represents the query intent in the complex soil remediation solution feature space, and accordingly define an optimized selection range. Within this range, the matching degree of each available soil remediation solution is further screened and evaluated to generate a semantic embedding representation that accurately reflects the query intent. In other words, the query method based on selection range ratio anchoring can significantly reduce the interference of irrelevant information by focusing on the most relevant subset of soil remediation solution contextual semantic features, thereby improving the efficiency of the target soil-remediation solution query response. In particular, in a specific example of the present application, such as Figure 4 As shown, the S4 includes: S41, determining a selection range ratio based on a set of embedded coding features of the parameterized description of the soil remediation scheme and an embedded coding feature of the parameterized description of the target soil; S42, based on the selection range ratio, performing dynamic semantic query encoding on the set of embedded coding features of the parameterized description of the target soil and the parameterized description of the soil remediation scheme to obtain the target soil-remediation scheme query response coding feature.

[0026] Specifically, the S41 determines the selection range ratio based on the set of embedded coding features of the soil remediation scheme parameterized description and the embedded coding features of the target soil parameterized description. That is, in the technical solution of the present application, first, the dynamic search anchor center of the set of embedded coding feature vectors of the soil remediation scheme parameterized description is extracted to obtain the soil remediation scheme dynamic search anchor vector. Specifically, the internal relationship score of the target soil parameterized description embedded coding vector relative to each soil remediation scheme parameterized description embedded coding vector in the set of the soil remediation scheme parameterized description embedded coding vector is first calculated to obtain a sequence of target soil-soil remediation scheme internal relationship score values; that is, by calculating the internal relationship score of the target soil parameterized description embedded coding vector relative to each soil remediation scheme parameterized description embedded coding vector in the set of the soil remediation scheme parameterized description embedded coding vector, the direct similarity and potential semantic connection between the target soil parameterized description embedded coding vector and each of the soil remediation scheme parameterized description embedded coding vectors are reflected to obtain a sequence of target soil-soil remediation scheme internal relationship score values. For example, certain soil properties (such as heavy metal concentrations) have a significant impact on the effectiveness of specific remediation technologies, but this relationship is often difficult to capture through simple distance metrics, and the calculation of internal relationship scores is intended to more comprehensively consider these complex interactions. Next, the soil remediation scheme parameterized description embedded coding vector corresponding to the maximum value in the sequence of target soil-soil remediation scheme internal relationship score values ​​is used as the soil remediation scheme dynamic search anchor vector. In this way, it is ensured that subsequent matching operations can be performed on the most representative basis, thereby improving the matching accuracy. In a specific example, the dynamic search anchor center of the set of embedded coding feature vectors of the soil remediation scheme parameterized description is extracted using the following dynamic search anchor center extraction formula to obtain the soil remediation scheme dynamic search anchor vector; wherein, the dynamic search anchor center extraction formula is: in, is a set of embedded coding vectors describing the parameterized description of the soil remediation scheme, They are the first, second, and third in the set of embedded coding vectors for the parameterized description of the soil remediation scheme. and The soil remediation scheme parameterized description is embedded in the coding vector, is the target soil parameterization description embedding encoding vector, and They are the parameterized description weight matrix of soil remediation scheme and the parameterized description weight matrix of target soil, for function, is the transposed vector of the internal relationship score reference vector between target soil and remediation scheme, is the first in the sequence of the internal relationship scores of the target soil and soil remediation scheme. The internal relationship score of the target soil and soil remediation solution, Indicates the embedded coding vector of the parameterized description of the soil remediation scheme corresponding to the maximum value in the sequence of the internal relationship score values ​​of the target soil-soil remediation scheme, Dynamic search for anchor vectors for soil remediation solutions.

[0027] Then, based on the characteristic distribution characteristics of the soil remediation scheme dynamic search anchor vector, the selection range ratio is determined, wherein the vector of the starting position of the selection range ratio is the soil remediation scheme dynamic search anchor vector. In a specific example, based on the characteristic distribution characteristics of the soil remediation scheme dynamic search anchor vector, the selection range ratio is determined by the following selection range ratio determination formula; wherein the selection range ratio determination formula is: in, and are the mean and variance of the dynamic search anchor quantity of the soil remediation scheme, represents the maximum value function, is a very small positive number, used to prevent the denominator from being 0. Indicates rounding up operation. To select the range ratio.

[0028] Specifically, the S42, based on the selection range ratio, performs dynamic semantic query encoding on the set of embedded coding features of the target soil parameterized description and the soil remediation scheme parameterized description to obtain the target soil-remediation scheme query response coding feature. That is, in the technical solution of the present application, first, based on the selection range ratio, a set of embedded coding vectors of the parameterized description of the dynamic search optimization soil remediation scheme is determined, wherein each dynamic search optimization soil remediation scheme parameterized description embedded coding vector in the set of embedded coding vectors of the parameterized description of the dynamic search optimization soil remediation scheme is a parameterized description embedded coding feature vector of the soil remediation scheme in the selection range ratio window. In the technical solution of the present application, each soil remediation scheme parameterized description embedded coding vector in the selection range ratio window is defined as a dynamic search optimization soil remediation scheme parameterized description embedded coding vector, as a potential matching candidate, constituting a set of dynamic search optimization soil remediation scheme parameterized description embedded coding vectors for performing more refined matching analysis. That is, by combining statistical analysis methods to focus on the remediation scheme features that have a high correlation with the target soil, while excluding irrelevant parts, the search space is narrowed and efficiency is improved. In this way, it is ensured that the selected dynamically searched optimized soil remediation scheme parameterized description embedded coding vector can retain the original information to the greatest extent and reduce noise interference. Then, the set of the target soil parameterized description embedded coding vector and the dynamically searched optimized soil remediation scheme parameterized description embedded coding vector is input into the dynamic semantic search encoder to obtain the target soil-remediation scheme query response coding vector as the target soil-remediation scheme query response coding feature. That is, by inputting the target soil parameterized description embedded coding vector and the set of the dynamically searched optimized soil remediation scheme parameterized description embedded coding vector into the dynamic semantic search encoder, the semantic interaction relationship between the parameters is captured, and the generated target soil-remediation scheme query response coding vector has a high degree of abstract generalization ability, so that the model can better understand the relationship between soil characteristics and remediation needs, thereby providing higher quality recommendation results. In a specific example, based on the selection range ratio, the set of the target soil parameterized description and the soil remediation scheme parameterized description embedded coding features are dynamically semantically queried and encoded using the following dynamic semantic query coding formula to obtain the target soil-remediation scheme query response coding feature; wherein, the dynamic semantic query coding formula is: in, is the vector of the starting position of the selected range ratio, that is, , is the vector of the end position of the selected range ratio, For each dynamic search optimization soil remediation scheme parameterized description embedded coding vector in the set of dynamic search optimization soil remediation scheme parameterized description embedded coding vectors, A set of embedding coding vectors describing parameterization for the dynamic search optimization soil remediation solution, is the norm of the vector, for and Dynamic semantic search cosine similarity between A query response encoding vector for the target soil-remediation solution.

[0029] In particular, the S5 recommends the remediation scheme of the target soil based on the target soil-remediation scheme query response coding features. In the technical solution of the present application, the target soil-remediation scheme query response coding vector is input into a remediation scheme recommender based on a classifier to obtain a recommendation result, and the recommendation result is the remediation scheme of the target soil. As a machine learning model, the classifier can learn the complex nonlinear relationship between the target soil and the remediation scheme according to the target soil-remediation scheme query response coding vector, and accordingly map the target soil-remediation scheme query response coding vector to a preset remediation scheme label, thereby realizing intelligent recommendation of remediation schemes. That is, the classifier can identify the most suitable remediation scheme for a specific type of soil based on the target soil-remediation scheme query response coding vector. In this way, not only the accuracy of the remediation scheme recommendation is improved, but also the scientificity and rationality of the decision-making are enhanced. Specifically, in the process of inputting the target soil-remediation solution query response encoding vector into the classifier-based remediation solution recommender to obtain the recommendation result, first, the target soil-remediation solution query response encoding vector is fully connected encoded using multiple fully connected layers of the classifier to obtain an encoded classification feature vector; then, the encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the recommendation result.

[0030] Preferably, when the set of the target soil parameterized description embedded coding vector and the soil remediation scheme parameterized description embedded coding vector respectively represent the embedded coding features of the target soil parameters and the low-dimensional embedded coding features of the soil remediation scheme parameterized description, when performing feature dynamic semantic search encoding based on selection range ratio anchoring, the complexity of the selection range ratio anchoring mechanism caused by different encoding modes under the semantic differences of parameter sources will lead to insufficient long-distance dynamic semantic search coding representation of the target soil-remediation scheme query response coding vector, thereby reducing the expression effect of the target soil-remediation scheme query response coding vector and affecting the accuracy of the recommendation results obtained by its input into the classifier-based remediation scheme recommender.

[0031] Therefore, in one example, when the target soil-remediation solution query response encoding vector is input into the classifier-based remediation solution recommender, the target soil-remediation solution query response encoding vector is optimized, and the optimization includes the following steps: Arrange the characteristic values ​​of the target soil-remediation solution query response coding vector in ascending order to form a target soil-remediation solution query response sequence coding vector; In response to the target soil-remediation solution query response sequence encoding vector The eigenvalue and The absolute value of the difference between the eigenvalues ​​is less than or equal to the distance difference hyperparameter , calculate the The eigenvalues ​​are similar to the The weighted sum between the eigenvalues ​​is the optimized Eigenvalue ; In response to the target soil-remediation solution query response sequence encoding vector The eigenvalue and The absolute value of the difference between the eigenvalues ​​is greater than the distance difference hyperparameter : Calculate the square root of the sum of squares of all eigenvalues ​​of the target soil-remediation solution query response encoding vector ; The square root Multiply by 2 and then divide by the square of the length of the target soil-remediation solution query response encoding vector to obtain the target soil-remediation solution query response spatial primitive value ; The target soil-remediation solution query response space primitive value is multiplied by the first After the eigenvalue, calculate the product with the first The weighted reduction between the eigenvalues ​​is the optimized Eigenvalue ; based on , the first Eigenvalue To obtain the optimized target soil-remediation scheme query response encoding vector.

[0032] In this way, in order to address the problem of insufficient global semantic search representation capability of the feature set of the target soil-remediation solution query response coding vector under a predetermined eigenvalue sequential distribution due to a long distance exceeding a predetermined local distribution interval threshold, a high-dimensional feature space primitive representation based on self-inner product fusion of the target soil-remediation solution query response coding vector is used to capture the complex structure of the global network interaction of its eigenvalues, thereby reconstructing the dynamic semantic search relationship between the eigenvalues ​​of the target soil-remediation solution query response coding vector by simulating scale-based high-dimensional feature space potential primitives, so as to achieve coding reconstruction of the real sequence distribution behavior of the target soil-remediation solution query response coding vector under long distances, improve the coding expression effect of the target soil-remediation solution query response coding vector, and improve the accuracy of the recommendation results obtained by inputting the classifier-based remediation solution recommender.

[0033] In summary, the soil remediation technology evaluation method according to the embodiment of the present application is explained, which extracts the key parameters of the target soil and obtains a series of parametrically described remediation schemes from a pre-established soil remediation scheme database, and uses embedded coding technology and vector coding technology based on deep learning to process these data to mine the semantic information of the parametric description of the target soil and the parametric description of the soil remediation scheme. Furthermore, by performing semantic queries on the target soil and each remediation scheme, a deep-level match between the target soil characteristics and the remediation scheme is achieved, and the remediation scheme for the target soil is intelligently recommended based on the matching results. In this way, the soil and remediation scheme can be accurately matched, the accuracy of soil remediation scheme screening can be improved, and the recommended results can be more in line with actual soil needs, thereby providing a direct and scientific action guide for soil remediation projects and ensuring the efficient implementation of remediation work.

[0034] Furthermore, a soil remediation technology evaluation system is also provided.

[0035] Figure 5 FIG. 1 is a block diagram of a soil remediation technology evaluation system according to an embodiment of the present application. Figure 5As shown, according to the soil remediation technology evaluation system 300 of the embodiment of the present application, it includes: a target soil parameter extraction module 310, which is used to extract the parameters of the target soil; a soil remediation scheme parameterized description extraction module 320, which is used to extract a set of soil remediation scheme parameterized descriptions from a soil remediation scheme database; an embedding coding module 330, which is used to embed the parameters of the target soil and the set of the soil remediation scheme parameterized descriptions to obtain a set of target soil parameterized description embedded coding features and soil remediation scheme parameterized description embedded coding features; a feature dynamic semantic query coding module 340, which is used to perform feature dynamic semantic query coding based on selection range ratio anchoring on the set of target soil parameterized description embedded coding features and the soil remediation scheme parameterized description embedded coding features to obtain target soil-remediation scheme query response coding features; a remediation scheme recommendation module 350, which is used to recommend a remediation scheme for the target soil based on the target soil-remediation scheme query response coding features.

[0036] As described above, the soil remediation technology assessment system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a soil remediation technology assessment algorithm. In a possible implementation, the soil remediation technology assessment system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the soil remediation technology assessment system 300 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the soil remediation technology assessment system 300 can also be one of the many hardware modules of the wireless terminal.

[0037] Alternatively, in another example, the soil remediation technology assessment system 300 and the wireless terminal may also be separate devices, and the soil remediation technology assessment system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0038] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A soil remediation technology assessment method, characterized in that: include: Extract the parameters of the target soil; extracting a set of parameterized descriptions of soil remediation schemes from a soil remediation scheme database; Embedding the set of the target soil parameters and the soil remediation scheme parameterized description to obtain a set of target soil parameterized description embedded coding features and a set of soil remediation scheme parameterized description embedded coding features; Performing feature dynamic semantic query encoding based on selection range ratio anchoring on the set of the target soil parameterized description embedded coding features and the soil remediation solution parameterized description embedded coding features to obtain a target soil-remediation solution query response coding feature; Based on the target soil-remediation solution query response coding features, a remediation solution for the target soil is recommended.

2. The soil remediation technology evaluation method according to claim 1, characterized in that: The parameters of the target soil and the set of the parameterized description of the soil remediation scheme are embedded and encoded to obtain a set of embedded coding features of the parameterized description of the target soil and a set of embedded coding features of the parameterized description of the soil remediation scheme, including: Performing embedding coding based on the Word2Vec model on the parameters of the target soil to obtain an embedding coding vector of the target soil parameterized description as the embedding coding feature of the target soil parameterized description; Each soil remediation scheme parameterized description in the set of soil remediation scheme parameterized descriptions is vectorized and encoded to obtain a set of soil remediation scheme parameterized description embedded coding vectors as a set of soil remediation scheme parameterized description embedded coding features.

3. The soil remediation technology evaluation method according to claim 2, characterized in that: The set of the target soil parameterized description embedded coding features and the soil remediation scheme parameterized description embedded coding features is subjected to feature dynamic semantic query coding based on the selection range ratio anchoring to obtain the target soil-remediation scheme query response coding features, including: Determining a selection range ratio based on the set of embedded coding features of the parameterized description of the soil remediation solution and the embedded coding features of the parameterized description of the target soil; Based on the selection range ratio, dynamic semantic query encoding is performed on the set of embedded coding features of the target soil parameterized description and the soil remediation solution parameterized description to obtain the target soil-remediation solution query response coding feature.

4. The soil remediation technology evaluation method according to claim 3, characterized in that: Determining a selection range ratio based on a set of embedded coding features of the parameterized description of the soil remediation scheme and embedded coding features of the parameterized description of the target soil includes: Extracting a dynamic search anchor center of a set of embedded encoding feature vectors of the soil remediation scheme parameterized description to obtain a soil remediation scheme dynamic search anchor vector; Based on the characteristic distribution characteristics of the soil remediation scheme dynamic search anchor vector, a selection range ratio is determined, wherein the vector of the starting position of the selection range ratio is the soil remediation scheme dynamic search anchor vector.

5. The soil remediation technology evaluation method according to claim 4, characterized in that: Extracting the dynamic search anchor center of the set of embedded encoding feature vectors of the soil remediation scheme parameterized description to obtain the soil remediation scheme dynamic search anchor vector, including: Calculating the internal relationship score value of the target soil parameterized description embedded coding vector relative to each soil remediation scheme parameterized description embedded coding vector in the set of soil remediation scheme parameterized description embedded coding vectors to obtain a sequence of target soil-soil remediation scheme internal relationship score values; The soil remediation scheme parameterized description embedding coding vector corresponding to the maximum value in the sequence of the target soil-soil remediation scheme internal relationship score values ​​is used as the soil remediation scheme dynamic search anchor vector.

6. The soil remediation technology evaluation method according to claim 5, characterized in that: Based on the selection range ratio, a set of embedded coding features of the target soil parameterized description and the soil remediation solution parameterized description is dynamically queried and coded to obtain the target soil-remediation solution query response coding feature, including: Based on the selection range ratio, determining a set of embedded coding vectors for parameterized description of dynamic search optimization soil remediation schemes, wherein each embedded coding vector for parameterized description of dynamic search optimization soil remediation schemes in the set of embedded coding vectors for parameterized description of dynamic search optimization soil remediation schemes is an embedded coding feature vector for parameterized description of soil remediation schemes in the selection range ratio window; The set of the target soil parameterized description embedded coding vector and the dynamically searched optimized soil remediation scheme parameterized description embedded coding vector is input into a dynamic semantic search encoder to obtain a target soil-remediation scheme query response coding vector as the target soil-remediation scheme query response coding feature.

7. The soil remediation technology evaluation method according to claim 6, characterized in that: Based on the target soil-remediation solution query response coding features, a remediation solution for the target soil is recommended, including: The target soil-remediation solution query response encoding vector is input into a classifier-based remediation solution recommender to obtain a recommendation result, which is a remediation solution for the target soil.

8. The soil remediation technology evaluation method according to claim 7, characterized in that: The target soil-remediation solution query response encoding vector is input into a classifier-based remediation solution recommender to obtain a recommendation result, which is a remediation solution for the target soil, including: Using multiple fully connected layers of the classifier to perform fully connected encoding on the target soil-remediation solution query response encoding vector to obtain an encoded classification feature vector; The encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the recommendation result.

9. A soil remediation technology evaluation system, characterized in that: include: A target soil parameter extraction module is used to extract the parameters of the target soil; A soil remediation scheme parameterized description extraction module is used to extract a set of soil remediation scheme parameterized descriptions from a soil remediation scheme database; An embedded coding module, used for embedding the set of the target soil parameters and the soil remediation scheme parameterized description to obtain a set of embedded coding features of the target soil parameterized description and embedded coding features of the soil remediation scheme parameterized description; A feature dynamic semantic query encoding module is used to perform feature dynamic semantic query encoding based on selection range ratio anchoring on the set of the target soil parameterized description embedded encoding features and the soil remediation solution parameterized description embedded encoding features to obtain a target soil-remediation solution query response encoding feature; The restoration scheme recommendation module is used to recommend a restoration scheme for the target soil based on the target soil-restoration scheme query response coding feature.

Images