In-situ remediation method for heavy metals in river sediment

Abstract

The application provides a river sediment heavy metal in-situ solidification remediation method, comprising the following steps: S1, detecting the content and physicochemical properties of the river sediment heavy metal; S2, designing a remediation material formula; S3, preparing and putting the remediation material; S4, mixing and stirring the sediment and the remediation material; S5, monitoring and evaluating the remediation effect; and S6, feeding back and adjusting the remediation effect. Through the fusion of multiple linear regression, curved surface fitting, self-organizing mapping and hybrid density network algorithm, the content and physicochemical property data of the sediment heavy metal are fully utilized, the remediation material formula can be accurately designed, compared with the traditional single algorithm or empirical formula design, the remediation material ratio can be more suitable for the actual pollution situation, and the utilization rate of the remediation material is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sludge treatment, and in particular to a river sediment heavy metal in-situ solidification remediation method. BACKGROUND

[0002] With the acceleration of industrialization and urbanization, a large amount of industrial wastewater containing heavy metals, domestic sewage, pesticides and fertilizers in agricultural production, and waste produced by mining and metal smelting are directly discharged into rivers without effective treatment, which makes river sediment become an important accumulation place of heavy metals. Cadmium (Cd), lead (Pb), mercury (Hg), chromium (Cr), arsenic (As) and other heavy metals continue to accumulate in the sediment, not only causing harm to benthic organisms and affecting the balance of river ecosystems, but also threatening human health through the food chain.

[0003] At present, the river sediment heavy metal remediation methods mainly include physical remediation, chemical remediation and biological remediation. Physical remediation such as dredging of sediment can directly remove contaminated sediment, but the engineering quantity is large, the cost is high, and it is easy to cause secondary damage to the river ecological environment; although some methods in chemical remediation can quickly reduce the activity of heavy metals, the chemical agents used may introduce new pollution; biological remediation has a long cycle and is greatly limited by environmental conditions. The existing remediation methods have certain defects in precision, efficiency and environmental friendliness, and it is difficult to meet the actual remediation demand, and a more scientific and efficient river sediment heavy metal in-situ solidification remediation method is urgently needed. SUMMARY

[0004] The present application provides a river sediment heavy metal in-situ solidification remediation method to solve one or more of the above problems.

[0005] To achieve the above purpose, the present application adopts the following technical solutions:

[0006] A river sediment heavy metal in-situ solidification remediation method, comprising the following steps:

[0007] S1. Detecting the heavy metal content and physicochemical properties of river sediment; collecting sediment samples, digesting the sediment samples, determining the heavy metal content in the sediment, and obtaining heavy metal content data; at the same time, determining the pH value, particle size distribution, and organic matter content of the sediment, and obtaining physicochemical property data;

[0008] S2. Repair material formula design; according to the heavy metal content data and the physicochemical property data obtained in S1, combining the solidification characteristics of different heavy metals, the preliminary ratio of the repair material is calculated by a multiple linear regression algorithm, and the intermediate ratio is obtained by a surface fitting algorithm optimization, the feature vectors containing the heavy metal content data, the physicochemical property data and the intermediate ratio are reduced and feature extracted by a self-organizing mapping algorithm to obtain a two-dimensional feature vector, and finally the two-dimensional feature vector is processed by a hybrid density network algorithm to obtain the final ratio of the repair material;

[0009] S3. Preparation and delivery of repair material; the total delivery amount is calculated according to the final ratio of the repair material, the repair material is prepared, and the prepared repair material is uniformly delivered into the river sediment;

[0010] S4. Mixing and stirring of sediment and repair material; the sediment after delivery of the repair material is stirred by a mechanical stirring device, the stirring speed and stirring time are controlled to make the repair material and the sediment uniformly mixed, and the uniformity of the distribution of the repair material in the sediment is detected after stirring is completed;

[0011] S5. Monitoring and evaluation of repair effect; sediment samples after repair are collected at a preset time interval after repair, the heavy metal content in the sediment is determined, the solidification rate of the heavy metal is calculated, and the physicochemical properties of the sediment are detected, the changes of the physicochemical properties of the sediment before and after repair are compared, and the repair effect evaluation result is obtained;

[0012] S6. Feedback and adjustment of repair effect; according to the repair effect evaluation result, if the heavy metal solidification rate does not reach the expected target, the reasons are analyzed and the repair material ratio or stirring parameters are adjusted, the mixing and stirring and the repair effect monitoring and evaluation are carried out again until the heavy metal solidification rate reaches the standard.

[0013] In the specification, the sediment samples in S1 are collected, specifically: grids are divided at the upstream, midstream and downstream of the river, as well as the center of the river channel and near the shore, and sediment samples with a depth of 0-50 cm are collected at the center of each grid, and the samples at different depths of the same sampling point are mixed uniformly.

[0014] In the specification, the heavy metal content in the sediment in S1 is determined, specifically: the contents of cadmium, lead, mercury, chromium and arsenic are determined by atomic absorption spectrometry and inductively coupled plasma mass spectrometry; the pH value is determined by a pH meter, the particle size distribution is determined by a laser particle size analyzer, and the organic matter content is determined by a loss on ignition method.

[0015] In the specification, in S2, the solidification characteristics of different heavy metals are as follows: for cadmium, a phosphate material is selected and mixed with cement as the repair material; for lead, a combination of iron oxide and lime is used as the repair material; for mercury, a sulfide is used in combination with bentonite as the repair material; for chromium, silicate cement and fly ash are selected as the repair material; and for arsenic, iron-containing materials and clay minerals are used as the repair material.

[0016] In the specification, in S2, the preliminary ratio of the repair material is calculated by a multiple linear regression algorithm, specifically: a multiple linear regression model is established, the heavy metal content data and the physicochemical property data are used as input variables, and the ratio of the repair material is used as an output variable, the model parameters are estimated by the least squares method, so that the error between the predicted repair material ratio and the actual effective ratio is minimized.

[0017] In the specification, in S2, the intermediate ratio is obtained by a surface fitting algorithm, specifically: a cubic polynomial surface fitting model is used, the square term of the heavy metal content, the interaction term and the coupling term of the physicochemical property are introduced, the preliminary ratio is nonlinearly optimized, and the nonlinear influence of the interaction between the heavy metals and the characteristics of the sediment is captured.

[0018] In the specification, in S2, the dimensionality reduction and feature extraction are performed by using a self-organizing mapping algorithm, specifically: the heavy metal content data, the physicochemical property data and the intermediate ratio are combined to form a high-dimensional feature vector, the high-dimensional feature vector is mapped to a two-dimensional grid through a self-organizing mapping network, and a two-dimensional coordinate reflecting the matching characteristics of the sediment and the repair material is extracted as a two-dimensional feature vector.

[0019] In the specification, in S3, when preparing the repair material, each repair material is weighed according to the ratio and then uniformly mixed by mechanical stirring; when the repair material is put in, special equipment such as a mud pump and a spraying device is used to uniformly put the repair material into the contaminated layer of the river sediment.

[0020] In the specification, in S4, the stirring speed and stirring time are controlled, specifically: the stirring speed is controlled at 50-150 r / min, the stirring time is determined according to the thickness of the sediment and the stirring speed coefficient of the stirring equipment, and the stirring speed coefficient is determined in advance through experiments.

[0021] In the specification, in S5, sediment samples are collected at time intervals of 1 day, 7 days and 30 days after the repair is completed, the heavy metal content after the repair is determined, the solidification rate is calculated, the formula is solidification rate=(initial heavy metal content-repaired heavy metal content) / initial heavy metal content*100%, and the changes in physicochemical properties such as pH value, particle size distribution and organic matter content of the sediment before and after the repair are compared.

[0022] In summary, the present application has at least the following beneficial effects:

[0023] 1. Precise formula design: By integrating multiple linear regression, surface fitting, self-organizing mapping and hybrid density network algorithms, the data of heavy metal content and physicochemical properties of sediment are fully utilized to accurately design the formula of remediation materials. Compared to traditional single algorithm or empirical formula design, this method can make the ratio of remediation materials more suitable for actual pollution conditions, improve the utilization rate of remediation materials, and avoid material waste.

[0024] 2. High-efficiency repair: From sediment detection, remediation material preparation and delivery, to mixing and stirring, effect monitoring and evaluation, and feedback adjustment, a complete and efficient repair process is formed. Precise control of parameters such as stirring speed and time ensures that remediation materials and sediment are fully mixed, accelerates heavy metal solidification reaction, and shortens the repair period. The mechanism of multiple monitoring and feedback adjustment can timely find and solve problems in the repair process, ensuring efficient progress of the repair work.

[0025] 3. Good repair effect: The repair effect is evaluated by calculating the heavy metal solidification rate and other indicators, combined with the precise formula design of multiple algorithms and sufficient mixing reaction, which can significantly reduce the content and activity of heavy metals in sediment, effectively reduce the risk of migration and transformation of heavy metals, and effectively control the heavy metal pollution of sediment, restore the ecological environment of river.

[0026] 4. Reduce environmental impact: In-situ solidification repair avoids the ecological damage caused by large-scale dredging of sediment, and the remediation materials are mostly environmentally friendly, reducing the risk of secondary pollution. While achieving pollution control, the original ecological system of the river is protected to the greatest extent.

[0027] 5. Data-driven optimization: The data between each step is closely related, and the output data of the previous step is used as the input of the subsequent step, realizing dynamic optimization of the repair process. From the data-driven formula design based on sediment detection data to the feedback adjustment of the formula and repair process based on repair effect data, a data closed loop is formed, continuously improving the adaptability of the repair scheme and the repair effect. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0029] Figure 1 The flowchart of the river sediment heavy metal in-situ solidification repair method involved in the present application.

[0030] Figure 2 The flowchart of the formula design of the remediation material involved in the present application.

[0031] Figure 3 A flowchart of a process for preparing and delivering the repair material involved in the present application.

[0032] Figure 4 A flowchart of a process for monitoring the repair effect and adjusting the feedback involved in the present application. DETAILED DESCRIPTION

[0033] In the following, only certain exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the embodiments of the present application. Therefore, the drawings and the description are considered to be exemplary in nature rather than limiting.

[0034] The following disclosure provides many different embodiments, or examples, for implementing different structures of the embodiments of the present application. For the purpose of simplifying the disclosure of the embodiments of the present application, the components and settings of specific examples are described in the following. Of course, they are only examples, and the purpose is not to limit the embodiments of the present application. In addition, the embodiments of the present application can repeatedly refer to numerals and / or reference letters in different examples, and such repetition is for the purpose of simplification and clarity, which itself does not indicate the relationship between the various embodiments and / or settings discussed.

[0035] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0036] As Figure 1 shown, the present embodiment provides a method for in-situ solidification remediation of heavy metals in river sediment, comprising the following steps:

[0037] S1. Detecting the content of heavy metals and the physicochemical properties of river sediment; collecting sediment samples, digesting the sediment samples, determining the content of heavy metals in the sediment, and obtaining the data of heavy metal content; at the same time, determining the pH value, particle size distribution, and organic matter content of the sediment, and obtaining the physicochemical property data;

[0038] S2. Designing the formula of the repair material; according to the data of heavy metal content and physicochemical properties obtained in S1, combining the solidification characteristics of different heavy metals, preliminarily calculating the preliminary ratio of the repair material through the multivariate linear regression algorithm, and then optimizing the intermediate ratio through the surface fitting algorithm, using the self-organizing mapping algorithm to reduce the dimension and extract the features of the feature vector containing the data of heavy metal content, physicochemical properties, and intermediate ratio, obtaining the two-dimensional feature vector, and finally processing the two-dimensional feature vector through the hybrid density network algorithm to obtain the final ratio of the repair material;

[0039] S3. Preparing and delivering the repair material; calculating the total delivery amount according to the final ratio of the repair material, preparing the repair material, and uniformly delivering the prepared repair material into the river sediment;

[0040] S4. Mixing and stirring the sediment and the remediation material; a mechanical stirring device is used to stir the sediment after the remediation material is added, the stirring speed and stirring time are controlled to make the remediation material and the sediment uniformly mixed, and the uniformity of the distribution of the remediation material in the sediment is detected after the stirring is completed;

[0041] S5. Monitoring and evaluation of the remediation effect; sediment samples after the remediation are collected at preset time intervals after the remediation is completed, the heavy metal content in the sediment is determined, the solidification rate of the heavy metal is calculated, the physicochemical properties of the sediment are detected, the changes in the physicochemical properties of the sediment before and after the remediation are compared, and the evaluation result of the remediation effect is obtained;

[0042] S6. Feedback and adjustment of the remediation effect; according to the evaluation result of the remediation effect, if the solidification rate of the heavy metal does not reach the expected target, the reasons are analyzed and the proportioning of the remediation material or the stirring parameters are adjusted, the mixing and stirring and the monitoring and evaluation of the remediation effect are re-performed until the solidification rate of the heavy metal reaches the standard.

[0043] In some embodiments, the sediment samples are collected in S1, specifically: grids are divided at upstream, midstream and downstream of the river, and at the center of the river channel and near the shore, and sediment samples with a depth of 0-50 cm are collected at the center of each grid, and the samples at different depths of the same sampling point are uniformly mixed.

[0044] In some embodiments, the heavy metal content in the sediment is determined in S1, specifically: the contents of cadmium, lead, mercury, chromium and arsenic are determined by atomic absorption spectrometry and inductively coupled plasma mass spectrometry; the physicochemical properties of the sediment are determined, specifically: the pH value is determined by a pH meter, the particle size distribution is determined by a laser particle size analyzer, and the organic matter content is determined by the loss on ignition method.

[0045] In some embodiments, in S2, the solidification characteristics of different heavy metals are combined, specifically: for cadmium, phosphate materials are mixed with cement as the remediation material; for lead, a combination of iron oxide and lime is used as the remediation material; for mercury, sulfide is used in combination with bentonite as the remediation material; for chromium, silicate cement and fly ash are selected as the remediation material; and for arsenic, iron-containing materials and clay minerals are used as the remediation material.

[0046] In some embodiments, the preliminary proportioning of the remediation material is calculated by a multiple linear regression algorithm in S2, specifically: a multiple linear regression model is established, the heavy metal content data and the physicochemical property data are used as input variables, the proportioning of the remediation material is used as an output variable, the model parameters are estimated by the least squares method, and the error between the predicted proportioning of the remediation material and the actual effective proportioning is minimized.

[0047] In some embodiments, in S2, the intermediate proportioning is optimized by a curved surface fitting algorithm, specifically: a cubic polynomial curved surface fitting model is adopted, a square term of heavy metal content, an interaction term and a coupling term of physicochemical properties are introduced, and the preliminary proportioning is nonlinearly optimized to capture the nonlinear influence of the interaction between heavy metals and the characteristics of the sediment.

[0048] In some embodiments, in S2, dimension reduction and feature extraction are performed by using a self-organizing mapping algorithm, specifically: the heavy metal content data, the physicochemical property data and the intermediate proportioning are combined to form a high-dimensional feature vector, the high-dimensional feature vector is mapped to a two-dimensional grid through a self-organizing mapping network, and the two-dimensional coordinates reflecting the matching characteristics of the sediment and the repair material are extracted as a two-dimensional feature vector.

[0049] In some embodiments, in S3, when the repair material is prepared, each repair material is weighed according to the proportioning and then uniformly mixed by mechanical stirring; and when the repair material is put, special equipment such as a slurry pump and a spraying device is used to uniformly put the repair material into the pollution layer of the river sediment.

[0050] In some embodiments, in S4, the stirring speed and stirring time are controlled, specifically: the stirring speed is controlled at 50-150 r / min, the stirring time is determined according to the sediment thickness and the stirring speed coefficient of the stirring equipment, and the stirring speed coefficient is determined through experiments in advance.

[0051] In some embodiments, in S5, sediment samples are collected at time intervals of 1 day, 7 days and 30 days after the repair is completed, the heavy metal content after the repair is determined, the solidification rate is calculated, the formula is solidification rate=(initial heavy metal content-repaired heavy metal content) / initial heavy metal content*100%, and the changes in physicochemical properties such as pH value, particle size distribution and organic matter content of the sediment before and after the repair are compared.

[0052] The technical concept of the application is as follows:

[0053] S1. Detection of heavy metal content and physicochemical properties of river sediment:

[0054] A columnar sampler is used to sample in different functional areas (upstream water source protection area, midstream urban influence area and downstream estuary mixing area) and different geomorphic units (main channel of river, floodplain and nearshore shoal) of the river according to a 50m*50m grid point method, 0-20cm and 20-50cm two layers of sediment are collected vertically at each sampling point, and each layer is collected 3 times to mix as one sample.

[0055] In laboratory detection, samples are treated by a nitric acid-hydrofluoric acid-perchloric acid complete digestion system, the heavy metal content such as cadmium (Cd) and lead (Pb) is determined by an atomic absorption spectrometer (AAS), and trace elements such as mercury (Hg) and arsenic (As) are determined by an inductively coupled plasma mass spectrometer (ICP-MS).

[0056] S1. Basic physicochemical properties of the sediment are detected synchronously: pH value is measured by glass electrode method, particle size distribution is measured by laser particle size analyzer (divided into three levels of clay <2 μm, silt 2-50 μm, sand >50 μm), and organic matter content is measured by muffle furnace ignition method (ignited at 550°C for 4h). The content of n heavy metals is denoted as C i (i = 1, 2, …, n, unit: mg / kg), and the physicochemical property index of zm is denoted as P j (j = 1, 2, …, zm, including pH value, clay content, organic matter content, etc.).

[0057] S2. Formulation design of repair materials:

[0058] Based on the detection data C i and P j , a multi-algorithm fusion formulation design model is constructed, and the chemical forms (water-soluble state, exchangeable state, carbonate-bound state, iron-manganese oxidation state, organic-bound state, residual state) and sediment mineral composition (quartz, feldspar, clay mineral content) of different heavy metals such as cadmium, lead, mercury, chromium, and arsenic are determined in four steps as shown in Figure 2 to determine the proportion of repair materials.

[0059] 1. Multivariate linear regression preliminary modeling:

[0060] A linear relationship model is established to describe the preliminary correlation between the proportion of repair materials and the characteristics of the sediment, and the expression is:

[0061]

[0062] wherein, is the preliminary mass proportion (%) of the kth repair material (k = 1, 2, …, l, l is the number of material types, such as hydroxyapatite, hematite, sodium sulfide, etc.); a k0 is a constant term, a ki , a kj are the regression coefficients of heavy metal content and physicochemical properties, respectively, and are optimized by least squares method, so that the root mean square error (RMSE) between the predicted value and the historical effective formulation is less than 5%.

[0063] 2. Nonlinear optimization of surface fitting:

[0064] Considering the nonlinear effects of heavy metal interaction and sediment characteristics, the preliminary proportion is corrected by using a cubic polynomial surface fitting:

[0065]

[0066] wherein, b uvw is the surface fitting coefficient, and the subscripts u, v, w correspond to Ci P j The polynomial degree of (u, v, w ∈ {0, 1, 2, 3}); by minimizing the fitting error of historical samples. Sure, This represents the actual effective proportion of the k-th type of repair material in the s-th historical sample (s = 1, 2, ..., N), where N is the total number of samples. To optimize the intermediate ratio; introduce a heavy metal concentration square term (e.g.) ), interactive items (such as C) Cd ·C Pb This corresponds to the case where v=1 and w=1. The coupling terms of physicochemical properties (such as the product of pH value and clay content) are used to capture the curing reaction mechanism under complex environments.

[0067] 3. Self-organizing map feature dimensionality reduction:

[0068] Build containing C i P j , (n+zm+l) dimensional eigenvectors Dimensionality reduction is performed using a Self-Organizing Map (SOM) network. The SOM maps high-dimensional data to a 5×5 two-dimensional grid through competitive learning, where each node di (di = 1, ..., 25) corresponds to a weight vector m. di =[m di1 m di2 , ..., m di(n+zm+l) The training objective is to minimize the Euclidean distance between the input vector and the best-matching unit.

[0069]

[0070] Where X (s) Let d(X) represent the feature vector of the s-th sample. (s) m di ) represents the input vector X (s) With weight vector m di The Euclidean distance is used; after training, each sample is mapped to a grid node q (q represents the mapped node coordinate index, q = 1, ..., 25, numbered in coordinate order), and two-dimensional coordinates reflecting the sediment-material matching characteristics are extracted. As input features for subsequent hybrid density networks.

[0071] 4. Probabilistic prediction for hybrid density networks:

[0072] Using Hybrid Density Networks (MDNs) to process the nonlinear characteristics of SOM output Constructing a multimodal output model:

[0073]

[0074] in, Let be the conditional probability density function, representing the conditional probability density function of the feature. The following is the ratio of the repair materials: The probability density, H = 3 (preset Gaussian mixture component number), π h The weight of the h-th component (∑π) h =1), Indicates Let μ be the Gaussian distribution function of the variable. h and For the mean and variance.

[0075] By maximizing the log-likelihood function The network parameters are trained, and the final output is the optimal allocation considering uncertainty. This value serves as the basis for the material proportioning in step S3.

[0076] S3. Preparation and application of repair materials (process as follows) Figure 3 As shown):

[0077] Parameter calculation: based on the final ratio output by S2 (The mass percentage of the k-th type of remediation material, k = 1, 2, ..., l), combined with the remediation area data delineated in S1, calculate the total amount of remediation material applied. First, obtain the planar area S of the remediation area through UAV mapping. Take the arithmetic mean of the contaminated layer depth at each sampling point in S1 as h0, and calculate the sediment volume V = S × h0 (unit: m). 3 Then query or measure the bulk density ρ of the k-th material. k (Unit: kg / m³) 3 The total amount of waste is calculated using the following formula:

[0078]

[0079] Proportioning and material mixing: Precise preparation is carried out in the temporary material preparation area set up on site. According to the calculated amount of each material, the repair material is accurately weighed using an electronic balance (accuracy ±0.1g) (total mass error controlled within ±0.5%), and put into a twin-shaft forced mixer (speed 80r / min) for 10min until the material is evenly dispersed (no obvious particle agglomeration).

[0080] Equipment deployment and spatial positioning: After preparation, a plunger-type mud pump (flow rate 100m³) is used. 3 / h) The remediation material slurry (solid content 30%–40%) is injected into the bottom sediment through a 100mm diameter delivery pipe. Before delivery, delivery holes (150mm diameter, penetrating 0–50cm of the contaminated layer) are arranged every 10m along the riverbed using a grid positioning method. The delivery depth is controlled by a depth sensor (accuracy ±2cm). During delivery, the actual delivery volume (M) of each material is recorded in real time. k A distribution histogram was plotted to provide baseline data for the analysis of the mixing uniformity of S4.

[0081] S4. Mix the base mud and repair materials:

[0082] Within 30 minutes after the initial sediment loading, a work vessel equipped with a low-speed submersible mixer (mixer speed 50-150 r / min) was used for zoned mixing. Before mixing, the sediment thickness h1 (unit: m, median thickness of sampling points) was measured using side-scan sonar. A mixing speed coefficient v was established through preliminary experiments. s (m / (r·min), representing the effective mixing depth per revolution per minute of the mixer, and the fitting formula v s =0.0003v + 0.01, where v is the mixer speed (r / min), calculate the mixing time:

[0083]

[0084] In actual operation, the initial setting was V1 = 100 r / min, and the conductivity was monitored in real time using a water quality sensor (stirring was stopped when the difference between adjacent sensors was <5%). During the stirring process, an underwater camera was used to check for any dead zones of accumulation. After the stirring was completed, 10 mixed samples were collected, the content of characteristic elements of the remediation material was measured, and the spatial distribution standard deviation (target <10%) was calculated. The results were used as the initial uniformity parameters for the S5 effect evaluation.

[0085] S5. Monitoring and Evaluation of Repair Results:

[0086] Sediment samples were collected 1 day, 7 days, and 30 days after remediation using the S1 grid method (densified to 25m × 25m), with three parallel samples per layer. Laboratory testing included:

[0087] 1. Heavy metal content determination: The content C after remediation was determined using the same method as S1. i ′(mg / kg).

[0088] 2. Retesting of physicochemical properties: Using the same testing methods as S1 (such as measuring pH value using the glass electrode method, measuring particle size distribution using a laser particle size analyzer, etc.), the physicochemical properties after repair were measured and recorded as P. j (j = 1, 2, ..., zm, units consistent with S1).

[0089] 3. Curing rate calculation:

[0090]

[0091] Where, η i The curing rate of the i-th heavy metal (e.g., cadmium target ≥85%, lead ≥80%) is considered stable when the curing rate fluctuates by less than 3% twice in 30 days.

[0092] 4. Microenvironment parameter detection: Establishing a matrix of physicochemical properties before and after remediation:

[0093]

[0094] Calculate the change in the index ΔP j =P j ′-P j pH value, oxidation-reduction potential (ORP), and cation exchange capacity (CEC) were measured to analyze changes in the sediment microenvironment.

[0095] Evaluation process: If η is at 30 days i If the target is met and the microenvironment parameters are normal (e.g., pH value changes within ±0.5), the repair is considered successful; otherwise, S6 feedback adjustment is triggered.

[0096] S6. Repair Results Feedback and Adjustments:

[0097] When S5 detects that the curing rate is not up to standard or the microenvironment is abnormal, adjust according to the following steps:

[0098] 1. Problem Root Cause Diagnosis:

[0099] Insufficient material reaction: XRD analysis showed that the peak value of the target cured product was less than 60% of the theoretical value, indicating that the mixing ratio was unreasonable.

[0100] Incomplete mixing: If the standard deviation of S4 uniformity test is >15% and there is a clogging zone, it is determined that the mixing is insufficient.

[0101] 2. Proportioning adjustment strategy (for materials with insufficient reaction):

[0102] The measured C of S5 i ′、P j ′ and S1 of C i P j The new input vector is used to retrain the hybrid density network of S2, and the corrected ratio is output. Calculation of supplementary investment:

[0103]

[0104] High-pressure jet grouting equipment was used to apply the material to key areas in a targeted manner, ensuring that the material penetrated into the pores of the bottom sediment.

[0105] 3. Stirring optimization strategy (for uneven mixing):

[0106] Using CFD simulation of the flow field, the low-velocity region (flow velocity < 0.05 m / s) was identified. The mixer speed was increased to 130 r / min, and the mixing time was extended by 20 min, employing a "zonal progressive mixing method" (alternating longitudinal and transverse mixing). After optimization, the uniformity was re-tested. If the standard deviation dropped below 10%, it was re-monitored in S5; otherwise, the speed was further adjusted to 150 r / min, and the time was extended by 30 min.

[0107] 4. Loop termination condition:

[0108] If the cure rate increase is less than 5% after two consecutive adjustments, or if the microenvironment remains abnormal, auxiliary technologies (such as electrokinetic remediation and biochar adsorption) are introduced to form a multi-technology coupled solution. Monitoring and feedback adjustments for remediation effectiveness are as follows: Figure 4 As shown.

[0109] In another embodiment, the S2 repair material formulation design can be carried out according to the following steps:

[0110] S21. Perform preliminary modeling using multiple linear regression to obtain preliminary quality ratios.

[0111] S22. Preliminary prediction using a mixed density network: The heavy metal content C in the sediment... i and physicochemical properties P j The constituent feature vectors serve as input to a hybrid density network (MDN). The output probability density function of the MDN is:

[0112]

[0113] Where H′ is the number of Gaussian mixture distributions, π′ h′ Let h′ be the weights of the Gaussian distribution, satisfying μ′ h′ and are the mean and variance of the h′-th Gaussian distribution, respectively;

[0114] By maximizing the log-likelihood function To obtain the preliminary predicted proportion of the k-th type of repair material in the s-th historical sample (s = 1, 2, ..., N, where N is the total number of samples), train the MDN parameters to obtain the actual effective proportion of the k-th type of repair material.

[0115] S23. Optimization using surface fitting: Initial proportions obtained from multiple linear regression. Preliminary prediction of sizing ratio using mixed density network and the heavy metal content C in the sediment i and physicochemical properties P jAs variables, construct the surface fitting function:

[0116]

[0117] Where U′, V′, W′, X′, and Y′ represent the highest degree of the polynomial corresponding to the dependent variable, and u′, v′, w′, x′, and y′ represent the degree of the polynomial corresponding to the dependent variable, respectively. u′v′w′x′y′ These are the surface fitting coefficients. The objective function is minimized. Adjust b using optimization algorithms (such as gradient descent). u′v′w′x′y′ The value of makes It more closely approximates the actual required ratio of repair materials. This refers to the ratio of the repair material k after surface fitting optimization.

[0118] S24. Dimensionality reduction and feature extraction using self-organizing maps:

[0119] The heavy metal content C in the sediment i Physicochemical properties P j and the optimized ratio after surface fitting Composition of feature vectors The Self-Organizing Map (SOM) algorithm is used to reduce the dimensionality and extract features from the feature vector X1′. SOM maps the high-dimensional feature vector onto a two-dimensional grid, and the mapped feature representation is as follows: (q′=1,…,f,f is the number of nodes in the two-dimensional grid). During the training of the SOM network, the weights between neurons are adjusted to achieve the objective function. Optimize;

[0120] Where X1′ (s) Let d(X1′) represent the feature vector of the s-th sample. (s) ,m′ di′ ) represents the input vector X1′(s) and the weight vector m′ of neuron di′. di′ The distance is used to iteratively update the weights to minimize E. SOM This enables feature extraction.

[0121] S25. Final prediction of mixed density network:

[0122] Feature representation obtained from self-organizing map As input to a hybrid density network, its output probability density function is: By maximizing the log-likelihood function L″= Training parameters yields the final proportion of repair material k.

[0123] The embodiments described above are for illustrative purposes only and are not intended to limit the invention. Therefore, any changes in numerical values ​​or substitutions of equivalent elements should still fall within the scope of this invention.

[0124] The above detailed description will enable those skilled in the art to understand that the present invention can indeed achieve the aforementioned objectives and has complied with the provisions of the Patent Law.

[0125] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention. The above descriptions are merely preferred embodiments of the invention and are not intended to limit the invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the invention should be included within the scope of protection of the invention.

[0126] It should be noted that the above description of the process is for illustrative purposes only and does not limit the scope of this specification. Those skilled in the art can make various modifications and changes to the process under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0127] The basic concepts have been described above. Obviously, for those skilled in the art who have read this application, the above disclosure is merely illustrative and does not constitute a limitation of this application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are suggested in this application, and therefore, such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this application.

[0128] Furthermore, this application uses specific terms to describe its embodiments. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different positions in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application can be appropriately combined.

[0129] Furthermore, those skilled in the art will understand that aspects of this application can be described and illustrated through several patentable types or situations, including any new and useful combination of processes, machines, products, or substances, or any new and useful improvements thereof. Therefore, aspects of this application can be implemented entirely in hardware, entirely in software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. All of the above hardware or software can be referred to as a “unit,” “module,” or “system.” Furthermore, aspects of this application can take the form of a computer program product embodied in one or more computer-readable media, wherein computer-readable program code is contained therein.

[0130] The computer program code required for the operation of each part of this application can be written in any one or more programming languages, including object-oriented programming languages ​​such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, and Python; general programming languages ​​such as C; Visual Basic, Fortran2103, Perl, COBOL2102, PHP, and ABAP; dynamic programming languages ​​such as Python, Ruby, and Groovy; or other programming languages. This program code can run entirely on the user's computer, or as a standalone software package on the user's computer, or partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer via any network, such as a local area network (LAN) or wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as Software as a Service (SaaS).

[0131] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this application are not intended to limit the order of the processes and methods of this application. Although some currently considered useful embodiments of the invention have been discussed in the foregoing disclosure by way of various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the substance and scope of the embodiments of this application. For example, although the implementation of the various components described above can be embodied in a hardware device, it can also be implemented as a purely software solution, such as an installation on an existing server or mobile device.

[0132] Similarly, it should be noted that, in order to simplify the description of the present application and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of the embodiments of the present application sometimes combines multiple features into a single embodiment, drawing, or description thereof. However, this approach of the present application should not be construed as reflecting an intention that the claimed subject matter requires more features than expressly recited in each claim. Rather, the subject of the invention should possess fewer features than in any single embodiment described above.

Claims

1. A method for in-situ solidification and remediation of heavy metals in riverbed sediment, characterized in that, Includes the following steps: S1. Detection of heavy metal content and physicochemical properties of river sediment; collect sediment samples, digest the sediment samples, determine the heavy metal content in the sediment, and obtain heavy metal content data; at the same time, determine the pH value, particle size distribution, organic matter content and physicochemical properties of the sediment, and obtain physicochemical property data. S2. Repair material formulation design; Based on the heavy metal content data and physicochemical property data obtained in S1, and combined with the curing characteristics of different heavy metals, a preliminary multiple linear regression model is performed to obtain the preliminary mass ratio. Preliminary prediction using a mixed density network: The heavy metal content in the sediment... and physicochemical properties The constituent feature vectors serve as the input to the Mixed Density Network (MDN); the output probability density function of the MDN is: ； in The number of mixture Gaussian distributions. For the first The weights are distributed according to a Gaussian distribution, satisfying... , and The first The mean and variance of a Gaussian distribution; By maximizing the log-likelihood function , For the first The actual effective ratio of the k-th repair material in a historical sample , Given the total number of samples, train the MDN parameters to obtain the preliminary predicted ratio of repair material k. ;Will , , and The intermediate ratio was obtained by optimizing the surface fitting algorithm. The self-organizing map algorithm was used to reduce the dimension and extract the features of the feature vector containing heavy metal content data, physicochemical property data and intermediate ratio to obtain a two-dimensional feature vector. Finally, the two-dimensional feature vector was processed by the hybrid density network algorithm to obtain the final ratio of the repair material. S3. Preparation and application of repair materials; Calculate the total amount to be applied according to the final ratio of the repair materials, prepare the repair materials, and evenly apply the prepared repair materials into the riverbed sediment; S4. Mixing and stirring the bottom mud and the repair material; Use mechanical mixing equipment to stir the bottom mud after the repair material is added, control the stirring speed and stirring time to ensure that the repair material and the bottom mud are evenly mixed, and check the uniformity of the distribution of the repair material in the bottom mud after the stirring is completed; S5. Monitoring and evaluation of the repair effect; Collect samples of the repaired sediment at preset time intervals after the repair is completed, determine the heavy metal content in the sediment, calculate the solidification rate of heavy metals, and at the same time test the physicochemical properties of the sediment. Compare the changes in the physicochemical properties of the sediment before and after the repair to obtain the evaluation results of the repair effect. S6. Feedback and adjustment of repair effect; Based on the repair effect evaluation results, if the heavy metal curing rate does not reach the expected target, analyze the reasons and adjust the ratio of repair materials or mixing parameters, and re-mix and re-evaluate the repair effect until the heavy metal curing rate reaches the target.

2. The method for in-situ solidification and remediation of heavy metals in river sediment according to claim 1, characterized in that, In S2, the intermediate ratio was optimized using a surface fitting algorithm. Specifically, a cubic polynomial surface fitting model was adopted, which introduced the square term of heavy metal content, the interaction term, and the coupling term of physicochemical properties to perform nonlinear optimization on the preliminary ratio, thereby capturing the nonlinear effects of the interaction between heavy metals and the characteristics of the sediment.

3. The method for in-situ solidification and remediation of heavy metals in riverbed sediment according to claim 2, characterized in that, In S2, the self-organizing mapping algorithm is used for dimensionality reduction and feature extraction. Specifically, the heavy metal content data, physicochemical property data and intermediate ratio are combined to form a high-dimensional feature vector. The high-dimensional feature vector is then mapped to a two-dimensional grid through the self-organizing mapping network. Two-dimensional coordinates reflecting the matching characteristics between the sediment and the remediation material are extracted as two-dimensional feature vectors.

4. The method for in-situ solidification and remediation of heavy metals in riverbed sediment according to claim 3, characterized in that, Optimization using surface fitting: Initial proportions obtained from multiple linear regression Preliminary prediction of sizing ratio using mixed density network and the content of heavy metals in the sediment and physicochemical properties As variables, construct the surface fitting function: ； in, , , , , These are the highest degrees of the polynomials corresponding to the variables. , , , , These represent the degrees of the polynomial of the dependent variable. The coefficients are the surface fitting coefficients; by minimizing the objective function Adjust using optimization algorithms The value of makes A closer approximation of the actual required repair material ratio; This refers to the ratio of the repair material k after surface fitting optimization; Dimensionality reduction and feature extraction were performed using self-organizing mapping: The heavy metal content of sediment was analyzed. Physicochemical properties and the optimized ratio after surface fitting Composition of feature vectors The self-organizing map (SOM) algorithm is used to analyze the eigenvectors. Dimensionality reduction and feature extraction are performed. SOM maps high-dimensional feature vectors onto a two-dimensional grid, and the mapped features are represented as follows: , , The number of nodes in the two-dimensional grid; during the training of the SOM network, the weights between neurons are adjusted to achieve the objective function. Optimize; in This represents the feature vector of the s-th sample. Represents the input vector With neurons weight vector The distance is used to iteratively update the weights to minimize the distance. To achieve feature extraction; Hybrid density network final prediction: feature representation obtained from self-organizing map As input to a hybrid density network, its output probability density function is: By maximizing the log-likelihood function Training parameters yields the final proportion of repair material k. .

5. The method for in-situ solidification and remediation of heavy metals in riverbed sediment according to claim 1, characterized in that, In S1, the heavy metal content in the sediment was determined by atomic absorption spectrometry and inductively coupled plasma mass spectrometry to determine the content of heavy metals such as cadmium, lead, mercury, chromium and arsenic. The physicochemical properties of the sediment were determined by pH measurement using a pH meter, particle size distribution measurement using a laser particle size analyzer, and organic matter content measurement using the loss on ignition method.

6. The method for in-situ solidification and remediation of heavy metals in river sediment according to claim 1, characterized in that, In S2, the curing characteristics of different heavy metals are specifically considered as follows: for cadmium, phosphate materials are mixed with cement as the repair material; for lead, a combination of iron oxide and lime is used as the repair material. For mercury, sulfides are used in combination with bentonite as remediation materials; For chromium, silicate cement and fly ash are selected as repair materials; For arsenic, iron-containing materials and clay minerals are used as remediation materials.

7. The method for in-situ solidification and remediation of heavy metals in riverbed sediment according to claim 1, characterized in that, In S2, the preliminary ratio of the repair materials is initially calculated using a multiple linear regression algorithm. Specifically, a multiple linear regression model is established, with heavy metal content data and physicochemical property data as input variables and the ratio of the repair materials as output variables. The model parameters are estimated using the least squares method to minimize the error between the predicted ratio of the repair materials and the actual effective ratio.

8. The method for in-situ solidification and remediation of heavy metals in river sediment according to claim 1, characterized in that, When preparing the remediation material in S3, each remediation material is weighed according to the ratio and then mixed evenly by mechanical stirring; when applying the remediation material, a mud pump and a spraying device are used to evenly apply the remediation material to the polluted layer of river bottom sediment.

9. The method for in-situ solidification and remediation of heavy metals in riverbed sediment according to claim 1, characterized in that, In S4, the stirring speed and stirring time are controlled. Specifically, the stirring speed is controlled between 50 and 150 r / min, and the stirring time is calculated and determined based on the bottom mud thickness and the stirring speed coefficient of the stirring equipment. The stirring speed coefficient is determined in advance through experiments.

10. The method for in-situ solidification and remediation of heavy metals in riverbed sediment according to claim 1, characterized in that, In S5, sediment samples were collected at time intervals of 1 day, 7 days, and 30 days after the completion of the repair. The heavy metal content after repair was measured, and the solidification rate was calculated. The formula is: solidification rate = (initial heavy metal content - heavy metal content after repair) / initial heavy metal content × 100%. At the same time, the changes in pH value, particle size distribution, organic matter content and physicochemical properties of the sediment before and after repair were compared.

Images