Rare earth mother liquor efficient recovery method and system based on surrounding water quality monitoring

Abstract

The invention relates to the technical field of water quality monitoring analysis and treatment, and discloses a rare earth mother liquor efficient recovery method and system based on surrounding water quality monitoring. The method comprises the following steps: performing component analysis on the rare earth mother liquor, and establishing a peripheral water quality monitoring network; based on the component and water quality data, the mother liquor is divided into A, B and C classes; the class A is precipitated by magnesium bicarbonate, the class B is mixed with the filtrate and is subjected to gradient extraction by a DOAM-PPA extraction agent, the class C is separated by ion exchange resin, and finally, the high-purity rare earth product is prepared by pH graded elution and crystallization. According to the method and the device, the problem of low resource utilization efficiency caused by lack of differentiated processing in the traditional technology is solved, and the technical problems that the existing process and environment monitoring are lack of a linkage mechanism and process parameter dynamic optimization cannot be realized are solved.

Description

Technical Field

[0001] The present application relates to the technical field of water quality monitoring, analysis and treatment, and in particular to a method and system for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring. Background Art

[0002] Rare earth elements (REEs), due to their unique optical, electronic, and magnetic properties, are widely used in modern industrial production and high-tech fields. The mother liquor produced during the rare earth separation process contains large quantities of REEs and other valuable metals, making it an important secondary resource. Currently, the main methods for treating REE mother liquors include chemical precipitation, extraction, and ion exchange. These methods utilize various physical and chemical principles to recover and reuse the REEs in the mother liquor.

[0003] However, existing rare earth mother liquor recovery technologies have significant shortcomings. While traditional chemical precipitation methods offer simplicity, they suffer from low recovery rates and struggle to fully recover rare earth elements. Conventional extraction methods, while offering high separation efficiency, are prone to forming difficult-to-handle emulsions and suffer from significant loss of extractant. Ion exchange methods offer excellent selectivity for rare earth elements, but they suffer from long processing cycles and high resin regeneration costs. More seriously, existing technologies suffer from widespread issues such as high energy consumption, high reagent consumption, and significant secondary pollution. Furthermore, there is a lack of effective linkage between the recovery process and environmental impact, resulting in significant pollution of the surrounding water environment during rare earth mother liquor treatment.

[0004] With increasingly stringent environmental standards, technological approaches that focus solely on rare earth recovery rates while ignoring environmental impacts are no longer sustainable. How to integrate surrounding water quality monitoring data with rare earth mother liquor recovery process parameters in real time to achieve a balance between efficient recovery and environmental protection has become a pressing technical challenge. Furthermore, existing technologies lack differentiated treatment strategies for rare earth mother liquors of varying properties, resulting in inefficient resource utilization. They also lack adaptive control mechanisms based on water environmental capacity, making it impossible to dynamically optimize process parameters based on environmental carrying capacity. Therefore, developing an efficient rare earth mother liquor recovery method that can intelligently classify, differentiate, and control rare earth mother liquors in real time based on water quality monitoring data is crucial. Summary of the Invention

[0005] The present application provides a method and system for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring, which is used to solve the problem of low resource utilization efficiency caused by the lack of differentiated treatment in traditional technologies, as well as the technical problem that the existing process and environmental monitoring lack a linkage mechanism and cannot achieve dynamic optimization of process parameters.

[0006] In the first aspect, the present application provides an efficient recovery method for rare earth mother liquor based on surrounding water quality monitoring, and the efficient recovery method for rare earth mother liquor based on surrounding water quality monitoring includes: performing inductively coupled plasma mass spectrometry analysis on the rare earth mother liquor, and establishing a water quality monitoring network around the plant to obtain mother liquor composition data and water quality baseline data; based on the mother liquor composition data and the water quality baseline data, the rare earth mother liquor is divided into Class A mother liquor, Class B mother liquor and Class C mother liquor, and differentiated treatment routes are adopted for different categories; magnesium bicarbonate is added to the Class A mother liquor for precipitation, and the Class B mother liquor is mixed with the filtrate and then gradient extracted with a DOAM-PPA extractant, and the process parameters are dynamically adjusted according to the water quality data; the extracted enriched liquid is mixed with the Class C mother liquor and separated by a D151 and D301 ion exchange resin system, and a high-purity rare earth product is obtained by crystallization after pH graded elution.

[0007] In a second aspect, the present application provides a rare earth mother liquor efficient recovery system based on surrounding water quality monitoring, the rare earth mother liquor efficient recovery system based on surrounding water quality monitoring comprising:

[0008] An analysis module is used to perform inductively coupled plasma mass spectrometry analysis on rare earth mother liquors. A water quality monitoring network is established around the plant to obtain mother liquor composition data and water quality baseline data.

[0009] a processing module, configured to classify the rare earth mother liquor into Class A mother liquor, Class B mother liquor, and Class C mother liquor based on the mother liquor composition data and the water quality baseline data, and adopt differentiated processing routes for different categories;

[0010] An extraction module is used to add magnesium bicarbonate to the Class A mother liquor to precipitate, mix the Class B mother liquor with the filtrate, perform gradient extraction with a DOAM-PPA extractant, and dynamically adjust process parameters based on water quality data;

[0011] The separation module is used to mix the extraction enrichment liquid with the Class C mother liquor and separate them through the D151 and D301 ion exchange resin systems, and then crystallize them after pH graded elution to obtain high-purity rare earth products.

[0012] In the third aspect, a rare earth mother liquor efficient recovery device based on surrounding water quality monitoring is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the rare earth mother liquor efficient recovery device based on surrounding water quality monitoring performs the above-mentioned rare earth mother liquor efficient recovery method based on surrounding water quality monitoring.

[0013] In a fourth aspect, a computer-readable storage medium is provided, wherein instructions are stored in the computer-readable storage medium, which, when executed on a computer, enables the computer to execute the above-mentioned method for efficient recovery of rare earth mother liquor based on surrounding water quality monitoring.

[0014] The technical solution provided in this application uses inductively coupled plasma mass spectrometry analysis and the establishment of a water quality monitoring network to accurately monitor and analyze the composition of the mother liquor and the surrounding water quality, providing a data foundation for the intelligent classification and differentiated treatment of rare earth mother liquors. Using mother liquor composition data and water quality baseline data, rare earth mother liquors are classified into Class A, Class B, and Class C. A differentiated treatment route is adopted, significantly improving resource utilization efficiency and avoiding the low processing efficiency caused by traditional single processes. An adaptive K-means clustering algorithm and a hierarchical decision tree algorithm are used for mother liquor classification, fully considering the multidimensional characteristics of rare earth element content, acidity, and impurity content, making the classification results more scientific and reasonable. The introduction of artificial intelligence algorithms such as bidirectional long-short-term memory networks, wavelet transforms, and convolutional neural networks enables deep mining and feature extraction of water quality data. The contribution of these algorithms in specific functional applications is reflected in their ability to effectively process time series data and accurately capture spatial distribution characteristics. The neural network time series model can identify periodic fluctuation characteristics and trends in water quality data, providing a decision-making basis for the dynamic adjustment of process parameters. The Class A mother liquor undergoes magnesium bicarbonate precipitation, the Class B mother liquor is mixed with the filtrate and then subjected to gradient extraction using the DOAM-PPA extractant, and the Class C mother liquor undergoes ion exchange. This targeted treatment strategy leverages the strengths of each process and improves overall recovery efficiency. During the magnesium bicarbonate precipitation process, precise temperature control and accurate calculation of the magnesium bicarbonate dosage ensure the completeness of the precipitation reaction while enabling the recycling of magnesium salts and CO2, significantly reducing production costs and environmental impact. The gradient extraction utilizes a three-stage process using the DOAM-PPA extractant, achieving efficient separation of rare earth elements under varying pH and temperature conditions. Compared to the traditional P507 extractant, it exhibits a higher separation factor and saturation capacity. The ion exchange system, combining the properties of both D151 and D301 resins, effectively treats low-concentration rare earth mother liquors.

[0015] The present invention realizes the real-time linkage adjustment of water quality monitoring data and process parameters, dynamically optimizes the extraction process parameters according to the changes in the rare earth ion content in the water quality data, and avoids the waste of reagents and environmental pollution caused by over-extraction. The pH gradient elution technology realizes the fine separation of rare earth elements, and the crystallization and roasting processes ensure the high purity of the final product. The introduction of graph neural networks uses monitoring sites as nodes and water flow directions as edges, and simulates the migration and diffusion process of pollutants through a message passing mechanism. This innovative application realizes the accurate assessment and prediction of water environment capacity and provides a scientific basis for the environmentally friendly adjustment of process parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 This is a schematic diagram of an embodiment of a method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring in an embodiment of the present application;

[0018] Figure 2 This is a schematic diagram of an embodiment of a rare earth mother liquor efficient recovery system based on surrounding water quality monitoring in an embodiment of the present application;

[0019] Figure 3 It is a schematic block diagram of the structure of a rare earth mother liquor efficient recovery device based on surrounding water quality monitoring in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present application embodiment provides a kind of rare earth mother liquor efficient recovery method and system based on peripheral water quality monitoring. The terms "first", "second", "third", "fourth" etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or precedence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in a sequence other than the content illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or equipment comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0021] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiments of the present application, an embodiment of a method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring includes:

[0022] Step S101: Perform inductively coupled plasma mass spectrometry analysis on the rare earth mother liquor, and simultaneously establish a water quality monitoring network around the plant to obtain mother liquor composition data and water quality baseline data;

[0023] Step S102: Based on the mother liquor composition data and the water quality baseline data, the rare earth mother liquor is divided into Class A mother liquor, Class B mother liquor, and Class C mother liquor, and differentiated treatment routes are adopted for different categories;

[0024] Step S103: adding magnesium bicarbonate to the Class A mother liquor to precipitate, mixing the Class B mother liquor with the filtrate, and performing gradient extraction using a DOAM-PPA extractant, and dynamically adjusting process parameters based on water quality data;

[0025] Step S104: The extract enrichment solution is mixed with the Class C mother liquor and separated by a D151 and D301 ion exchange resin system. After pH graded elution, the mixture is crystallized to obtain a high-purity rare earth product.

[0026] It is understandable that the execution subject of this application can be a rare earth mother liquor efficient recovery system based on surrounding water quality monitoring, or a terminal or server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0027] Specifically, before the rare earth mother liquor is recovered and processed, the mother liquor is subjected to multi-element analysis using inductively coupled plasma mass spectrometry. The mother liquor samples are pretreated using a multi-spectral excitation sequence, and the content of each element is obtained through step ionization scanning. At the same time, an array of adaptive monitoring stations is established at key hydrological nodes upstream and downstream of the plant, and dual-wavelength fluorescence spectroscopy technology is used to monitor changes in the rare earth element content in the water in real time. The acquired multi-element spectral distribution data is subjected to noise reduction processing using a self-correcting multivariate analysis algorithm to eliminate background interference and random noise and improve data reliability. The dynamic distribution curve of the rare earth element content is analyzed using a neural network time series model, which includes a bidirectional long and short-term memory network, wavelet transform multi-scale decomposition, and a convolutional neural network. Ultimately, a digital twin model of water environment capacity is constructed to predict water quality changes.

[0028] Based on the analyzed mother liquor composition data and water quality baseline data, rare earth mother liquor is classified and processed. Specifically, a multivariate cluster analysis system uses an adaptive K-means clustering algorithm to spatially partition the three-dimensional characteristics of rare earth content, acidity, and impurity content, generating a characteristic distribution map. A hierarchical decision tree algorithm is applied to this map for boundary optimization. Classification criteria are set based on the environmental capacity thresholds in the water quality baseline data, and a mother liquor classification decision model is established. This model uses real-time intelligent classification to categorize rare earth mother liquor into Class A with high rare earth and low impurities, Class B with medium rare earth and medium impurities, and Class C with low rare earth and high impurities. Physical diversion is achieved through an intelligent control pipeline system, allowing for different treatment routes.

[0029] Different treatment methods are used for different types of mother liquor. Class A mother liquor is thermostatically controlled in a double-layer reactor, precisely adjusting the temperature to 40±1°C. A magnesium bicarbonate solution calculated to be 1.8 times the molar number of rare earth ions is added for precipitation reaction, resulting in rare earth carbonate precipitate, filtrate, and recyclable CO2 gas, thus realizing the recycling of magnesium salts. The filtrate is mixed with Class B mother liquor, the pH value is adjusted to 1.5±0.1, and after treatment with an oxidant, it is introduced into a three-stage countercurrent extraction device. DOAM-PPA extractant is used as the extractant to form a gradient extraction environment. This extractant has a higher separation coefficient and saturation capacity. Based on the real-time changes in the rare earth ion content in the water quality data, the extraction process parameters, including extraction time, phase ratio, and pH value, are dynamically adjusted through a multivariable adaptive control system.

[0030] The extract concentrate is mixed with the Class C mother liquor in a specific ratio, pretreated with a reducing agent, and then filtered through a filter to produce a clear, uniform ion exchange feed solution. This feed solution is then introduced into a series system consisting of a D151 strongly acidic cation exchange resin and a D301 weakly basic anion exchange resin for exchange adsorption. A pH gradient elution technique is employed on the saturated resin bed, sequentially eluting with acidic solutions of varying pH values ​​to produce rare earth enriched solutions of varying compositions. These enriched solutions are then divided into light rare earth, medium rare earth, and heavy rare earth groups, and subjected to differentiated temperature-controlled crystallization treatments. Subsequently, high-purity rare earth products are produced through calcination and conversion.

[0031] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0032] The rare earth mother liquor is pretreated by a multi-spectral excitation sequence, and the pretreated mother liquor is subjected to step ionization scanning by an inductively coupled plasma mass spectrometer to obtain multi-element spectrum distribution data;

[0033] An array of adaptive monitoring stations was set up at key hydrological nodes upstream and downstream of the plant to conduct real-time high-frequency dual-wavelength fluorescence spectroscopy monitoring of the water, generating a dynamic distribution curve of rare earth element content.

[0034] The multi-element spectrum distribution data were subjected to noise reduction by applying a self-correcting multivariate analysis algorithm, and quantitative conversion was performed in combination with the standard substance calibration curve to obtain the mother liquor composition data.

[0035] The dynamic distribution curve of rare earth element content is analyzed and processed through a neural network time series model to identify periodic fluctuation characteristics and trends, construct a digital twin model of water environment capacity, and obtain water quality baseline data with predictive function.

[0036] Specifically, before the rare earth mother liquor is recovered and processed, the mother liquor is subjected to multi-element analysis using inductively coupled plasma mass spectrometry. This method uses a multi-spectral excitation sequence to pretreat the mother liquor sample and obtains the content of each element through step ionization scanning. At the same time, an adaptive monitoring station array is established at key hydrological nodes upstream and downstream of the plant, and dual-wavelength fluorescence spectroscopy technology is used to monitor changes in the rare earth element content in the water in real time. The obtained multi-element spectral distribution data is subjected to noise reduction processing using a self-correcting multivariate analysis algorithm to eliminate background interference and random noise and improve data reliability. The dynamic distribution curve of the rare earth element content is analyzed using a neural network time series model, which includes a bidirectional long and short-term memory network, wavelet transform multi-scale decomposition, and a convolutional neural network. Ultimately, a digital twin model of water environment capacity is constructed to predict water quality changes.

[0037] Based on the analyzed mother liquor composition data and water quality baseline data, rare earth mother liquor is classified and processed. Specifically, a multivariate cluster analysis system uses an adaptive K-means clustering algorithm to spatially partition the three-dimensional characteristics of rare earth content, acidity, and impurity content, generating a characteristic distribution map. A hierarchical decision tree algorithm is applied to this map for boundary optimization. Classification criteria are set based on the environmental capacity thresholds in the water quality baseline data, and a mother liquor classification decision model is established. This model uses real-time intelligent classification to categorize rare earth mother liquor into Class A with high rare earth and low impurities, Class B with medium rare earth and medium impurities, and Class C with low rare earth and high impurities. Physical diversion is achieved through an intelligent control pipeline system, allowing for different treatment routes.

[0038] Different treatment methods are used for different types of mother liquor. Class A mother liquor is thermostatically controlled in a double-layer reactor, precisely adjusting the temperature to 40±1°C. A magnesium bicarbonate solution calculated to be 1.8 times the molar number of rare earth ions is added for precipitation reaction, resulting in rare earth carbonate precipitate, filtrate, and recyclable CO2 gas, thus realizing the recycling of magnesium salts. The filtrate is mixed with Class B mother liquor, the pH value is adjusted to 1.5±0.1, and after treatment with an oxidant, it is introduced into a three-stage countercurrent extraction device. DOAM-PPA extractant is used as the extractant to form a gradient extraction environment. This extractant has a wider pH operating window and higher separation selectivity. Based on the real-time changes in the rare earth ion content in the water quality data, the extraction process parameters, including extraction time, phase ratio, and pH value, are dynamically adjusted through a multivariable adaptive control system.

[0039] Finally, the extract is mixed with the Class C mother liquor in a specific ratio, pretreated with a reducing agent, and filtered through a filter to produce a clear, homogeneous ion exchange feed solution. This feed solution is then introduced into a series system consisting of a D151 strongly acidic cation exchange resin and a D301 weakly basic anion exchange resin for exchange adsorption. A pH gradient elution technique is employed on the saturated resin bed, sequentially eluting with acidic solutions of varying pH values ​​to produce rare earth enriched solutions of varying compositions. These enriched solutions are then divided into light rare earth, medium rare earth, and heavy rare earth groups, which are then subjected to differentiated temperature-controlled crystallization treatments. Subsequently, high-purity rare earth products are produced through calcination and conversion. The entire process utilizes the magnesium bicarbonate method to efficiently recycle magnesium salts and CO2, eliminating the environmental pollution associated with conventional processes. Furthermore, the use of the DOAM-PPA extractant significantly improves the separation efficiency of heavy rare earth elements.

[0040] In a specific embodiment, the step of analyzing and processing the dynamic distribution curve of the rare earth element content through a neural network time series model may specifically include the following steps:

[0041] The dynamic distribution curve of rare earth element content is input into a bidirectional long short-term memory network for temporal feature extraction. The bidirectional long short-term memory network contains a three-layer hidden layer structure, and the number of memory units in each layer is set to twice the number of rare earth elements. The temporal encoding feature vector is obtained;

[0042] Apply wavelet transform to the temporal coding feature vector for multi-scale decomposition, extract the fluctuation characteristics of different frequency domains, and assign weights to each frequency component through the self-attention mechanism to obtain a multi-frequency domain weighted feature matrix;

[0043] The multi-frequency domain weighted feature matrix is ​​input into a convolutional neural network for spatial feature learning. The convolutional neural network is composed of three convolutional layers and two pooling layers alternating with convolution kernel sizes of 3×3, 5×5, and 7×7, respectively, to obtain the spatial distribution pattern characteristics of water quality.

[0044] A graph neural network model is constructed based on the characteristics of the spatial distribution pattern of water quality. The monitoring sites are used as nodes and the water flow direction is used as edges. The migration and diffusion process of pollutants is simulated through the message passing mechanism to obtain water quality baseline data with predictive function.

[0045] Specifically, the dynamic distribution curves of rare earth element content collected from an array of monitoring stations upstream and downstream of the plant were input into a bidirectional long short-term memory (BiLSTM) network for time series feature extraction. The bidirectional nature of the BiLSTM network is reflected in its simultaneous consideration of information from both historical and future time points. Time series features are abstracted layer by layer through a three-layer hidden layer structure. For the 14 rare earth elements detected (including lanthanum, cerium, neodymium, and praseodymium), each layer contains 28 memory cells (twice the number of rare earth elements) to ensure sufficient time series information is captured. Each BiLSTM memory cell consists of three components: an input gate, a forget gate, and an output gate. Through selective memory and forgetting, it achieves long-term dependency learning and ultimately outputs a time series encoding feature vector. The resulting time series encoding feature vector is then subjected to multi-scale wavelet transform decomposition, breaking down the time series signal into fluctuation characteristics in different frequency domains. Compared to the traditional Fourier transform, the wavelet transform provides information in both the time and frequency domains, making it more suitable for analyzing non-stationary signals. In the specific operation, the Daubechies wavelet basis function is used for a five-level decomposition to obtain approximate coefficients and detail coefficients for different frequency bands. Then, each frequency component is assigned a weight through the self-attention mechanism. The self-attention mechanism calculates the correlation score between each frequency component. Frequency components with high correlation obtain higher weights, thus forming a multi-frequency domain weighted feature matrix.

[0046] The multi-frequency weighted feature matrix is ​​input into a convolutional neural network (CNN) for spatial feature learning. This CNN consists of three alternating convolutional layers and two pooling layers. The first convolutional layer uses a 3×3 convolution kernel to extract local subtle features, the second convolutional layer uses a 5×5 convolution kernel to capture medium-scale features, and the third convolutional layer uses a 7×7 convolution kernel to capture large-scale spatial patterns. The pooling layer uses a max pooling operation to reduce the dimensionality of the feature map while retaining significant features. Through this progressive feature extraction method, the CNN outputs the spatial distribution pattern of water quality, reflecting the spatial correlation between different monitoring points.

[0047] Based on the spatial distribution patterns of water quality, a graph neural network (GNN) model was constructed, with monitoring sites as nodes and water flow directions as edges, forming a water area topology. The GNN simulates the migration and diffusion of pollutants in a water system through a message-passing mechanism. Information transmission between nodes follows the direction of water flow, and changes in the status of upstream nodes affect the status of downstream nodes. After multiple rounds of iterative updates, the GNN generates predictive water quality baseline data, capable of predicting changes in rare earth element content at each monitoring site under different operating conditions.

[0048] In a rare earth separation plant application, rare earth element (REE) content data was collected for 30 consecutive days from five monitoring points located 500 and 1000 meters upstream and 500, 1000, and 2000 meters downstream of the plant, generating time series curves for 14 rare earth elements. After processing using a BiLSTM network, the original 10,080-dimensional data (24 hours × 30 days × 14 elements) was compressed into a 392-dimensional time series encoding feature vector. A wavelet transform decomposed the feature vector into five sub-bands. A self-attention mechanism calculated the weights of low-frequency components at approximately 0.6, mid-frequency components at approximately 0.3, and high-frequency components at approximately 0.1, forming a multi-frequency weighted feature matrix. CNN processing generated a feature map reflecting the spatial relationships between the five monitoring points. The resulting GNN model achieved over 90% accuracy in predicting REE content at downstream monitoring points within a 24-hour prediction window, providing reliable data support for the precise classification of rare earth mother liquor and process parameter optimization.

[0049] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0050] The mother liquor composition data was input into the multivariate cluster analysis system, and the adaptive K-means clustering algorithm was used to spatially divide the three-dimensional characteristics of rare earth element content, acidity and impurity content to obtain the characteristic distribution map of rare earth mother liquor;

[0051] A hierarchical decision tree algorithm was applied to optimize the boundaries of the characteristic distribution map, and the classification criteria were set based on the environmental capacity threshold in the water quality baseline data to obtain a mother liquor classification decision model.

[0052] The rare earth mother liquor is intelligently classified in real time through the mother liquor classification decision model, and is automatically divided into high rare earth and low impurity category A, medium rare earth and medium impurity category B, and low rare earth and high impurity category C according to the total rare earth content, acidity and impurity content parameters, to obtain the classified and marked mother liquor flow;

[0053] The classified and marked mother liquor flow is physically diverted through an intelligent control pipeline system, with a magnesium bicarbonate precipitation treatment route assigned to Class A mother liquor, a DOAM-PPA extraction treatment route assigned to Class B mother liquor, and an ion exchange treatment route assigned to Class C mother liquor, resulting in three differentiated treatment process paths.

[0054] Specifically, the mother liquor composition data obtained by inductively coupled plasma mass spectrometry analysis was input into a multivariate cluster analysis system, which used an adaptive K-means clustering algorithm to process the data. This adaptive K-means clustering algorithm is an improvement on the traditional K-means algorithm, automatically adjusting the number of cluster centers based on data distribution characteristics without the need for manually specifying the K value. The algorithm uses three key parameters of the rare earth mother liquor—total rare earth content, acidity, and impurity content—as three-dimensional features to construct a feature space. By iteratively calculating the Euclidean distance from sample points to each cluster center and continuously adjusting the cluster center positions, it ultimately forms a characteristic distribution map of the rare earth mother liquor. The resulting characteristic distribution map is then subjected to boundary optimization using a hierarchical decision tree algorithm. This tree-based classification algorithm uses information gain or Gini impurity metrics to select the optimal partitioning attributes and construct decision rules layer by layer. In this method, the hierarchical decision tree first examines the primary attribute, total rare earth content, followed by acidity and impurity content, forming a hierarchical classification structure. At the same time, the environmental capacity threshold in the water quality baseline data is combined in the construction of the decision tree, and the environmental carrying capacity is used as a constraint to ensure that the classification results take into account both the recycling value and the environmental impact, thereby obtaining an environmentally friendly mother liquor classification decision model.

[0055] The above-mentioned mother liquor classification decision model performs real-time intelligent classification of rare earth mother liquor. Based on the combined values ​​of three parameters—total rare earth content, acidity, and impurity content—it is automatically divided into three categories: Category A, which represents high rare earth and low impurities, with a total rare earth content greater than 10 g / L, an acidity less than 1 mol / L, and an impurity content less than 5%; Category B, which represents medium rare earth and medium impurities, with a total rare earth content between 5-10 g / L, an acidity between 1-3 mol / L, and an impurity content between 5-15%; and Category C, which represents low rare earth and high impurities, with a total rare earth content less than 5 g / L, an acidity greater than 3 mol / L, and an impurity content greater than 15%. This multi-parameter classification method enables a comprehensive assessment of the mother liquor's value and treatment difficulty, resulting in a classified and labeled mother liquor stream.

[0056] The classified and labeled mother liquor streams are physically diverted through an intelligent control piping system consisting of solenoid valves, flow meters, and a multi-channel distributor. The system controls the valve opening and closing states based on the classification tag signals, directing different types of mother liquor to the corresponding treatment routes. Specifically, Class A mother liquors are assigned to a magnesium bicarbonate precipitation treatment route. Magnesium bicarbonate precipitants are environmentally friendly and recyclable, making them suitable for direct precipitation and recovery of mother liquors with high rare earth content. Class B mother liquors are assigned to a DOAM-PPA extraction treatment route. DOAM-PPA is a novel extractant with a wider pH operating window and higher separation selectivity, demonstrating excellent extraction performance for medium-concentration rare earth elements. Class C mother liquors are assigned to an ion exchange treatment route. Ion exchange resins are capable of effectively enriching rare earth elements under low-concentration, high-impurity conditions. This differentiated treatment strategy optimizes the process path for each mother liquor, achieving the recycling of magnesium salts and CO2, and improving overall recovery efficiency and environmental performance.

[0057] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0058] The Class A mother liquor is thermostatically controlled in a double-layer reactor to precisely adjust the temperature to 40±1°C, and a magnesium bicarbonate solution having a calculated amount of 1.8 times the molar number of rare earth ions is added to perform a precipitation reaction to obtain a rare earth carbonate precipitate, a filtrate, and recyclable CO2 gas;

[0059] The filtrate is mixed with the type B mother liquor, and the pH value is adjusted to 1.5±0.1. An oxidant is added to the mixed liquor for pretreatment to obtain a pre-extraction liquid with a stable valence state;

[0060] The pre-extraction liquid was introduced into a three-stage countercurrent extraction device, DOAM-PPA was used as the extractant, and different pH values ​​and temperature conditions were set in the first, second and third stages respectively to form a gradient extraction environment to obtain an organic phase enriched in rare earth elements;

[0061] According to the real-time changes in the rare earth ion content in the water quality data, the extraction process parameters are dynamically adjusted through a multivariable adaptive control system, including adjusting the extraction time, phase ratio and pH value. At the same time, magnesium salts and CO2 are recycled to obtain extraction enrichment liquid and recycled auxiliary materials.

[0062] Specifically, the process is carried out in a double-layer reactor, with an inner layer made of polytetrafluoroethylene for corrosion protection and an outer layer made of stainless steel for mechanical strength. A constant-temperature water circulation system is introduced into the middle layer, precisely controlling the reaction temperature to 40±1°C. Temperature control is crucial: too low a temperature will slow the reaction rate, while too high a temperature will encourage some light rare earth elements to form soluble complexes. When adding magnesium bicarbonate solution, the dosage is calculated as 1.8 times the molar number of rare earth ions. This ratio takes into account the stoichiometric ratio of magnesium bicarbonate to rare earth elements and their consumption during the reaction. The specific calculation method is to first determine the concentration of each rare earth element in the mother liquor through inductively coupled plasma mass spectrometry, convert it to a total molar number, and multiply it by 1.8 to obtain the required moles of magnesium bicarbonate. After the magnesium bicarbonate precipitation reaction is completed, the rare earth carbonate precipitate, filtrate, and recyclable CO2 gas are separated using a vacuum filtration device, enabling the recycling of magnesium salts. The filtrate, which still contains some unprecipitated rare earth elements, is mixed with a Type B mother liquor containing impurities from the intermediate rare earth elements. A pH meter is used to monitor the pH of the mixture in real time, and acid or base is added to adjust the pH to 1.5±0.1, the optimal balance between extraction efficiency and stability. An oxidizing agent (such as H2O2 or NaClO) is then added to pretreat the mixture, ensuring that the ferrous iron in the solution is oxidized to ferric iron and that the cerium trivalent remains stable, thus preventing precipitation or emulsification during the extraction process and producing a stable pre-extraction liquid.

[0063] The pre-extraction liquid after pretreatment was introduced into a three-stage countercurrent extraction apparatus consisting of 20 extraction units connected in series, divided into three functional sections. DOAM-PPA (di-n-octylaminomethyl)phenylphosphite) was used as the extractant, which exhibits excellent selectivity for heavy rare earth elements, high extraction capacity, and a wide pH operating window. The key to this three-stage gradient extraction lies in setting different process parameters in each section: the first section (units 1-8) is the crude extraction section, with a pH controlled at 2.5±0.1 and a temperature of 40±1°C; the second section (units 9-16) is the fine extraction section, with a pH of 3.0±0.1 and a temperature of 45±1°C; and the third section (units 17-20) is the wash section, using dilute nitric acid at a pH of 1.5±0.1 and a temperature of 25±2°C. This gradient configuration allows the rare earth elements to be gradually enriched in the organic phase during the extraction process, and the DOAM-PPA extractant maintains efficient extraction performance over a wide pH range.

[0064] During the extraction process, rare earth ion content data collected in real time through the plant's water quality monitoring network serves as feedback and is input into a multivariable adaptive control system. Based on a fuzzy logic control algorithm, this system dynamically adjusts extraction process parameters based on changing trends in water quality data. The specific adjustment logic is as follows: when rare earth content at downstream monitoring points increases, the system automatically extends the extraction time or reduces the processing rate. When the rare earth ion content in the aqueous phase at the extraction unit outlet exceeds a set threshold, the pH value or phase ratio (the volume ratio of the organic phase to the aqueous phase) of the corresponding unit is adjusted. The phase ratio adjustment range is typically 0.8:1 to 1.2:1, and the pH value adjustment range is ±0.2 units. This closed-loop control strategy links the extraction process with environmental monitoring, while fully leveraging the technical advantages of the magnesium bicarbonate method and the DOAM-PPA extractant, ensuring efficient recovery while minimizing environmental impact and achieving circular economy benefits.

[0065] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0066] The extract enriched liquid is mixed with the Class C mother liquor in a specific ratio, a reducing agent is added to the mixed liquid for pretreatment, and the mixed liquid is filtered through a filter to obtain a clear and uniform ion exchange feed liquid;

[0067] The ion exchange feed liquid is introduced into a series system composed of D151 and D301 ion exchange resins, and the exchange adsorption is carried out at an appropriate flow rate to obtain a saturated resin bed enriched with rare earth elements;

[0068] The pH gradient elution technology is used on the saturated resin bed, and acidic solutions with different pH values ​​are used in sequence for graded elution to obtain rare earth enriched solutions with different compositions;

[0069] The rare earth enriched liquid is divided into light rare earth group, medium rare earth group and heavy rare earth group, and subjected to differentiated temperature-controlled crystallization treatment respectively. After subsequent roasting and transformation, high-purity rare earth products are obtained.

[0070] Specifically, the enriched liquid obtained from the initial extraction process is mixed with a Type C mother liquor (low in rare earth elements and high in impurities) at a volume ratio of 2:1. This ratio, determined based on the low rare earth content but large volume of the Type C mother liquor, ensures the ion exchange system's processing efficiency while fully utilizing the rare earth resources within it. A reducing agent, Na₂SO₃, is added to the mixture at a dosage of 1% of the mixture's volume. This reduces some rare earth elements to stable valence states, specifically reducing tetravalent cerium to trivalent cerium, thereby preventing precipitate formation during the subsequent ion exchange process. The pretreated mixture is then filtered through a 0.22μm membrane filter to remove insoluble impurities, resulting in a clear, homogeneous ion exchange feed solution.

[0071] The prepared ion exchange feed solution is introduced into a series-connected ion exchange system consisting of a D151 strongly acidic cation exchange resin column and a D301 weakly basic anion exchange resin column. The D151 resin is a sulfonic acid-based macroporous cation exchange resin with a strongly acidic group, -SO3H, that exhibits excellent selective adsorption for trivalent rare earth ions. The D301 resin is a tertiary amine-based weakly basic anion exchange resin with a -NR2 functional group, primarily used to remove anionic impurities from the solution. During the ion exchange process, the feed flow rate is maintained at 2.0 BV / h (BV represents the resin bed volume, meaning the volume of liquid passing through per hour is twice the resin bed volume). This flow rate ensures sufficient exchange contact time while balancing treatment efficiency. As the exchange process progresses, the D151 resin gradually saturates. When the rare earth ion concentration at the outlet reaches 5% of the inlet concentration, indicating near saturation, adsorption ceases and the elution phase begins.

[0072] A pH gradient elution technique was used on a saturated D151 resin bed, using hydrochloric acid solutions with pH values ​​of 4.0, 3.0, 2.0, and 1.0, respectively. The principle of pH gradient elution is to separate rare earth elements by exploiting the differences in their desorption capacity from the resin under different pH conditions. Specifically, elution was first performed with hydrochloric acid at pH 4.0, primarily eluting light rare earth elements such as lanthanum, cerium, praseodymium, and neodymium. Elution was then performed with hydrochloric acid at pH 3.0, primarily eluting medium rare earth elements such as samarium, europium, and gadolinium. Elution was then performed with hydrochloric acid at pH 2.0, primarily eluting some medium and heavy rare earth elements. Finally, elution was performed with hydrochloric acid at pH 1.0, primarily eluting heavy rare earth elements such as dysprosium, holmium, erbium, thulium, ytterbium, and lutetium. During the elution process, the changes in rare earth ion concentration in the eluent are tracked in real time through an online monitoring device. When the rare earth ion concentration in a certain level of eluent drops below 0.1g / L, the eluent is switched to the next level to achieve precise segmented collection and obtain rare earth enriched solutions with different compositions.

[0073] The collected rare earth enrichment solution was divided into light rare earth, medium rare earth, and heavy rare earth groups according to their composition. Different temperature-controlled crystallization processes were applied to each group. The light rare earth group was precipitated by adding magnesium bicarbonate at a temperature of 60±1°C (a molar ratio of 2.8:1). The medium rare earth group was precipitated by adding magnesium bicarbonate at a pH of 1.5±0.1 at a temperature of 65±1°C (a molar ratio of 3.0:1). The heavy rare earth group was precipitated by adding magnesium bicarbonate at a pH of 1.8±0.1 at a temperature of 70±1°C (a molar ratio of 3.2:1). The crystallization process lasted 12-16 hours. By controlling parameters such as crystallization temperature, time, and pH, the growth rate and morphology of the crystals were influenced, resulting in a uniform, high-purity rare earth carbonate precipitate. CO2 gas and magnesium salt solution were also recovered for recycling. After crystallization, the product is filtered, washed, dried, and calcined at 800±10°C for four hours. The carbonate decomposes and converts into rare earth oxides, resulting in the final high-purity rare earth product. This entire process recycles magnesium salts and CO2, significantly reducing production costs and eliminating the environmental pollution associated with traditional processes.

[0074] In a specific embodiment, the process of performing the step of sequentially using acidic solutions with different pH values ​​for graded elution may specifically include the following steps:

[0075] The saturated resin bed is pre-washed by using deionized water at a constant flow rate to flush the surface of the saturated resin bed to remove free ions that are not fully adsorbed, thereby obtaining a washed and balanced resin bed system;

[0076] A high pH hydrochloric acid solution is introduced into the resin bed system in the washing equilibrium to perform the first stage elution. While the online monitoring device tracks the changes in the rare earth ion concentration at the outlet in real time, the eluate collection point is precisely controlled to obtain a first eluate component rich in light rare earth elements;

[0077] A medium pH hydrochloric acid solution is introduced into the resin bed system that has completed the first stage elution to perform the second stage elution, and the eluent flow rate is dynamically adjusted according to the change trend of the outlet liquid conductivity to obtain a second eluent component rich in medium rare earth elements;

[0078] A low pH hydrochloric acid solution is introduced into the resin bed system that has completed the second stage elution for final elution. By controlling the temperature and flow rate parameters to optimize the elution efficiency, a third eluate component rich in heavy rare earth elements is obtained, while achieving precise separation of rare earth enriched liquids of different compositions.

[0079] Specifically, pH gradient elution technology is a key step in achieving fine separation of rare earth elements. The saturated resin bed is first pre-washed with deionized water with a resistivity greater than 18 MΩ·cm at a constant flow rate of 1.0 BV / h. The purpose of this pre-wash is to remove residual feed liquid and inadequately adsorbed free ions from the interstices of the resin bed, preventing these ions from interfering with the separation during elution. The rinsing process continues until the effluent conductivity remains below 50 μS / cm and remains stable, indicating that the resin bed has reached a washing equilibrium state and the resin bed system is ready for subsequent elution operations.

[0080] A high pH hydrochloric acid solution with a pH of 4.0 is introduced into the washing and balancing resin bed system for the first stage of elution. The elution flow rate is controlled at 0.8BV / h. The low flow rate ensures that the eluent is in full contact with the resin bed. The online monitoring device tracks the changes in the rare earth ion concentration at the outlet in real time. The device includes a conductivity sensor, a UV absorption detector, and an automatic sampling and analysis system. When the rare earth ion concentration in the outlet liquid begins to rise rapidly, the automatic collection system is activated. The collection standard is set to a total rare earth concentration in the eluent greater than 0.5g / L, and collection is continued until the concentration drops below 0.5g / L. At this time, most of the light rare earth elements (mainly lanthanum, cerium, praseodymium, and neodymium) have been eluted, forming a first eluent component rich in light rare earth elements.

[0081] A medium pH hydrochloric acid solution with a pH value of 3.0 is introduced into the resin bed system that has completed the first stage of elution for the second stage of elution. At this time, the eluent flow rate is dynamically adjusted according to the trend of changes in the outlet liquid conductivity. The specific adjustment strategy is: when the conductivity rises rapidly, the flow rate is reduced to 0.6BV / h to increase the contact time between the eluent and the resin; when the conductivity reaches a peak and begins to decline, the flow rate is adjusted to 1.0BV / h to speed up the elution speed. The conductivity change curve collects data through an online monitoring system, and the slope change is calculated in real time with a preset algorithm to determine the position of the conductivity change inflection point, thereby accurately controlling the collection timing. The second stage of elution mainly obtains medium rare earth elements (samarium, europium, and gadolinium), forming a second eluent component rich in medium rare earth elements.

[0082] A low-pH hydrochloric acid solution with a pH of 1.5 is introduced into the resin bed system after the second-stage elution for the final elution. During this stage, both temperature and flow rate are controlled simultaneously: the temperature is raised to 35±2°C, and the flow rate is controlled at 1.2 BV / h. The increased temperature enhances the breakage of complex bonds between heavy rare earth elements and the resin, accelerating their desorption, while the higher flow rate improves elution efficiency. The collection criteria for the final elution stage are similarly based on changes in rare earth ion concentration and conductivity in the outlet liquid, but an additional spectral analysis confirmation step is added to ensure that the collected solution is enriched in heavy rare earth elements (dysprosium, holmium, erbium, thulium, ytterbium, and lutetium), forming the third eluent fraction and achieving precise separation of the rare earth-enriched solution.

[0083] In a rare earth separation plant application, a fully saturated D151 resin bed (resin volume 5L) was subjected to pH gradient elution. A pre-wash was performed with 20L of deionized water. The first elution phase began after the effluent conductivity decreased from an initial 3200μS / cm to 45μS / cm. Elution was performed using hydrochloric acid at a pH of 4.0. The online monitoring device recorded a rapid increase in rare earth ion concentration to a peak of 2.8g / L before beginning to decline. Collection was stopped when the concentration dropped to 0.48g / L, yielding 12L of light rare earth-rich eluate. The second phase, using hydrochloric acid at a pH of 3.0, involved reducing the flow rate from the standard 0.8BV / h to 0.65BV / h based on the slope of the conductivity curve. Approximately 8L of medium rare earth-rich eluate was collected at the peak concentration. In the final stage, elution was performed at 35°C using a hydrochloric acid solution with a pH of 1.5 at a flow rate of 1.2 BV / h, collecting 6 L of heavy rare earth-rich eluate. ICP-MS analysis revealed high separation of rare earth elements among the three eluate fractions, with light rare earth elements accounting for 90% of the first fraction, medium rare earth elements exceeding 80% of the second fraction, and heavy rare earth elements exceeding 75% of the third fraction. This effectively addresses the severe cross-contamination issues associated with traditional rare earth separation techniques.

[0084] The above describes the rare earth mother liquor efficient recovery method based on surrounding water quality monitoring in the embodiment of the present application. The following describes the rare earth mother liquor efficient recovery system based on surrounding water quality monitoring in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of a rare earth mother liquor efficient recovery system based on surrounding water quality monitoring includes:

[0085] Analysis module 201 is used to perform inductively coupled plasma mass spectrometry analysis on rare earth mother liquor and establish a water quality monitoring network around the plant to obtain mother liquor composition data and water quality baseline data;

[0086] A processing module 202 is configured to classify the rare earth mother liquor into Class A mother liquor, Class B mother liquor, and Class C mother liquor based on the mother liquor composition data and the water quality baseline data, and adopt differentiated processing routes for different categories;

[0087] Extraction module 203, for adding magnesium bicarbonate to the Class A mother liquor to precipitate, mixing the Class B mother liquor with the filtrate, performing gradient extraction using a DOAM-PPA extractant, and dynamically adjusting process parameters based on water quality data;

[0088] The separation module 204 is used to separate the extracted enriched liquid and the Class C mother liquor through a D151 and D301 ion exchange resin system, and then crystallize after pH grading and elution to obtain a high-purity rare earth product.

[0089] above Figure 2 From the perspective of modular functional entities, the rare earth mother liquor efficient recovery system based on peripheral water quality monitoring in the embodiment of the present invention is described in detail. The following describes in detail the rare earth mother liquor efficient recovery equipment based on peripheral water quality monitoring in the embodiment of the present invention from the perspective of hardware processing.

[0090] Reference Figure 3 In the embodiment of the present invention, there is also provided a rare earth mother liquor efficient recovery device based on surrounding water quality monitoring. The rare earth mother liquor efficient recovery device based on surrounding water quality monitoring can be a server, and its internal structure can be as follows: Figure 3 As shown. The rare earth mother liquor efficient recovery equipment based on peripheral water quality monitoring includes a processor, a memory, a display screen, an input device, a network interface and a database connected via a system bus. Among them, the computer-designed processor is used to provide computing and control capabilities. The memory of the rare earth mother liquor efficient recovery equipment based on peripheral water quality monitoring includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the rare earth mother liquor efficient recovery equipment based on peripheral water quality monitoring is used to store the corresponding data in this embodiment. The network interface of the rare earth mother liquor efficient recovery equipment based on peripheral water quality monitoring is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0091] Those skilled in the art will understand that Figure 3 The structure shown in is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the rare earth mother liquor efficient recovery equipment based on surrounding water quality monitoring to which the solution of the present invention is applied.

[0092] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions. When the instructions are run on a computer, the computer executes the steps of the rare earth mother liquor efficient recovery method based on surrounding water quality monitoring.

[0093] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0094] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a rare earth mother liquor efficient recovery device based on peripheral water quality monitoring (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program code.

[0095] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A rare earth mother liquor efficient recovery method based on surrounding water quality monitoring, characterized in that: The method comprises: Conduct inductively coupled plasma mass spectrometry analysis on rare earth mother liquors and establish a water quality monitoring network around the plant to obtain mother liquor composition data and water quality baseline data; Based on the mother liquor composition data and the water quality baseline data, the rare earth mother liquor is divided into Class A mother liquor, Class B mother liquor and Class C mother liquor, and differentiated treatment routes are adopted for different categories; Magnesium bicarbonate is added to the Class A mother liquor to precipitate, and the Class B mother liquor is mixed with the filtrate and then subjected to gradient extraction using a DOAM-PPA extractant, and process parameters are dynamically adjusted according to water quality data; The extracted enriched liquid is mixed with the Class C mother liquor and separated by D151 and D301 ion exchange resin systems, and then crystallized after pH graded elution to obtain a high-purity rare earth product.

2. The method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring according to claim 1, characterized in that: The rare earth mother liquor is subjected to inductively coupled plasma mass spectrometry analysis, and a water quality monitoring network is established around the plant to obtain mother liquor composition data and water quality baseline data, including: The rare earth mother liquor is pretreated by a multi-spectral excitation sequence, and the pretreated mother liquor is subjected to step ionization scanning by an inductively coupled plasma mass spectrometer to obtain multi-element spectrum distribution data; An array of adaptive monitoring stations is set up at key hydrological nodes upstream and downstream of the plant to conduct real-time high-frequency dual-wavelength fluorescence spectroscopy monitoring of the water body to obtain a dynamic distribution curve of rare earth element content; Applying a self-correcting multivariate analysis algorithm to the multi-element spectrum distribution data for noise reduction, and performing quantitative conversion in combination with a standard substance calibration curve to obtain mother liquor composition data; The dynamic distribution curve of the rare earth element content is analyzed and processed through a neural network time series model to identify periodic fluctuation characteristics and trends, construct a digital twin model of water environment capacity, and obtain water quality baseline data with predictive function.

3. The method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring according to claim 2, characterized in that: The dynamic distribution curve of the rare earth element content is analyzed and processed through a neural network time series model to identify periodic fluctuation characteristics and trends, build a water environment capacity digital twin model, and obtain water quality baseline data with predictive function, including: Inputting the dynamic distribution curve of the rare earth element content into a bidirectional long short-term memory network to extract time series features, wherein the bidirectional long short-term memory network includes a three-layer hidden layer structure, and the number of memory units in each layer is twice the number of rare earth elements, to obtain a time series coding feature vector; Applying wavelet transform to the temporal coding feature vector for multi-scale decomposition, extracting the fluctuation characteristics of different frequency domains, assigning weights to each frequency component through a self-attention mechanism, and obtaining a multi-frequency domain weighted feature matrix; The multi-frequency domain weighted feature matrix is ​​input into a convolutional neural network for spatial feature learning. The convolutional neural network is composed of three convolutional layers and two pooling layers alternating with each other, and the convolution kernel sizes are 3×3, 5×5, and 7×7, respectively, to obtain the spatial distribution pattern characteristics of water quality; Based on the characteristics of the water quality spatial distribution pattern, a graph neural network model is constructed, with monitoring sites as nodes and water flow directions as edges. The pollutant migration and diffusion process is simulated through a message passing mechanism to obtain water quality baseline data with predictive function.

4. The method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring according to claim 1, characterized in that: Based on the mother liquor composition data and the water quality baseline data, the rare earth mother liquor is divided into Class A mother liquor, Class B mother liquor and Class C mother liquor, and differentiated treatment routes are adopted for different categories, including: The mother liquor composition data is input into a multivariate cluster analysis system, and an adaptive K-means clustering algorithm is used to spatially divide the three-dimensional characteristics of rare earth element content, acidity and impurity content to obtain a characteristic distribution map of the rare earth mother liquor; Applying a hierarchical decision tree algorithm to the characteristic distribution map to perform boundary optimization, and setting classification criteria in combination with the environmental capacity threshold in the water quality baseline data to obtain a mother liquor classification decision model; The rare earth mother liquor is intelligently classified in real time by the mother liquor classification decision model, and is automatically divided into high rare earth and low impurity category A, medium rare earth and medium impurity category B, and low rare earth and high impurity category C according to the total rare earth content, acidity, and impurity content parameters to obtain a classified and marked mother liquor stream; The classified and marked mother liquor flows are physically diverted through an intelligent control pipeline system, with a magnesium bicarbonate precipitation treatment route assigned to the Class A mother liquor, a DOAM-PPA extraction treatment route assigned to the Class B mother liquor, and an ion exchange treatment route assigned to the Class C mother liquor, thereby obtaining three differentiated treatment process paths.

5. The method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring according to claim 1, characterized in that: The method comprises adding magnesium bicarbonate to the Class A mother liquor for precipitation, mixing the Class B mother liquor with the filtrate, performing gradient extraction with a DOAM-PPA extractant, and dynamically adjusting process parameters according to water quality data, including: The Class A mother liquor is thermostatically controlled through a double-layer reactor to precisely adjust the temperature to 40±1° C., and a magnesium bicarbonate solution having a calculated amount of 1.8 times the molar number of rare earth ions is added to perform a precipitation reaction to obtain a rare earth carbonate precipitate, a filtrate, and recyclable CO2 gas; The filtrate is mixed with the Class B mother liquor, and the pH value is adjusted to 1.5±0.

1. An oxidant is added to the mixed liquor for pretreatment to obtain a pre-extraction liquid with a stable valence state; The pre-extraction liquid is introduced into a three-stage countercurrent extraction device, DOAM-PPA is used as an extractant, and different pH values ​​and temperature conditions are set in the first, second and third stages respectively to form a gradient extraction environment to obtain an organic phase enriched in rare earth elements; According to the real-time changes in the rare earth ion content in the water quality data, the extraction process parameters are dynamically adjusted through a multivariable adaptive control system, including adjusting the extraction time, phase ratio and pH value, while recycling magnesium salts and CO2 to obtain extraction enrichment liquid and recycled auxiliary materials.

6. The method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring according to claim 1, characterized in that: The extraction enrichment solution is mixed with the Class C mother liquor and then separated by a D151 and D301 ion exchange resin system, and then crystallized after pH graded elution to obtain a high-purity rare earth product, including: The extract enriched liquid is mixed with the Class C mother liquor in a specific ratio, a reducing agent is added to the mixed liquid for pretreatment, and the mixed liquid is filtered through a filter to obtain a clear and uniform ion exchange feed liquid; The ion exchange feed liquid is introduced into a series system consisting of the D151 and D301 ion exchange resins, and an appropriate flow rate is controlled to perform exchange adsorption to obtain a saturated resin bed enriched in rare earth elements; The saturated resin bed is subjected to pH gradient elution technology, and acidic solutions with different pH values ​​are sequentially used for graded elution to obtain rare earth enriched solutions with different compositions; The rare earth enriched solution is divided into a light rare earth group, a medium rare earth group and a heavy rare earth group, which are subjected to differentiated temperature-controlled crystallization treatments respectively, and then calcined and converted to obtain the high-purity rare earth product.

7. The method for efficiently recovering rare earth mother liquor based on surrounding water quality monitoring according to claim 6, characterized in that: The saturated resin bed is subjected to pH gradient elution technology, and acidic solutions with different pH values ​​are sequentially used for graded elution to obtain rare earth enriched solutions with different compositions, including: Pre-washing the saturated resin bed, using deionized water at a constant flow rate to flush the surface of the saturated resin bed to remove inadequately adsorbed free ions, thereby obtaining a washed and balanced resin bed system; A high pH hydrochloric acid solution is introduced into the washing and balancing resin bed system to perform the first stage elution, and the eluate collection point is precisely controlled while the online monitoring device tracks the change of the outlet rare earth ion concentration in real time to obtain a first eluate component rich in light rare earth elements; A medium pH hydrochloric acid solution is introduced into the resin bed system that has completed the first stage elution to perform the second stage elution, and the eluent flow rate is dynamically adjusted according to the change trend of the outlet liquid conductivity to obtain a second eluent component rich in medium rare earth elements; A low pH hydrochloric acid solution is introduced into the resin bed system that has completed the second stage elution for final elution. The elution efficiency is optimized by controlling the temperature and flow rate parameters to obtain a third eluate component rich in heavy rare earth elements, while achieving precise separation of the rare earth enriched liquids of different compositions.

8. A rare earth mother liquor efficient recovery system based on surrounding water quality monitoring, characterized in that: For realizing the rare earth mother liquor efficient recovery method based on surrounding water quality monitoring according to any one of claims 1 to 7, the rare earth mother liquor efficient recovery system based on surrounding water quality monitoring comprises: An analysis module is used to perform inductively coupled plasma mass spectrometry analysis on rare earth mother liquors. A water quality monitoring network is established around the plant to obtain mother liquor composition data and water quality baseline data. a processing module, configured to classify the rare earth mother liquor into Class A mother liquor, Class B mother liquor, and Class C mother liquor based on the mother liquor composition data and the water quality baseline data, and adopt differentiated processing routes for different categories; An extraction module is used to add magnesium bicarbonate to the Class A mother liquor to precipitate, mix the Class B mother liquor with the filtrate, perform gradient extraction with a DOAM-PPA extractant, and dynamically adjust process parameters based on water quality data; The separation module is used to mix the extraction enrichment liquid with the Class C mother liquor and separate them through the D151 and D301 ion exchange resin systems, and then crystallize them after pH graded elution to obtain high-purity rare earth products.

9. A rare earth mother liquor efficient recovery device based on surrounding water quality monitoring, characterized in that: It includes a memory and a processor, the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the rare earth mother liquor efficient recovery method based on surrounding water quality monitoring as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the processor is caused to execute the rare earth mother liquor efficient recovery method based on surrounding water quality monitoring according to any one of claims 1 to 7.

Images