Rare earth extraction process dynamic control system based on data acquisition

Abstract

The invention relates to the technical field of rare earth extraction process intelligent control, and particularly discloses a rare earth extraction process dynamic control system based on data acquisition, which comprises an emulsification risk real-time monitoring and dividing module, a high-risk early warning processing module and a low-risk area emulsification trend prediction and self-adaptive control module, the method comprises the following steps: collecting interfacial tension and turbidity data through a high-precision sensor, extracting features by combining empirical mode decomposition and empirical wavelet transform, constructing an emulsification trend feature value, and dividing a risk area; a support vector machine regression prediction model is established for the low-risk area data, and quantitative prediction of the emulsification development trend is achieved; and when the predicted score exceeds the limit, dynamically adjusting the stirring rate, the phase flow rate and the demulsifier addition amount by adopting a fuzzy PID algorithm to realize closed-loop optimization control of the extraction process.

Description

Technical Field

[0001] The invention relates to the technical field of intelligent control of rare earth extraction process, and in particular to a dynamic control system for rare earth extraction process control based on data acquisition. Background Art

[0002] In the rare earth extraction process, solvent extraction is a key separation and purification link and is widely used in industrial production. In this process, the organic phase and the aqueous phase are fully contacted in the mixing tank to achieve the selective transfer of metal ions, and then enter the clarification stage to complete the phase separation. However, in actual operation, due to the influence of factors such as stirring intensity, phase composition changes, and interference from impurity ions, emulsification can easily occur, making it difficult to effectively separate the two phases, thereby affecting the mass transfer efficiency, reducing the extraction efficiency, and even causing process interruptions. Therefore, how to achieve real-time monitoring, accurate identification, and dynamic regulation of the emulsification trend has become a technical problem that needs to be solved urgently to improve the stability and automation level of the rare earth extraction process.

[0003] Existing systems usually use traditional time domain or frequency domain methods for feature extraction, which makes it difficult to effectively capture the nonlinear and non-stationary characteristics of the initial emulsification signal, resulting in delayed emulsification state recognition and high misjudgment rate. In particular, in terms of trend recognition of the transformation from low emulsification risk areas to high risk areas, traditional methods cannot provide sufficient warning time and accuracy, which restricts the early intervention and adaptive adjustment of control strategies, affecting the stability and safety of the overall process operation. Summary of the invention

[0004] The object of the present invention is to provide a rare earth extraction process control dynamic control system based on data acquisition to solve the above-mentioned background problems.

[0005] The purpose of the present invention can be achieved through the following technical solutions: Dynamic control system for rare earth extraction process control based on data acquisition, including: An emulsification risk real-time monitoring and division module, which is used to collect the interfacial tension data and the turbidity data of the mixed phase in the rare earth extraction process in real time, and dynamically analyze the data in the control system. According to the analysis results, the process nodes are divided into high emulsification risk areas and low emulsification risk areas; A high-risk early warning processing module, which triggers a real-time early warning mechanism based on high emulsification risk areas; Low-risk area emulsification trend prediction and adaptive control module. The low-risk area emulsification trend prediction and adaptive control module is based on the low-emulsification risk area, extracts the abnormal characteristic value of the interfacial tension and the turbidity fluctuation characteristic value of the mixed phase in the rare earth extraction process, constructs the interfacial tension abnormal characteristic value and the turbidity fluctuation characteristic value into a comprehensive risk characteristic vector, and inputs it into the emulsification development trend prediction model for analysis to predict the future emulsification development trend. According to the prediction result, the extraction operation parameters are dynamically adjusted to realize the adaptive optimization control of the extraction process; Among them, the extraction operation parameters include: stirring rate, phase flow rate and demulsifier addition amount.

[0006] As a further scheme of the present invention: the dynamic analysis of data in the control system specifically includes: Perform trend analysis on the interfacial tension and calculate the abnormal characteristic value of the interfacial tension according to its change amplitude; Calculate the turbidity fluctuation characteristic value according to the change of the mixed phase turbidity; Perform normalization calculation on the abnormal characteristic value of the interfacial tension and the turbidity fluctuation characteristic value to obtain the emulsification trend characteristic value, which is used to divide the process nodes into high-emulsification risk areas and low-emulsification risk areas.

[0007] As a further scheme of the present invention: the process of obtaining the abnormal characteristic value of the interfacial tension is as follows: Obtain the interfacial tension time series data collected in real time during the rare earth extraction process; Perform empirical mode decomposition on the interfacial tension time series, and adaptively decompose it into multiple intrinsic mode function components and a trend residue term; Perform Hilbert transform on each intrinsic mode function component, construct an analytical signal and extract the instantaneous amplitude and instantaneous frequency information, and then calculate the energy density distribution of each intrinsic mode function; Select the first intrinsic mode functions with energy fluctuation trends, calculate the difference between the maximum and minimum values of the energy density of these intrinsic mode functions to obtain the energy density deviation value, and calculate the ratio of the energy density deviation value to the energy mean value of the selected first intrinsic mode functions to obtain the abnormal characteristic value of the interfacial tension.

[0008] As a further scheme of the present invention: the process of obtaining the turbidity fluctuation characteristic value is as follows: Obtain the mixed phase turbidity time series data collected in real time during the rare earth extraction process; Perform empirical wavelet transform decomposition on the mixed phase turbidity time series data to obtain a set of intrinsic mode components; Select multiple modal components reflecting the turbidity fluctuation characteristics and calculate the energy distribution of the selected modal components; Calculate the ratio of the energy distribution of each selected intrinsic mode component to the total energy distribution of all selected intrinsic mode components to obtain the energy ratio of each selected intrinsic mode component, and sum the energy ratios of all selected intrinsic mode components to obtain the turbidity fluctuation eigenvalue.

[0009] As a further solution of the present invention: the process nodes are divided into a high emulsification risk area and a low emulsification risk area, specifically including: Judge whether the emulsification trend eigenvalue of each process node is greater than or equal to a preset threshold. If so, it is recorded as a high emulsification risk area; if not, it is recorded as a low emulsification risk area.

[0010] As a further solution of the present invention: constructing the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue into a comprehensive risk feature vector and inputting it into the emulsification development trend prediction model for analysis, specifically including: Obtain the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue of the low emulsification risk area, construct the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue into a comprehensive risk feature vector as the input of the emulsification development trend prediction model, minimize the error between the predicted emulsification risk score and the actual emulsification risk score as the prediction target of the emulsification development trend prediction model, train the emulsification development trend model, and output the predicted emulsification risk score according to the trained emulsification development trend model. The emulsification development trend model is a support vector machine model.

[0011] As a further solution of the present invention: the training process of the emulsification development trend prediction model is as follows: In the model training stage, first extract the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue of the low-risk area from the historical data and construct them into a comprehensive risk feature vector as the model input; use the historical actual emulsification risk score as the model output label, adopt the support vector machine regression algorithm, use the radial basis function as the kernel function, adjust the penalty coefficient, kernel parameter and error tolerance to construct an optimal regression model. The goal is to minimize the error between the emulsification risk score predicted by the model and the actual score. Introduce a cross-validation strategy to optimize the hyperparameters during the training process, and use the mean square error and coefficient of determination indicators to evaluate the model performance. Deploy the trained support vector machine model to the control system to receive the collected feature data in real time and output the predicted score of the future emulsification risk, so as to realize the intelligent prediction and dynamic regulation of the emulsification phenomenon in the rare earth extraction process.

[0012] As a further solution of the present invention: predicting the future emulsification development trend, specifically including: Determine whether the future emulsification risk score is greater than or equal to a preset threshold. If so, there is a tendency for the future emulsification risk to transform from a low emulsification risk to a high emulsification risk. If not, there is no such tendency for the future emulsification risk to transform from a low emulsification risk to a high emulsification risk.

[0013] According to the prediction result, dynamically adjust the extraction operation parameters to achieve adaptive optimization control of the extraction process, specifically including: If the emulsification risk score output by the emulsification development trend prediction model is greater than or equal to the set threshold, the system automatically triggers an adaptive regulation mechanism to dynamically adjust three key operation parameters: stirring rate, phase flow rate, and demulsifier dosage. Among them, the stirring rate is stepped down according to the rising trend of the emulsification risk to reduce the interfacial disturbance and inhibit the formation of emulsion droplets; the phase flow rate is linearly adjusted according to the change trend of the interfacial tension anomaly characteristics to maintain the two-phase flow balance; the demulsifier dosage establishes a feedback control strategy based on the turbidity fluctuation characteristic value and the historical demulsification effect, and increases the dosing ratio as needed to enhance the demulsification efficiency.

[0014] Advantages of the present invention: (1) By integrating two types of key process parameters, namely the interfacial tension anomaly characteristic value and the turbidity fluctuation characteristic value, the present invention constructs an emulsification trend characteristic value with physical significance and engineering interpretability, and on this basis, realizes the dynamic identification of the emulsification state and the risk area division in the rare earth extraction process. Specifically, the system uses the empirical mode decomposition method to adaptively decompose the interfacial tension time series, extracts the energy density characteristics of each order of intrinsic mode function, and obtains the interfacial tension anomaly characteristic value by calculating the ratio of the energy fluctuation deviation to the mean value, thereby effectively capturing the change trend of the two-phase interface stability; at the same time, the empirical wavelet transform is used to perform frequency-domain mode decomposition on the turbidity signal, select the mode component reflecting the main characteristics of the mixed-phase turbidity fluctuation, and calculate the turbidity fluctuation characteristic value based on its energy distribution ratio, further enhancing the sensitivity to the tiny disturbance in the initial stage of emulsification. After normalizing the above two characteristic values and fusing them with a preset ratio, a comprehensive emulsification trend characteristic value is formed as the basis for judging the emulsification risk state of the process node. This method not only overcomes the lag and uncertainty problems brought by traditional reliance on single parameter or manual experience judgment, but also improves the accuracy, sensitivity and real-time response ability of emulsification risk identification through the introduction of multi-source feature fusion and advanced signal processing means, providing a solid data support and decision-making basis for the subsequent early warning and control strategy implementation.

[0015] (2) The present invention constructs an emulsion development trend prediction model based on support vector machine regression, fully excavating the non-linear mapping relationship between the abnormal eigenvalue of interfacial tension and the turbidity fluctuation eigenvalue in the historical data of the low-emulsion-risk region, so as to achieve high-precision prediction of the future emulsion risk score. In the model training stage, the system takes the extracted comprehensive risk feature vector as the input variable, and the actual emulsion risk score generated by expert evaluation or historical records as the output label, uses the radial basis function as the kernel function, and optimizes hyperparameters such as the penalty coefficient kernel parameter and the error tolerance through a cross-validation strategy to construct a regression model with strong generalization ability. The model aims to minimize the mean square error between the predicted score and the actual score, and can accurately capture the development direction of the emulsion trend under complex working conditions.

[0016] When the emulsion risk score predicted by the model approaches or exceeds the set threshold, the system automatically triggers an adaptive control mechanism, and combines the fuzzy PID control algorithm to coordinately optimize and adjust key operating parameters such as the stirring rate, the phase flow rate, and the demulsifier addition amount. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below with reference to the accompanying drawings.

[0018] Figure 1 It is a flow block diagram of a dynamic control system for rare earth extraction process based on data acquisition according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Please refer to Figure 1 As shown, the present invention is a dynamic control system for rare earth extraction process based on data acquisition, including: An emulsion risk real-time monitoring and classification module, which is used to collect the interfacial tension data and the turbidity data of the mixed phase in the rare earth extraction process in real time, and dynamically analyze the data in the control system. According to the analysis results, the process nodes are divided into high-emulsion-risk regions and low-emulsion-risk regions; A high-risk early warning processing module, which triggers a real-time early warning mechanism based on the high-emulsion-risk region; Low-risk area emulsification trend prediction and adaptive control module. The low-risk area emulsification trend prediction and adaptive control module is based on the low-emulsification risk area, extracts the interfacial tension abnormal characteristic value and turbidity fluctuation characteristic value of the mixed phase in the rare earth extraction process, constructs the interfacial tension abnormal characteristic value and turbidity fluctuation characteristic value into a comprehensive risk characteristic vector, and inputs it into the emulsification development trend prediction model for analysis to predict the future emulsification development trend. According to the prediction result, the extraction operation parameters are dynamically adjusted to achieve the adaptive optimization control of the extraction process; Among them, the extraction operation parameters include: stirring rate, phase flow rate, and demulsifier addition amount.

[0021] In the emulsification risk real-time monitoring and division module, the interfacial tension data and turbidity data of the mixed phase in the rare earth extraction process are collected in real time, and the data is dynamically analyzed in the control system. According to the analysis result, the process nodes are divided into high-emulsification risk areas and low-emulsification risk areas, specifically including: In the rare earth extraction process, the real-time collection of interfacial tension data is realized by a high-precision interfacial tension sensor installed in the mixing tank. The interfacial tension sensor adopts an online measurement device based on the capillary rise method, which can continuously obtain the time series signal of the interfacial tension between the two phases changing with time; the sensor is communicatively connected to the control system, converts the collected analog signal into a digital signal, and performs filtering and temperature compensation processing to improve the measurement accuracy and stability.

[0022] The turbidity data of the mixed phase is monitored online by an optical turbidity detector set at the outlet end of the mixed phase. The turbidity detector measures the light scattering intensity caused by suspended particles in the mixed liquid by the scattered light method (such as a 90° scattering angle) to obtain the dynamic response curve of turbidity changing with time; the collected turbidity signal is transmitted to the control system through a signal amplification module and an A / D conversion module for subsequent feature extraction and emulsification state evaluation.

[0023] Perform trend analysis on the interfacial tension and calculate the interfacial tension abnormal characteristic value according to its change amplitude; Calculate the turbidity fluctuation characteristic value according to the change of the turbidity of the mixed phase; Perform normalization calculation processing on the interfacial tension abnormal characteristic value and turbidity fluctuation characteristic value to obtain the emulsification trend characteristic value, which is used to divide the process nodes into high-emulsification risk areas and low-emulsification risk areas; Among them, the calculation expression of the emulsification trend characteristic value is: ; In the formula, represents the emulsification trend characteristic value, represents the interfacial tension abnormal characteristic value, represents the turbidity fluctuation characteristic value, and represents a preset proportionality coefficient, and and are both greater than 0.

[0024] The process of obtaining the interfacial tension abnormal eigenvalue is as follows: Obtain the interfacial tension time series data collected in real time during the rare earth extraction process; Perform empirical mode decomposition on the interfacial tension time series, and adaptively decompose it into multiple intrinsic mode function components and a trend residue term; Perform Hilbert transform on each intrinsic mode function component, construct an analytic signal and extract instantaneous amplitude and instantaneous frequency information, and then calculate the energy density distribution of each intrinsic mode function; Among them, the calculation expression of the analytic signal is: ; Among them, the calculation expression of the instantaneous amplitude is: ; Among them, the calculation expression of the instantaneous frequency is: ; Among them, the calculation expression of the energy density is: ; In the formula, represents the analytic signal of the th intrinsic mode function, represents the instantaneous amplitude of the th intrinsic mode function, represents the complex exponential function, represents the imaginary unit, represents the th instantaneous phase of the intrinsic mode function, represents the real part signal of the th intrinsic mode function, is the Hilbert transform result of the th intrinsic mode function, representing the orthogonal component of the signal, represents the number of intrinsic mode functions, represents the derivative of the instantaneous phase with respect to time ; represents time, represents the th instantaneous frequency of the intrinsic mode function; Select the first intrinsic mode functions with the main energy fluctuation trend, calculate the difference between the maximum and minimum values of the energy density of these intrinsic mode functions to obtain the energy density deviation value, and use the energy density deviation value and the selected first The ratio of the energy mean values of the intrinsic mode functions is calculated to obtain the interfacial tension anomaly eigenvalue.

[0025] The process of obtaining the turbidity fluctuation eigenvalue is as follows: Obtain the turbidity time series data of the mixed phase collected in real time during the rare earth extraction process; Perform empirical wavelet transform decomposition on the turbidity time series data of the mixed phase to obtain a set of intrinsic mode components. The calculation expression is: ; In the formula, represents the th intrinsic mode component, represents the inverse Fourier transform, represents the spectrum of the turbidity time series data of the mixed phase, represents the th filter function, represents time, represents the frequency variable; Select multiple mode components that reflect the main fluctuation characteristics of turbidity, and calculate the energy distribution of the selected mode components. The calculation expression is: ; In the formula, represents the energy distribution of the th selected intrinsic mode component, represents the th selected intrinsic mode component; Respectively calculate the ratio of the energy distribution of each selected intrinsic mode component to the total energy distribution of all selected intrinsic mode components to obtain the energy ratio of each selected intrinsic mode component. Sum up the energy ratios of all selected intrinsic mode components to obtain the turbidity fluctuation eigenvalue.

[0026] The division of the process nodes into high emulsification risk areas and low emulsification risk areas specifically includes: Judge whether the emulsification trend eigenvalue of each process node is greater than or equal to the preset threshold. If so, it is recorded as a high emulsification risk area; if not, it is recorded as a low emulsification risk area.

[0027] In the high-risk early warning processing module, the high-risk early warning processing module triggers a real-time early warning mechanism based on the high emulsification risk area, specifically including: In the high-risk early warning processing module, when the emulsification trend eigenvalue is greater than or equal to the threshold, the system determines that the current process node is in the high emulsification risk area and triggers a real-time early warning mechanism; the control system issues a warning signal through the sound and light alarm device and pops up a corresponding early warning prompt window on the human-machine interface.

[0028] It should be noted that this module realizes the dynamic recognition and precise classification of the emulsification state of the process node by collecting the interfacial tension and turbidity data of the mixed phase in the rare earth extraction process in real time, and extracting the abnormal characteristic value of the interfacial tension and the fluctuation characteristic value of the turbidity based on the empirical mode decomposition and empirical wavelet transform technologies, respectively, and then constructing a normalized emulsification trend characteristic value. This module fully combines the advantages of signal processing and multi-parameter fusion analysis, improving the accuracy and real-time performance of emulsification risk judgment. Compared with the traditional method that relies on manual experience or single-parameter evaluation, the present invention can capture the change of emulsification trend earlier and more sensitively. Especially, an automatic warning mechanism is introduced based on the determination of the high emulsification risk area. Through the linkage of the control system with the sound and light alarm and the human-machine interface prompt, the safety and intelligent level of the rare earth extraction process are significantly improved, and it has good engineering application prospects and innovation value.

[0029] In the emulsification trend prediction and adaptive control module in the low-risk area, based on the low emulsification risk area, the abnormal characteristic value of the interfacial tension and the fluctuation characteristic value of the turbidity of the mixed phase in the rare earth extraction process are extracted, and the abnormal characteristic value of the interfacial tension and the fluctuation characteristic value of the turbidity are constructed into a comprehensive risk characteristic vector and input into the emulsification development trend prediction model for analysis to predict the future emulsification development trend. According to the prediction result, the extraction operation parameters are dynamically adjusted to realize the adaptive optimization control of the extraction process, specifically including: Obtain the abnormal characteristic value of the interfacial tension and the fluctuation characteristic value of the turbidity in the low emulsification risk area, construct the abnormal characteristic value of the interfacial tension and the fluctuation characteristic value of the turbidity into a comprehensive risk characteristic vector as the input of the emulsification development trend prediction model, and use the minimization of the error between the predicted emulsification risk score and the actual emulsification risk score as the prediction target of the emulsification development trend prediction model to train the emulsification development trend model. According to the trained emulsification development trend model, output the predicted emulsification risk score, and the emulsification development trend model is a support vector machine model.

[0030] The training process of the emulsification development trend prediction model is as follows: During the model training phase, first, extract the interfacial tension anomaly eigenvalue and turbidity fluctuation eigenvalue of the low-risk area from historical data, combine these two features into a comprehensive risk feature vector as the model input; use the historical actual emulsification risk score as the model output label, adopt the support vector machine regression algorithm, with the radial basis function as the kernel function, and construct the optimal regression model by adjusting the penalty coefficient, kernel parameter, and error tolerance. The goal is to minimize the error between the emulsification risk score predicted by the model and the actual score. Introduce the cross-validation strategy to optimize the hyperparameters during the training process, and use the mean squared error and coefficient of determination metrics to evaluate the model performance. Deploy the trained support vector machine model to the control system, receive the collected feature data in real time, and output the predicted score of the future emulsification risk, so as to realize the intelligent prediction and dynamic regulation of the emulsification phenomenon in the rare earth extraction process.

[0031] The prediction of the future emulsification development trend specifically includes: Judge whether the future emulsification risk score is greater than or equal to the preset threshold. If so, there is a trend that the future emulsification risk may transform from low emulsification risk to high emulsification risk. If not, there is no such trend that the future emulsification risk transforms from low emulsification risk to high emulsification risk.

[0032] According to the prediction result, dynamically adjust the extraction operation parameters to achieve the adaptive optimization control of the extraction process, specifically including: If the emulsification risk score output by the emulsification development trend prediction model is greater than or equal to the set threshold, the system automatically triggers the adaptive regulation mechanism to dynamically adjust the three key operation parameters of the stirring rate, phase flow rate, and demulsifier addition amount respectively; among them, the stirring rate is stepped down according to the rising trend of the emulsification risk to reduce the interfacial disturbance and inhibit the formation of emulsion droplets; the phase flow rate is linearly adjusted according to the change trend of the interfacial tension anomaly feature to maintain the two-phase flow balance; the demulsifier addition amount establishes a feedback control strategy based on the turbidity fluctuation eigenvalue and the historical demulsification effect, and increases the dosing ratio as needed to enhance the demulsification efficiency.

[0033] During the dynamic adjustment process, the system uses the fuzzy PID control algorithm to co-optimize and control the above three operation parameters, takes the emulsification risk score deviation and its change rate as input variables, combines the preset empirical control rule base, generates the corresponding control output signal, and drives the actuator to precisely adjust the rotation speed of the stirring motor, the frequency of the delivery pump, and the dosing rate of the metering pump, so as to realize the real-time response and process stable control of the emulsification phenomenon in the rare earth extraction process, and improve the intelligent level and operation stability of the overall production process.

[0034] Working principle of the present invention: It aims to achieve real-time monitoring, risk identification and intelligent regulation of emulsification phenomena during the extraction process. The system, through the emulsification risk real-time monitoring and classification module, uses a high-precision interfacial tension sensor and an optical turbidity detector to online collect key parameters of the mixed phase, and combines empirical mode decomposition and empirical wavelet transform technologies to extract interfacial tension abnormal eigenvalue and turbidity fluctuation eigenvalue respectively. Further, through fusion and normalization processing, an emulsification trend eigenvalue is obtained. According to the set threshold, the process nodes are divided into high and low emulsification risk areas; for the high-risk area, the system triggers an audible and visual alarm and a human-machine interface prompt to achieve immediate early warning; for the low-risk area, the low-risk area emulsification trend prediction and adaptive control module constructs an emulsification development trend prediction model based on support vector machine regression, uses the comprehensive risk feature vector as the input, predicts the future emulsification risk score, and accordingly dynamically adjusts key operating parameters such as stirring rate, phase flow rate and demulsifier addition amount, and adopts a fuzzy PID control algorithm to achieve multi-parameter collaborative optimization regulation, so as to achieve the integrated control goal of intelligent identification, trend prediction and adaptive regulation of emulsification phenomena in the rare earth extraction process, significantly improving the process stability, safety and intelligent level, and having good industrial application value and technological innovation.

[0035] The above has described in detail an embodiment of the present invention, but the content described is only the preferred embodiment of the present invention and cannot be considered as limiting the implementation scope of the present invention. All equivalent changes and improvements made according to the scope of the present invention application shall still fall within the scope covered by the patent of the present invention.

Claims

1. A dynamic control system for rare earth extraction process based on data acquisition, characterized in that, Including: An emulsification risk real-time monitoring and classification module, which is used to collect the interfacial tension data and turbidity data of the mixed phase in the rare earth extraction process in real time, dynamically analyze the data in the control system, and divide the process nodes into high emulsification risk areas and low emulsification risk areas according to the analysis results; A high-risk early warning processing module, which triggers a real-time early warning mechanism based on the high emulsification risk area; A low-risk area emulsification trend prediction and adaptive control module, which extracts the abnormal characteristic value of the interfacial tension and the turbidity fluctuation characteristic value of the mixed phase in the rare earth extraction process based on the low emulsification risk area, constructs a comprehensive risk characteristic vector from the abnormal characteristic value of the interfacial tension and the turbidity fluctuation characteristic value, and inputs it into the emulsification development trend prediction model for analysis, predicts the future emulsification development trend, and dynamically adjusts the extraction operation parameters according to the prediction results to achieve the adaptive optimization control of the extraction process; Among them, the extraction operation parameters include: stirring rate, phase flow rate, and demulsifier addition amount.

2. The dynamic control system for rare earth extraction process based on data acquisition according to claim 1, characterized in that, The dynamic analysis of the data in the control system specifically includes: Performing a trend analysis on the interfacial tension and calculating the abnormal characteristic value of the interfacial tension according to its change amplitude; Calculating the turbidity fluctuation characteristic value according to the change of the turbidity of the mixed phase; Performing a normalization calculation process on the abnormal characteristic value of the interfacial tension and the turbidity fluctuation characteristic value to obtain the emulsification trend characteristic value, which is used to divide the process nodes into high emulsification risk areas and low emulsification risk areas.

3. The dynamic control system for rare earth extraction process based on data acquisition according to claim 2, characterized in that The process of obtaining the abnormal characteristic value of the interfacial tension is as follows: Obtaining the interfacial tension time series data collected in real time during the rare earth extraction process; Performing empirical mode decomposition on the interfacial tension time series, and adaptively decomposing it into multiple intrinsic mode function components and a trend residue term; Performing Hilbert transform on each intrinsic mode function component, constructing an analytic signal and extracting the instantaneous amplitude and instantaneous frequency information, and then calculating the energy density distribution of each intrinsic mode function; Select the first few intrinsic mode functions of the energy fluctuation trend, calculate the difference between the maximum and minimum values of the energy density of these intrinsic mode functions to obtain the energy density deviation value, and calculate the ratio of the energy density deviation value to the average energy of the first few intrinsic mode functions selected to obtain the interfacial tension anomaly eigenvalue.

4. The dynamic control system for rare earth extraction process based on data acquisition according to claim 2, wherein The process of obtaining the turbidity fluctuation characteristic value is as follows: Obtaining the mixed phase turbidity time series data collected in real time during the rare earth extraction process; Performing empirical wavelet transform decomposition on the mixed phase turbidity time series data to obtain a set of intrinsic mode components; Selecting multiple mode components reflecting the turbidity fluctuation characteristics and calculating the energy distribution of the selected mode components; Calculating the ratio of the energy distribution of each selected intrinsic mode component to the total energy distribution of all selected intrinsic mode components respectively to obtain the energy ratio of each selected intrinsic mode component, and summing up the energy ratios of all selected intrinsic mode components to obtain the turbidity fluctuation characteristic value.

5. The dynamic control system for rare earth extraction process based on data acquisition according to claim 1, wherein The process of dividing the process nodes into high emulsification risk areas and low emulsification risk areas specifically includes: Judging whether the emulsification trend characteristic value of each process node is greater than or equal to a preset threshold. If so, it is recorded as a high emulsification risk area; if not, it is recorded as a low emulsification risk area.

6. The dynamic control system for rare earth extraction process based on data acquisition according to claim 1, wherein Constructing the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue into a comprehensive risk feature vector and inputting it into the emulsion development trend prediction model for analysis, specifically including: Obtaining the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue of the low-emulsion-risk region, constructing the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue into a comprehensive risk feature vector as the input of the emulsion development trend prediction model, minimizing the error between the predicted emulsion risk score and the actual emulsion risk score as the prediction target of the emulsion development trend prediction model, training the emulsion development trend model, and according to the trained emulsion development trend model, outputting the predicted emulsion risk score. The emulsion development trend model is a support vector machine model.

7. The dynamic control system for rare earth extraction process based on data acquisition according to claim 6, characterized in that The training process of the emulsion development trend prediction model is as follows: In the model training stage, first extract the interfacial tension anomaly eigenvalue and the turbidity fluctuation eigenvalue of the low-risk region from the historical data and construct them into a comprehensive risk feature vector as the model input; use the historical actual emulsion risk score as the model output label, adopt the support vector machine regression algorithm, use the radial basis function as the kernel function, and construct the optimal regression model by adjusting the penalty coefficient, the kernel parameter and the error tolerance. The goal is to minimize the error between the emulsion risk score predicted by the model and the actual score. Introduce the cross-validation strategy to optimize the hyperparameters during the training process, and use the mean square error and the coefficient of determination to evaluate the model performance. Deploy the trained support vector machine model to the control system to receive the collected feature data in real time and output the predicted score of the future emulsion risk, so as to realize the intelligent prediction and dynamic regulation of the emulsion phenomenon in the rare earth extraction process.

8. The dynamic control system for rare earth extraction process based on data acquisition according to claim 1, characterized in that, Predicting the future emulsion development trend specifically includes: Judging whether the future emulsion risk score is greater than or equal to the preset threshold. If so, there is a tendency for the future emulsion risk to transform from low emulsion risk to high emulsion risk. If not, there is no tendency for the future emulsion risk to transform from low emulsion risk to high emulsion risk.

9. The dynamic control system for rare earth extraction process based on data acquisition according to claim 1, characterized in that, According to the prediction result, dynamically adjusting the extraction operation parameters to achieve the adaptive optimization control of the extraction process, specifically including: If the emulsion risk score output by the emulsion development trend prediction model is greater than or equal to the set threshold, the system automatically triggers the adaptive regulation mechanism to dynamically adjust the three key operation parameters of the stirring rate, the phase flow rate and the demulsifier addition amount respectively; among them, the stirring rate is gradually reduced according to the rising trend of the emulsion risk to reduce the interfacial disturbance and inhibit the formation of emulsion droplets; the phase flow rate is linearly adjusted according to the change trend of the interfacial tension anomaly feature to maintain the two-phase flow balance; the demulsifier addition amount establishes a feedback control strategy based on the turbidity fluctuation eigenvalue and the historical demulsification effect, and increases the dosing ratio as needed to enhance the demulsification efficiency.

Images