Copper extraction effect analysis method and system

Abstract

The invention discloses a copper extraction effect analysis method and system, and the method comprises the steps: carrying out the continuous scanning of the surface of a copper sample through an optical sensor, and obtaining the chromaticity change time sequence data in a refining process; and carrying out segmentation processing on the data by utilizing a preset threshold value, and extracting dynamic change trend characteristics and key point locations. And in combination with conductivity and hardness value data recorded by a real-time monitoring system, a polynomial regression model is constructed, and correlation between copper purity and physical characteristics is established. And through historical data calibration and mean filtering optimization, the mapping relation between chromaticity change and copper purity is determined. And finally, according to the optimized chromaticity characteristic value and the calibrated correlation model, calculating the quantitative score of the real-time refining effect, and judging the grading basis of the copper purity. According to the method, the real-time evaluation of the copper purity and the dynamic optimization of the refining process are realized, and the refining efficiency and the product quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of metal extraction, and particularly discloses a method and system for analyzing the effect of copper extraction. Background Art

[0002] Copper extraction and refining technology, as a core area of modern metallurgical industry, is of irreplaceable importance in promoting the development of materials science, electronics industry and energy technology. The high purity of copper is not only the key to enhancing its electrical conductivity, thermal conductivity and corrosion resistance, but also the cornerstone for meeting the requirements of high-end applications. However, there are still many problems in the current effect evaluation and optimization during the copper refining process, and technical breakthroughs are urgently needed to meet the growing industrial demands. Existing methods for analyzing the effect of copper extraction mostly rely on offline detection means, such as chemical titration or spectral analysis. Although these methods are accurate, they are time-consuming, unable to achieve real-time monitoring, and difficult to comprehensively reflect the dynamic changes of copper during the refining process. In addition, traditional physical property tests are usually carried out independently, lacking a systematic correlation with copper purity, resulting in low evaluation efficiency of the refining effect and difficulty in meeting the real-time requirements of large-scale production.

[0003] In this context, the core challenges faced by copper refining technology have gradually emerged. First, as an intuitive indicator during the refining process, the real-time monitoring technology of chromaticity change is not yet mature, and it is difficult to accurately capture the corresponding relationship between minute changes and purity. Second, although physical parameters such as conductivity and hardness are closely related to copper purity, existing research lacks effective correlation models and cannot quickly infer the refining effect through these indirect indicators. Finally, the index grading of refined copper still mainly relies on manual experience, with insufficient intelligence, resulting in poor consistency of grading results and difficulty in meeting the requirements of fine production. The unresolved technical factors directly lead to the unique problems of low control precision, low efficiency in the refining process, and large fluctuations in product quality.

[0004] Therefore, how to achieve real-time monitoring of chromaticity change during the copper refining process, establish a precise correlation model between copper purity and physical properties such as conductivity and hardness, and on this basis, conduct intelligent grading of various indicators of refined copper has become the key issue for improving the analysis efficiency of copper extraction effect and the stability of refining quality. The solution to this problem will provide important support for the intelligent and real-time development of copper refining technology. Summary of the Invention

[0005] The present invention provides a method and system for analyzing the effect of copper extraction, aiming to solve at least one defect existing in the existing methods for analyzing the effect of copper extraction.

[0006] One aspect of the present invention relates to a method for analyzing the effect of copper extraction, including the following steps: Obtain the chromaticity change data of the copper sample during the refining process, continuously scan the copper surface through an optical sensor, and obtain the time series data of chromaticity change; According to the time series data of chromaticity change, use a preset threshold to segment the copper extraction data and judge the trend characteristics of dynamic change; Extract key points from the trend characteristics of dynamic change, and record the conductivity and hardness value data at the corresponding time points through a real-time monitoring system; For the extracted conductivity and hardness value data, construct a correlation model, use the polynomial regression algorithm to calculate the functional relationship between copper purity and physical properties, and obtain preliminary correlation parameters; After obtaining the preliminary correlation parameters, calibrate the correlation model through historical refining effect data to determine the mapping coefficient between chromaticity change and copper purity; If the mapping coefficient exceeds the preset range, perform secondary filtering on the chromaticity change data, use the mean filtering algorithm to eliminate noise interference, and obtain the optimized chromaticity characteristic value; According to the optimized chromaticity characteristic value and the calibrated correlation model, calculate the quantization score of the real-time refining effect and judge the grading basis of copper purity.

[0007] Furthermore, the steps of obtaining the chromaticity change data of the copper sample during the refining process and continuously scanning the copper surface through an optical sensor to obtain the time series data of chromaticity change include: Use a preset threshold to judge the eigenvalue sequence. If the eigenvalue exceeds the threshold, determine the abnormal chromaticity change point; For the abnormal chromaticity change point, obtain the corresponding time series segment to obtain the time distribution of the abnormal process; Analyze the fluctuations in the refining process through the time distribution and judge the change trend in process monitoring; According to the change trend, use the support vector machine algorithm to classify the refining process and determine the process status category; Extract the key time points from the process status category to obtain the basis for optimizing and adjusting the refining process.

[0008] Furthermore, the steps of using a preset threshold to segment the copper extraction data according to the time series data of chromaticity change and judging the trend characteristics of dynamic change include: Obtain the time series data of chromaticity change, use a preset threshold to segment the data, and obtain the segmented data units; For the segmented data units, use the sequence analysis method to determine the chromaticity change pattern within each segment; Detect the occurrence points of dynamic change through the chromaticity change pattern to obtain the change detection result; If the change detection result exceeds the preset threshold, corresponding trend features are extracted in combination with the judgment logic; Based on the extracted trend features, the K-means algorithm is used to cluster the features to obtain the feature classification result; Obtain the feature classification result, and judge the trend direction of the dynamic change through the continuity analysis of the time series; For the trend direction, the change detection result is fused to determine the final description of the trend feature.

[0009] Furthermore, the steps of extracting key points from the trend features of the dynamic change and recording the conductivity and hardness value data at the corresponding time points through the real-time monitoring system include: Obtain the conductivity value and hardness value data corresponding to the key points through the real-time monitoring system to obtain the time series record; For the time series record, a feature extraction method is used to determine the analysis result of the dynamic change points; If the analysis result of the points shows dynamic change, the directionality of the change is judged through the trend feature to obtain the directionality description; According to the directionality description, obtain the fluctuation range of the conductivity value and hardness value, and determine the fluctuation feature; For the fluctuation feature, the K-means algorithm is used to cluster the key points to obtain the point classification result; Through the point classification result, fuse the continuity analysis of the time series to judge the stability of the trend feature and obtain the final description; Extract the significant features of the dynamic change from the final description to determine the distribution law of the key points.

[0010] Furthermore, for the extracted conductivity and hardness value data, the steps of constructing an association model and using the polynomial regression algorithm to calculate the functional relationship between copper purity and physical properties to obtain the preliminary association parameters include: For the conductivity value and hardness value data, use the polynomial regression algorithm to construct the functional relationship between copper purity and physical properties to obtain the association parameters; Through the association parameters, calculate the fluctuation range between copper purity and conductivity value, and determine the fluctuation feature; According to the fluctuation feature, analyze the change trend between physical properties and hardness value to obtain the trend description; If the trend description shows that the change directions are the same, fuse the distribution characteristics of the conductivity value and hardness value to obtain the distribution law; Through the distribution law, use the K-means algorithm to cluster copper purity and physical properties to determine the classification result; According to the classification result, obtain the stability feature of the functional relationship and judge the stability description; For the stability description, extract the significance distribution of physical properties and determine the significance law.

[0011] Further, after obtaining the preliminary correlation parameters, the steps of calibrating the correlation model with historical refining effect data and determining the mapping coefficient between chromaticity change and copper purity include: Obtain the distribution characteristics of the refining effect from historical data and determine the distribution law; Extract the fluctuation range of chromaticity change according to the distribution law to obtain the change characteristics; Fuse the distribution characteristics of copper purity for the change characteristics to determine the purity distribution; Adjust the correlation parameters through the purity distribution to obtain the parameter adjustment result; If the parameter adjustment result is consistent with the historical data, calibrate the model using the linear regression algorithm to obtain the mapping coefficient; Analyze the corresponding relationship between chromaticity change and copper purity according to the mapping coefficient to determine the corresponding characteristics; Verify the stability of the model calibration through the corresponding characteristics and judge the stability description.

[0012] Further, if the mapping coefficient exceeds the preset range, the steps of performing secondary filtering on the chromaticity change data and using the mean filtering algorithm to remove noise interference to obtain the optimized chromaticity characteristic value include: If the mapping coefficient exceeds the preset range, process the chromaticity change data using the mean filtering algorithm to obtain the optimized characteristic value by removing noise interference; Determine the fluctuation range of chromaticity change through the optimized characteristic value and obtain the distribution characteristics within the fluctuation range; Judge the stability of chromaticity change according to the distribution characteristics to obtain the stability description result; If the stability description result meets the preset threshold, process the change data using the smoothing algorithm to obtain the smoothed data sequence; Obtain the trend characteristics of chromaticity change through the smoothed data sequence and determine the distribution law of the trend characteristics; Adjust the mapping coefficient according to the distribution law of the trend characteristics to obtain the adjusted mapping parameter; Verify the optimized characteristics of chromaticity change through the adjusted mapping parameter and judge the matching degree of the optimized characteristics.

[0013] Further, according to the optimized chromaticity characteristic value and the calibrated correlation model, the steps of calculating the quantization score of the real-time refining effect and judging the grading basis of copper purity include: Calculate the quantization score through the optimized chromaticity characteristic and the calibrated correlation model to obtain the preliminary result of the real-time refining effect; Compare the quantization score with a preset threshold. If it exceeds the threshold, process the data input using a mean filtering algorithm to obtain a smoothed feature sequence; Analyze the distribution characteristics of copper purity through the smoothed feature sequence to determine the preliminary range of the classification basis; Match the distribution characteristics within the preliminary range with the correlation model to obtain the calibrated parameters after matching; Adjust the real-time refining process according to the calibrated parameters after matching to obtain the adjusted refined data; Calculate the final quantization score through the adjusted refined data and the feature analysis result to judge the classification result of copper purity.

[0014] Another aspect of the present invention relates to a copper extraction effect analysis system applied to the above copper extraction effect analysis method. The copper extraction effect analysis system includes: A first acquisition module for acquiring the chromaticity change data of the copper sample during the refining process, continuously scanning the copper surface through an optical sensor to obtain the time series data of chromaticity change; A first judgment module for segmenting the copper extraction data using a preset threshold according to the time series data of chromaticity change to judge the trend characteristics of dynamic change; A recording module for extracting key points from the trend characteristics of dynamic change and recording the conductivity and hardness value data at the corresponding time points through a real-time monitoring system; A second acquisition module for constructing a correlation model for the extracted conductivity and hardness value data, using a polynomial regression algorithm to calculate the functional relationship between copper purity and physical properties to obtain preliminary correlation parameters; A determination module for calibrating the correlation model through historical refining effect data after obtaining the preliminary correlation parameters to determine the mapping coefficient between chromaticity change and copper purity; A third acquisition module for, if the mapping coefficient exceeds the preset range, performing secondary filtering on the chromaticity change data, using a mean filtering algorithm to remove noise interference to obtain optimized chromaticity feature values; A second judgment module for calculating the quantization score of the real-time refining effect according to the optimized chromaticity feature values and the calibrated correlation model to judge the classification basis of copper purity.

[0015] Further, the first acquisition module includes: A first determination unit for judging the eigenvalue sequence using a preset threshold. If the eigenvalue exceeds the threshold, determine the abnormal chromaticity change point; A first acquisition unit for, for the abnormal chromaticity change point, acquiring the corresponding time series segment to obtain the time distribution of the abnormal process; A judgment unit, configured to analyze the fluctuations in the refining process through time distribution analysis and judge the change trend in process monitoring; A second determination unit, configured to classify the refining process by using a support vector machine algorithm according to the change trend and determine the process status category; A second acquisition unit, configured to extract key time points from the process status category to obtain a basis for optimizing and adjusting the refining process.

[0016] The beneficial effects achieved by the present invention are as follows: The present invention provides a method and system for analyzing the effect of copper extraction. By continuously scanning the surface of a copper sample through an optical sensor, time series data of chromaticity changes in the refining process is obtained. The data is segmented by using a preset threshold to extract dynamic change trend features and key points. Combining the conductivity and hardness value data recorded by a real-time monitoring system, a polynomial regression model is constructed to establish the correlation between copper purity and physical properties. Through historical data calibration and mean filtering optimization, the mapping relationship between chromaticity change and copper purity is determined. Finally, according to the optimized chromaticity characteristic values and the calibrated correlation model, a quantitative score of the real-time refining effect is calculated to judge the grading basis of copper purity. The present invention realizes the real-time evaluation of copper purity and the dynamic optimization of the refining process, improving the refining efficiency and product quality. Description of the Drawings

[0017] Figure 1 It is a schematic flowchart of an embodiment of the method for analyzing the effect of copper extraction according to the present invention. Detailed Embodiment

[0018] In order to better understand the above technical solutions, the above technical solutions will be described in detail below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0019] As Figure 1 shown, a first embodiment of the present invention proposes a method for analyzing the effect of copper extraction, including the following steps: Step S100: Obtain the chromaticity change data of the copper sample in the refining process, and continuously scan the copper surface through an optical sensor to obtain time series data of chromaticity changes.

[0020] The chromaticity change data of the copper sample includes hue change data and saturation difference data. Hue change refers to the change in the surface or solution color of the copper sample after different treatments (such as oxidation, alloying, or chemical reaction) (such as metallic luster → oxidized black / blue-green). Saturation difference is the change in the depth of color. For example, the red color of pure copper may turn into brown-black after oxidation, and the saturation is significantly reduced.

[0021] The time - series data of chromaticity change is a dataset that records the dynamic changes of color attributes over time, and is usually used to analyze the fluctuation laws of color parameters (such as hue, saturation, brightness, etc.) at continuous time points.

[0022] Step S200: According to the time - series data of chromaticity change, segment the copper extraction data using a preset threshold, and judge the trend characteristics of the dynamic change.

[0023] In the copper extraction production process, segmented data processing is a key technical means to achieve process optimization and efficiency improvement, mainly carried out around the process flow characteristics and data characteristics.

[0024] The trend characteristics of dynamic change refer to the regular patterns that continuously evolve over time or space.

[0025] Step S300: Extract key points from the trend characteristics of dynamic change, and record the conductivity and hardness value data at the corresponding time points through a real - time monitoring system.

[0026] Electrical Conductivity is a physical quantity that measures the copper's ability to conduct electricity, defined as the ability of a material to transmit current under a unit electric field strength, and numerically is the reciprocal of resistivity.

[0027] Hardness reflects the copper's ability to resist plastic deformation (such as indentation, scratching), and is directly related to the mechanical strength of the material.

[0028] Step S400: For the extracted conductivity and hardness value data, construct a correlation model, use the polynomial regression algorithm to calculate the functional relationship between copper purity and physical properties, and obtain preliminary correlation parameters.

[0029] A correlation model refers to a systematic method that quantitatively correlates key parameters (such as current density, electrolyte composition, temperature, etc.) in the copper extraction process with the final products (such as copper purity, precipitation efficiency) by establishing a mathematical model or theoretical framework. Its core goal is to optimize process conditions and predict extraction effects.

[0030] Step S500: After obtaining the preliminary correlation parameters, calibrate the correlation model through historical refining effect data to determine the mapping coefficient between chromaticity change and copper purity.

[0031] Historical refining effect data refers to the quantitative records of process efficiency, product purity, and resource utilization rate accumulated through long - term practice and technological iteration in the copper smelting and refining process.

[0032] The mapping coefficient between chromaticity change and copper purity refers to the mathematical relationship that characterizes the change in copper material purity through the quantitative change of chromaticity parameters (such as hue, saturation).

[0033] Step S600: If the mapping coefficient exceeds the preset range, perform secondary filtering on the chromaticity change data, and use the mean filtering algorithm to eliminate noise interference to obtain the optimized chromaticity eigenvalue.

[0034] Secondary filtering refers to a technique that performs noise suppression, feature enhancement, or frequency-selective processing on signals or data through two consecutive or nested filtering operations (or a filter with a second-order transfer function). Its core goal is to improve signal quality or extract target information. It is widely used in industrial inspection, communication, image processing, and chemical analysis (such as copper purity detection).

[0035] Step S700: According to the optimized chromaticity eigenvalue and the calibrated correlation model, calculate the quantization score of the real-time refining effect to determine the grading basis of copper purity.

[0036] The quantization score of the real-time refining effect refers to a comprehensive evaluation value generated by combining dynamic algorithms through real-time collection of refining process parameters and product quality data, and is used to characterize the optimization level of refining efficiency, purity control, and resource utilization rate during the production process.

[0037] The grading basis of copper purity is a rule system for dividing the quality grades of copper materials by standardizing the content of the main component (copper) and impurity elements in copper materials, combined with physical property requirements (such as electrical conductivity, ductility, etc.). Its core is to distinguish the quality of copper materials for different industrial uses through quantization indicators to ensure that the material properties match the scenario requirements.

[0038] Furthermore, for the copper extraction effect analysis method provided in this embodiment, step S100 includes: Step S110: Use a preset threshold to judge the eigenvalue sequence. If the eigenvalue exceeds the threshold, determine the abnormal chromaticity change point.

[0039] Extract the surface characteristics from the initial time series data to obtain the eigenvalue sequence of chromaticity change; use a preset threshold to judge the eigenvalue sequence. If the eigenvalue exceeds the threshold, determine the abnormal chromaticity change point.

[0040] The abnormal chromaticity change point refers to the area of unexpected color mutation caused by composition, structure, or environmental factors under specific processing or use conditions of the material. Its chromaticity value significantly deviates from the normal range and is usually related to microscopic mechanisms such as oxidation, impurity enrichment, or phase change.

[0041] In the monitoring of the chromaticity change of the refined copper sample, first, the copper surface is scanned by a high-precision optical sensor at a sampling frequency of 100 frames per second. The sensor uses the CIE Lab color space model to record the values of three channels: L* (brightness), a* (red-green axis), and b* (yellow-blue axis). For example, in the initial state, L* = 83 ± 2, a* = 2 ± 1, and b* = 5 ± 1 are measured.

[0042] Step S120: For the abnormal chromaticity change points, obtain the corresponding time series segments to get the time distribution of the abnormal process.

[0043] The collected time series data is denoised by the Kalman filtering algorithm. The filtering parameters are set as the process noise covariance Q = 01 and the observation noise covariance R = 05. After processing, the data standard deviation is reduced to 30% of the original value.

[0044] Step S130: Analyze the fluctuations in the refining process through the time distribution to judge the change trend in process monitoring.

[0045] The dynamic time warping (DTW) algorithm is used to align the chromaticity curves of different batches. The width of the warping window is set to 50 data points. After warping, the similarity error is reduced from 12% to 5%.

[0046] Step S140: According to the change trend, use the support vector machine algorithm to classify the refining process and determine the process state category.

[0047] To quantify the chromaticity change trend, the first derivative is used to calculate the chromaticity change rate per minute. For example, the change rate of the b* value in the oxidation stage is 15 units / minute. At the same time, the periodic characteristics are analyzed by the fast Fourier transform (FFT), and a significant spectral peak is found at 5 Hz, corresponding to the periodic growth of the surface oxide film.

[0048] Step S150: Extract the key time points from the process state category to obtain the basis for optimizing and adjusting the refining process.

[0049] A multiple linear regression model is established with temperature (X1) and humidity (X2) as independent variables and the chromaticity change rate (Y) as the dependent variable, obtaining the regression equation Y = 21X1 + 08X2 - 34 (R² = 92). The model residuals are confirmed to conform to the normal distribution through the Shapiro-Wilk test (p > 05).

[0050] During the whole process, the data is processed in real time by the Internet of Things edge computing node, and the processing delay is controlled within 200 milliseconds.

[0051] Preferably, for the method for analyzing the effect of copper extraction provided in this embodiment, step S200 includes: Step S210: Obtain the time series data of chromaticity changes, and segment the data through a preset threshold to obtain the segmented data units.

[0052] Collect the chromaticity time series data. For example, record the absorbance value of the sample at a wavelength of 450 nm every 5 seconds through a spectrometer, and continuously monitor for 120 minutes to obtain a sequence containing 1440 data points.

[0053] Step S220: For the segmented data units, use the sequence analysis method to determine the chromaticity change patterns within each segment.

[0054] Use the sliding window method to smooth the original data. The window width is set to 15 data points (corresponding to 25 minutes), and the high-frequency noise is eliminated by calculating the moving average value of the data within the window.

[0055] Step S230: Detect the occurrence points of dynamic changes through the chromaticity change patterns to obtain the change detection results.

[0056] Set the dynamic threshold to ±10% of the baseline value, and the baseline value takes the average value of 45 ± 12 during the first 30-minute stable period.

[0057] Step S240: If the change detection result exceeds the preset threshold, combine the judgment logic to extract the corresponding trend features.

[0058] Segmentation is triggered when 3 consecutive data points exceed the threshold range. Use the Bayesian Online Change Point Detection (BOCD) model in the change point detection algorithm, set the prior probability to 01, calculate the posterior probability in real time, and determine it as an effective segmentation point when the posterior probability is greater than 95.

[0059] Step S250: According to the extracted trend features, use the K-means algorithm to cluster the features to obtain the feature classification results.

[0060] Conduct trend analysis on each data segment. Use the least squares method to fit a linear equation. If the absolute value of the slope is less than 001, it is determined as a stable segment; if it is greater than 005, it is determined as a significantly changing segment. For example, a segmentation point is detected at 76 minutes, and the fitting slope of the subsequent 30-minute data reaches 0087, which is determined as an upward trend segment. The coefficient of determination R² = 92 of this segment indicates a good fit.

[0061] Step S260: Obtain the feature classification results, and judge the trend direction of the dynamic changes through the continuity analysis of the time series.

[0062] Construct a finite state automaton in combination with the segmentation results, define three states: stable, rising, and falling. When two consecutive segments belong to the rising trend and the slope difference is less than 20%, they are merged into the same state period.

[0063] Step S270: For the trend direction, fuse the change detection results to determine the final trend feature description.

[0064] The final output is an analysis report containing 6 feature segments, including parameters such as the start time, duration, trend type, and change rate of each segment, providing a quantitative basis for subsequent process control.

[0065] Furthermore, for the copper extraction effect analysis method provided in this embodiment, step S300 includes: Step S310: Obtain the conductivity value and hardness value data corresponding to the key points through the real-time monitoring system to obtain a time series record.

[0066] Collect data synchronously through a conductivity sensor and a hardness detector, and set the sampling frequency to record once every 10 seconds for continuous monitoring for 180 minutes to form a time series containing 1080 groups of conductivity-hardness paired data.

[0067] Step S320: For the time series record, use a feature extraction method to determine the analysis result of the dynamically changing points.

[0068] Use the Savitzky-Golay filter for data smoothing, with a polynomial order of 3 and a window width of 21 data points (corresponding to 5 minutes), suppressing random fluctuations while retaining trend features.

[0069] Step S330: If the point analysis result shows dynamic changes, then judge the directionality of the changes through trend features to obtain a directionality description.

[0070] The initial value of the baseline conductivity is 5 mS / cm, the hardness is 150 mg / L, and the dynamic threshold is set at ±15% of the baseline value.

[0071] Step S340: According to the directionality description, obtain the fluctuation range of the conductivity value and the hardness value to determine the fluctuation characteristics.

[0072] Use the CUSUM (Cumulative Sum Control) algorithm to detect key points, set the cumulative sum threshold to 5 times the standard deviation, and trigger an event mark when the statistic exceeds the threshold. For example, when it is monitored that the conductivity suddenly increases to 8 mS / cm at the 52nd minute and the hardness synchronously rises to 210 mg / L, the system automatically records this point.

[0073] Step S350: For the fluctuation characteristics, use the K-means algorithm to cluster the key points to obtain the point classification result.

[0074] Perform a local regression analysis on the data segments 20 minutes before and after the marked points, calculate the trend slope using the Theil-Sen estimator, and determine a significant mutation when the conductivity change rate exceeds 02 mS / cm / min and the hardness change rate exceeds 3 mg / L / min.

[0075] Step S360: Through the point classification results, fuse the continuity analysis of the time series, judge the stability of the trend characteristics, and obtain the final description.

[0076] Spatially aggregate the mutation points through the DBSCAN clustering algorithm, set the neighborhood radius ε = 15 minutes and the minimum number of samples min_samples = 2, and merge the continuous mutations with a time interval less than 15 minutes into the same key event.

[0077] Step S370: Extract the significant features of the dynamic changes from the final description to determine the distribution law of the key points.

[0078] Finally, output the time coordinates of the key events, the peak change amounts of conductivity and hardness, the duration, and the clustering attribution number to form a structured event log for system decision-making calls.

[0079] Preferably, for the method for analyzing the effect of copper extraction provided in this embodiment, step S400 includes: Step S410: For the conductivity value and hardness value data, use the polynomial regression algorithm to construct the functional relationship between copper purity and physical properties to obtain the correlation parameters.

[0080] Collect the laboratory data of copper samples, set the conductivity range to 3 - 12 mS / cm, the hardness to 100 - 300 mg / L, corresponding to the copper purity standard value of 85% - 99%, and collect a total of 500 sample points.

[0081] Step S420: Through the correlation parameters, calculate the fluctuation range between copper purity and conductivity value to determine the fluctuation characteristics.

[0082] Use the least squares method to fit a second-order polynomial model in the form of purity = + × conductivity + × hardness + × conductivity² + × conductivity × hardness + × hardness², solve the normal equation through QR decomposition to obtain the coefficient matrix β = [82, 15, -003, -008, 0004, 00002].

[0083] Step S430: According to the fluctuation characteristics, analyze the change trend between physical properties and hardness value to obtain the trend description.

[0084] To verify the robustness of the model, k-fold cross-validation (k = 5) was adopted, and the root mean square error RMSE = 45% and the coefficient of determination R² = 93 were calculated.

[0085] Step S440: If the trend description shows the same change direction, then fuse the distribution characteristics of the conductivity value and the hardness value to obtain the distribution law.

[0086] The trend description showing the same change direction means that when analyzing the trend of a certain object, it is found that the change directions of the two indicators of conductivity and hardness are the same. For example, both of them may increase or decrease simultaneously with the change of a certain factor (such as time, position, etc.).

[0087] Specifically, in the analysis process, it is necessary to carefully observe the trend of the conductivity value changing with time or conditions, and at the same time record the change of the hardness value. By comparing the fluctuation patterns of the two, the potential relationship between them can be found. For example, when the conductivity value gradually increases, the hardness value may increase or decrease accordingly, forming a specific synchronous change pattern. The identification of this pattern helps to predict future change trends and provides a scientific basis for optimizing material properties.

[0088] Step S450: Through the distribution law, use the K-means algorithm to cluster the copper purity and physical properties to determine the classification result.

[0089] For the real-time monitoring data of conductivity 2 mS / cm and hardness 185 mg / L, substituting them into the model to predict the purity of 97%, and the deviation from the measured value of 93% by X-ray fluorescence method is within the allowable range.

[0090] Step S460: According to the classification result, obtain the stability characteristics of the functional relationship and judge the stability description.

[0091] Perform the Durbin-Watson test on the residual sequence (statistic 92) to confirm that there is no autocorrelation in the error.

[0092] Step S470: For the stability description, extract the significant distribution of the physical properties to determine the significant law.

[0093] Solidify the model parameters into the production database, and when new data is input, the purity calculation is automatically triggered, and the result is written into the quality analysis report.

[0094] Preferably, for the method for analyzing the effect of copper extraction provided in this embodiment, step S500 includes: Step S510: Obtain the distribution characteristics of the refining effect through historical data to determine the distribution law.

[0095] In the copper refining process, there is a non-linear relationship between the chromaticity change and the copper purity, and a mapping model needs to be established for accurate calibration.

[0096] Step S520: Extract the fluctuation range of the chromaticity change according to the distribution law to obtain the change characteristics.

[0097] Collect 300 groups of historical refining data, with the chromaticity value range of 50 - 150 CIELAB units, corresponding to the copper purity of 88% - 95%.

[0098] Step S530: Integrate the distribution characteristics of copper purity for the change characteristics to determine the purity distribution.

[0099] Adopt the Gaussian process regression algorithm, select the radial basis function as the kernel function (length scale 2, variance 5), and construct the chromaticity - purity mapping model. Optimize the hyperparameters through maximum likelihood estimation to obtain the standard deviation of the diagonal elements of the covariance matrix as 8.

[0100] Step S540: Adjust the correlation parameters through the purity distribution to obtain the parameter adjustment result.

[0101] When the input real - time chromaticity monitoring value is 120 CIELAB, the chromaticity - purity mapping model outputs the mean value of the purity probability distribution as 92% and the standard deviation as 3%.

[0102] Step S550: If the parameter adjustment result is consistent with the historical data, then use the linear regression algorithm to calibrate the model to obtain the mapping coefficient.

[0103] Adopt Bayesian optimization to adjust the current density of the electrolytic cell, and control the deviation between the predicted purity and the target value of 99% within the range of ±5%.

[0104] Step S560: Analyze the corresponding relationship between the chromaticity change and the copper purity according to the mapping coefficient to determine the corresponding characteristics.

[0105] The chromaticity - purity mapping model automatically updates the training set every 8 hours. After the new data eliminates outliers through the 3σ criterion, the posterior distribution is recalculated. When the chromaticity sensor detects a mutation (gradient change > 5 CIELAB / minute), trigger the on - line spectral analyzer for secondary verification, and set the data consistency threshold to 90%.

[0106] Step S570: Verify the stability of the model calibration through the corresponding characteristics and judge the stability description.

[0107] Finally, write the calibrated parameters into the PLC controller to adjust the refining process parameters in real - time, so that the chromaticity fluctuation is stable within the range of ±2 CIELAB.

[0108] Furthermore, for the copper extraction effect analysis method provided in this embodiment, step S600 includes: Step S610: If the mapping coefficient exceeds the preset range, the chromaticity change data is processed using a mean filter algorithm to obtain optimized eigenvalues by removing noise interference.

[0109] During the copper refining process, if the mapping coefficient of the chromaticity change data exceeds the preset range (such as 8 - 2), the system automatically starts secondary filtering processing. Using the mean filter algorithm with a window size set to 5, the continuously collected chromaticity data is smoothed to remove noise interference.

[0110] Step S620: Determine the fluctuation range of the chromaticity change through the optimized eigenvalues, and obtain the distribution characteristics within the fluctuation range.

[0111] For example, when the input chromaticity values are [115, 118, 120, 122, 125] CIELAB, the output after filtering is 120 CIELAB, effectively reducing data fluctuations.

[0112] Step S630: Judge the stability of the chromaticity change according to the distribution characteristics to obtain the stability description result.

[0113] The system extracts the optimized chromaticity eigenvalues through principal component analysis (PCA), retains the first two principal components, and the cumulative variance contribution rate reaches 95%.

[0114] Step S640: If the stability description result meets the preset threshold, the change data is processed using a smoothing algorithm to obtain a smoothed data sequence.

[0115] Based on the extracted eigenvalues, the system uses the support vector machine (SVM) algorithm, selects the polynomial kernel (order 3, penalty coefficient 5), and constructs the mapping relationship between the chromaticity features and process parameters.

[0116] Step S650: Obtain the trend characteristics of the chromaticity change through the smoothed data sequence, and determine the distribution law of the trend characteristics.

[0117] Input the optimized chromaticity eigenvalues, and the model outputs the electrolytic cell temperature adjustment range of 1200 - 1250 °C and the current density adjustment range of 200 - 220 A / m².

[0118] Step S660: Adjust the mapping coefficient according to the distribution law of the trend characteristics to obtain the adjusted mapping parameters.

[0119] When the change rate of the chromaticity eigenvalues exceeds 3% / minute, the system triggers the on - line temperature sensor for data correction to ensure the accuracy of process parameter adjustment.

[0120] Step S670: Verify the optimized features of the chromaticity change through the adjusted mapping parameters, and judge the matching degree of the optimized features.

[0121] The optimized process parameters are sent to the execution unit in real time through the DCS system to ensure the stability of the refining process.

[0122] Preferably, for the copper extraction effect analysis method provided in this embodiment, step S700 includes: Step S710: Calculate the quantization score through the optimized chromaticity features and the calibrated correlation model to obtain a preliminary result of the real-time refining effect.

[0123] Based on the optimized chromaticity feature values, the system calculates the quantization score of the real-time refining effect through the calibrated correlation model.

[0124] Step S720: Compare the quantization score with a preset threshold. If it exceeds the threshold, the mean filtering algorithm is used to process the data input to obtain a smoothed feature sequence.

[0125] The quantization score is calculated using the weighted average method, and the weight coefficients are assigned according to the importance of the chromaticity feature values, which are 4, 3, and 3 respectively.

[0126] Step S730: Analyze the distribution characteristics of copper purity through the smoothed feature sequence to determine a preliminary range of the classification basis.

[0127] For example, when the chromaticity feature values are [85, 92, 88], the quantization score is 85×4 + 92×3 + 88×3 = 886. The system determines the classification basis of copper purity according to the quantization score. The score range of 9 - 0 is the first-level purity, 8 - 9 is the second-level purity, and 7 - 8 is the third-level purity.

[0128] Step S740: Match the distribution characteristics within the preliminary range with the correlation model to obtain the calibrated parameters after matching.

[0129] If the quantization score is lower than 7, the system automatically triggers an exception handling mechanism to perform clustering analysis on the chromaticity feature values through the Gaussian mixture model (GMM) to identify and remove abnormal data points.

[0130] Step S750: Adjust the processing flow of real-time refining according to the calibrated parameters after matching to obtain the refined data after adjustment.

[0131] The system uses the random forest algorithm (RF) to re-model the data after removing anomalies. The feature selection uses the Gini coefficient, the number of trees is set to 100, and the maximum depth is 10.

[0132] Step S760: Calculate the final quantization score through the refined data after adjustment and the feature analysis result to judge the classification result of copper purity.

[0133] After re - modeling, the system updates the quantization score and re - judges the copper purity grade. For example, when the recalculated quantization score is 91, the system determines that the copper purity is of the first grade. Finally, the system associates the grading result with the process parameters and adjusts the current density and temperature of the electrolytic cell through a fuzzy logic controller (FLC) to ensure the stability of copper purity.

[0134] The present invention relates to an effect analysis system for copper extraction, which is applied to the above - mentioned effect analysis method for copper extraction. The effect analysis system for copper extraction includes a first acquisition module, a first judgment module, a recording module, a second acquisition module, a determination module, a third acquisition module, and a second judgment module. Among them, the first acquisition module is used to acquire the chromaticity change data of the copper sample during the refining process, continuously scan the copper surface through an optical sensor, and obtain the time - series data of chromaticity change; the first judgment module is used to segment the copper extraction data according to the preset threshold based on the time - series data of chromaticity change, and judge the trend characteristics of dynamic change; the recording module is used to extract key points from the trend characteristics of dynamic change, and record the conductivity and hardness value data at the corresponding time points through a real - time monitoring system; the second acquisition module is used to construct an association model for the extracted conductivity and hardness value data, and calculate the functional relationship between copper purity and physical properties by using a polynomial regression algorithm to obtain preliminary association parameters; the determination module is used to calibrate the association model through historical refining effect data after obtaining the preliminary association parameters, and determine the mapping coefficient between chromaticity change and copper purity; the third acquisition module is used to perform secondary filtering on the chromaticity change data if the mapping coefficient exceeds the preset range, and use a mean filtering algorithm to eliminate noise interference to obtain optimized chromaticity characteristic values; the second judgment module is used to calculate the quantization score of the real - time refining effect according to the optimized chromaticity characteristic values and the calibrated association model, and judge the grading basis of copper purity.

[0135] Furthermore, for the effect analysis system for copper extraction provided in this embodiment, the first acquisition module includes a first determination unit, a first acquisition unit, a judgment unit, a second determination unit, and a second acquisition unit. Among them, the first determination unit is used to judge the eigenvalue sequence according to the preset threshold. If the eigenvalue exceeds the threshold, an abnormal chromaticity change point is determined; the first acquisition unit is used to obtain the corresponding time - series segment for the abnormal chromaticity change point to obtain the time distribution of the abnormal process; the judgment unit is used to analyze the fluctuation of the refining process through the time distribution and judge the change trend in process monitoring; the second determination unit is used to classify the refining process by using a support vector machine algorithm according to the change trend and determine the process state category; the second acquisition unit is used to extract key time points from the process state category to obtain the basis for optimizing and adjusting the refining process.

[0136] This embodiment provides a method and system for analyzing the effect of copper extraction. Compared with the prior art, by continuously scanning the surface of the copper sample with an optical sensor, the time series data of the chromaticity change during the refining process is obtained. The data is segmented using a preset threshold to extract the dynamic change trend features and key points. Combining the conductivity and hardness value data recorded by the real-time monitoring system, a polynomial regression model is constructed to establish the correlation between copper purity and physical properties. Through historical data calibration and mean filtering optimization, the mapping relationship between chromaticity change and copper purity is determined. Finally, according to the optimized chromaticity characteristic values and the calibrated correlation model, the quantization score of the real-time refining effect is calculated to determine the grading basis of copper purity. This embodiment realizes the real-time evaluation of copper purity and the dynamic optimization of the refining process, improving the refining efficiency and product quality.

[0137] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications falling within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A method for analyzing the effect of copper extraction, characterized in that: The following steps are involved: Obtain the color change data of the copper sample during the refining process, and continuously scan the copper surface with an optical sensor to obtain the time series data of the color change; According to the time series data of chromaticity changes, the copper extraction data is segmented using a preset threshold to determine the trend characteristics of dynamic changes; Extract key points from the trend characteristics of dynamic changes, and record the conductivity and hardness value data at the corresponding time points through the real-time monitoring system; Based on the extracted conductivity and hardness value data, a correlation model was constructed, and the functional relationship between copper purity and physical properties was calculated using a polynomial regression algorithm to obtain preliminary correlation parameters; After obtaining the preliminary correlation parameters, the correlation model is calibrated through historical refining effect data to determine the mapping coefficient between chromaticity change and copper purity; If the mapping coefficient exceeds the preset range, the chromaticity change data is subjected to secondary filtering, and the noise interference is eliminated by using a mean filtering algorithm to obtain an optimized chromaticity characteristic value; According to the optimized chromaticity characteristic value and the calibrated correlation model, the quantitative score of the real-time refining effect is calculated to determine the grading basis of copper purity.

2. The method for analyzing the effect of copper extraction as claimed in claim 1, wherein: The step of obtaining the color change data of the copper sample during the refining process, and continuously scanning the copper surface by an optical sensor to obtain the time series data of the color change comprises: The eigenvalue sequence is judged using a preset threshold value, and if the eigenvalue exceeds the threshold value, the abnormal chromaticity change point is determined; For abnormal chromaticity change points, the corresponding time series segments are obtained to obtain the time distribution of the abnormal process; Analyze the fluctuation of the refining process through time distribution and determine the changing trend in process monitoring; According to the changing trend, the support vector machine algorithm is used to classify the refining process and determine the process status category; Extract key time points from process status categories to obtain the basis for optimizing and adjusting the refining process.

3. The effect analysis method of copper extraction as claimed in claim 1, characterized in that: The step of segmenting the copper extraction data according to the time series data of chromaticity change by using a preset threshold value to determine the trend characteristics of dynamic change includes: Acquire the time series data of chromaticity change, segment the data by a preset threshold value, and obtain segmented data units; For the segmented data units, a sequence analysis method is used to determine the chromaticity change pattern within each segment; Through the chromaticity change pattern, the occurrence point of dynamic change is detected to obtain the change detection result; If the change detection result exceeds the preset threshold, the corresponding trend features are extracted in combination with the judgment logic; According to the extracted trend features, the K-means algorithm is used to cluster the features and obtain the feature classification results; Obtain feature classification results and determine the trend direction of dynamic changes through continuity analysis of time series; According to the trend direction, the change detection results are integrated to determine the final trend feature description.

4. The method for analyzing the effect of copper extraction as claimed in claim 1, characterized in that: The step of extracting key points from the trend characteristics of dynamic changes and recording the conductivity and hardness value data at corresponding time points through a real-time monitoring system includes: The conductivity and hardness data corresponding to the key points are obtained through the real-time monitoring system to obtain time series records; For time series records, feature extraction methods are used to determine the analysis results of dynamically changing points; If the point analysis results show dynamic changes, the direction of the change is determined by the trend characteristics to obtain a directional description; According to the directional description, the fluctuation range of the conductivity value and the hardness value is obtained, and the fluctuation characteristics are determined; According to the fluctuation characteristics, the K-means algorithm is used to cluster the key points and obtain the point classification results; Through the point classification results, the continuity analysis of the time series is integrated to determine the stability of the trend characteristics and obtain the final description; Extract the significant features of dynamic changes from the final description and determine the distribution pattern of key points.

5. The method for analyzing the effect of copper extraction as claimed in claim 1, characterized in that: The steps of constructing a correlation model for the extracted conductivity and hardness value data, calculating the functional relationship between copper purity and physical properties using a polynomial regression algorithm, and obtaining preliminary correlation parameters include: Based on the conductivity and hardness data, a polynomial regression algorithm is used to construct the functional relationship between copper purity and physical properties, and the correlation parameters are obtained; By correlating parameters, the fluctuation range between copper purity and conductivity values ​​is calculated to determine the fluctuation characteristics; According to the fluctuation characteristics, the changing trend between physical properties and hardness values ​​is analyzed to obtain trend description; If the trend description shows that the change direction is consistent, the distribution characteristics of the conductivity value and the hardness value are integrated to obtain the distribution law; Based on the distribution law, K-means algorithm is used to cluster the copper purity and physical properties to determine the classification results; According to the classification results, the stability characteristics of the functional relationship are obtained and the stability description is determined; For stability description, the significant distribution of physical properties is extracted and the significant laws are determined.

6. The method for analyzing the effect of copper extraction according to claim 1, wherein: After obtaining the preliminary correlation parameters, the correlation model is calibrated by historical refining effect data, and the steps of determining the mapping coefficient between the chromaticity change and the copper purity include: Obtain the distribution characteristics of the refining effect through historical data and determine the distribution law; Extract the fluctuation range of chromaticity change according to the distribution law and obtain the change characteristics; According to the change characteristics, the distribution characteristics of copper purity are integrated to determine the purity distribution; Adjust the associated parameters through purity distribution to obtain parameter adjustment results; If the parameter adjustment results are consistent with historical data, the linear regression algorithm is used to calibrate the model and obtain the mapping coefficients; The corresponding relationship between the chromaticity change and the copper purity is analyzed according to the mapping coefficient to determine the corresponding characteristics; Verify the stability of the model calibration through the corresponding features and judge the stability description.

7. The method for analyzing the effect of copper extraction according to claim 1, wherein: If the mapping coefficient exceeds the preset range, the steps of performing secondary filtering on the chromaticity change data, using a mean filtering algorithm to remove noise interference, and obtaining an optimized chromaticity characteristic value include: If the mapping coefficient exceeds the preset range, the chromaticity change data is processed using a mean filter algorithm to obtain an optimized eigenvalue by eliminating noise interference; Determine the fluctuation range of chromaticity change through the optimized eigenvalues, and obtain the distribution characteristics within the fluctuation range; The stability of chromaticity change is determined based on the distribution characteristics, and the stability description result is obtained; If the stability description result meets the preset threshold, the change data is processed by a smoothing algorithm to obtain a smoothed data sequence; Obtain the trend characteristics of chromaticity change through the smoothed data sequence and determine the distribution law of the trend characteristics; Adjusting the mapping coefficient according to the distribution law of the trend characteristics to obtain adjusted mapping parameters; The optimized features of the chromaticity change are verified by adjusting the mapping parameters to determine the matching degree of the optimized features.

8. The method for analyzing the effect of copper extraction as claimed in claim 1, characterized in that: The step of calculating the quantitative score of the real-time refining effect based on the optimized chromaticity characteristic value and the calibrated correlation model and determining the classification basis of the copper purity includes: The quantitative scores were calculated by using the optimized chromaticity features and the calibrated association model to obtain preliminary results of the real-time refinement effect; The quantitative score is compared with the preset threshold. If it exceeds the threshold, the data input is processed by the mean filter algorithm to obtain a smoothed feature sequence. The distribution characteristics of copper purity are analyzed through the smoothed characteristic sequence to determine the preliminary range of classification basis; Matching the distribution characteristics within the preliminary range with the correlation model to obtain the matched calibration parameters; adjusting the processing flow of real-time refinement according to the matched calibration parameters to obtain adjusted refined data; The final quantitative score is calculated through the adjusted refined data and characteristic analysis results to determine the grading result of copper purity.

9. A copper extraction effect analysis system, applied to the copper extraction effect analysis method according to any one of claims 1 to 8, characterized in that: The effect analysis system of copper extraction comprises: The first acquisition module is used to acquire the color change data of the copper sample during the refining process, and continuously scans the copper surface through an optical sensor to obtain the time series data of the color change; The first judgment module is used to segment the copper extraction data according to the time series data of chromaticity change using a preset threshold value to judge the trend characteristics of dynamic change; The recording module is used to extract key points from the trend characteristics of dynamic changes and record the conductivity and hardness value data at the corresponding time points through the real-time monitoring system; The second acquisition module is used to construct a correlation model for the extracted conductivity and hardness value data, calculate the functional relationship between copper purity and physical properties using a polynomial regression algorithm, and obtain preliminary correlation parameters; A determination module, for calibrating the association model through historical refining effect data after obtaining the preliminary association parameters, and determining a mapping coefficient between chromaticity change and copper purity; A third acquisition module is used to perform secondary filtering on the chromaticity change data if the mapping coefficient exceeds a preset range, and adopt a mean filtering algorithm to eliminate noise interference to obtain an optimized chromaticity feature value; The second judgment module is used to calculate the quantitative score of the real-time refining effect according to the optimized chromaticity characteristic value and the calibrated correlation model, and judge the classification basis of the copper purity.

10. The copper extraction effect analysis system according to claim 9, characterized in that: The first acquisition module includes: A first determination unit, configured to use a preset threshold to judge the characteristic value sequence, and if the characteristic value exceeds the threshold, determine an abnormal chromaticity change point; A first acquisition unit is used to acquire a corresponding time series segment for an abnormal chromaticity change point, and obtain a time distribution of the abnormal process; A judgment unit is used to analyze the fluctuation of the refining process through time distribution and judge the changing trend in process monitoring; A second determination unit is used to classify the refining process according to the change trend by using a support vector machine algorithm to determine the process state category; The second acquisition unit is used to extract key time points from the process state categories to obtain the optimization adjustment basis of the refining process.

Images