A method for preparing refined tellurium from copper telluride slag

Abstract

The present application relates to tellurium preparation technical field, specifically to a kind of for copper telluride slag preparation refined tellurium method;Temporal decomposition is carried out to concentration time curve, and outlier residual point is obtained according to the discrete characteristics of residual point in residual curve;According to the distance characteristics of outlier residual point and adjacent residual point, the data variation characteristics of neighborhood residual point, obtain slow fluctuation value;According to the residual point of outlier residual point and other target product, obtain change correlation degree;According to the distribution characteristics of suspected oxygen pressure abnormal point, clustering is carried out, and according to the area characteristics of point cluster, the density characteristics and distribution characteristics of residual point in cluster obtain oxygen pressure abnormal time.This application obtains oxygen pressure abnormal trend value according to the distribution characteristics and abnormal degree of oxygen pressure abnormal time and adjusts and early warns oxygen pressure, improves the timeliness and accuracy of monitoring, and guarantees the preparation effect of tellurium.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tellurium preparation, in particular to a method for preparing refined tellurium from copper telluride slag. BACKGROUND

[0002] The preparation of refined tellurium from copper telluride slag refers to mixing copper telluride slag with an oxidizing agent for oxygen pressure alkaline leaching to obtain alkaline leaching slag containing sodium tellurate and copper oxide, and then separating copper and part of rare elements from the alkaline leaching slag by sulfidation leaching, so that tellurium is enriched in the sulfidation leaching solution, and finally the sulfidation leaching solution is reduced to recover refined tellurium. In the preparation process, if the reaction parameters of the oxygen pressure alkaline leaching process are not controlled properly, it will directly lead to low leaching rate, impurity dissolution or incomplete oxidation of tellurium, so real-time monitoring is needed during the reaction process to ensure normal reaction. The oxygen pressure parameter in the oxygen pressure alkaline leaching process has a great influence on the reaction. The existing method usually monitors whether the oxygen pressure parameter is abnormal according to the oxygen pressure threshold value, but the range of threshold value monitoring is large, and when the threshold value is approached, the abnormal reaction may have been carried out for a period of time, resulting in low accuracy and timeliness of monitoring the oxygen pressure state by the oxygen pressure threshold value, and finally affecting the preparation effect of refined tellurium. SUMMARY

[0003] In order to solve the technical problem that monitoring the oxygen pressure parameter in the oxygen pressure alkaline leaching process by the oxygen pressure threshold value will result in low accuracy and timeliness of monitoring and affect the preparation effect of refined tellurium, the purpose of the present application is to provide a method for preparing refined tellurium from copper telluride slag, and the technical solution adopted is as follows:

[0004] Obtain a concentration time sequence curve of a target product in the oxygen pressure alkaline leaching process;

[0005] Time sequence decompose the concentration time sequence curve to obtain a residual curve, obtain an outlier residual point according to the dispersion characteristics of the residual points in the residual curve, obtain a slow fluctuation value according to the distance characteristics of the outlier residual point and adjacent residual points and the data change characteristics of the neighborhood residual points of the outlier residual point, obtain a change correlation degree according to the adjacent data change difference characteristics of the outlier residual point and the residual points of the same time of other target products, obtain a first oxygen pressure abnormality degree of the outlier residual point according to the slow fluctuation value and the change correlation degree, and obtain a suspected oxygen pressure abnormal point according to the first oxygen pressure abnormality degree;

[0006] Cluster according to the distribution characteristics of the suspected oxygen pressure abnormal point to obtain different point clusters, obtain a second oxygen pressure abnormality degree of the suspected oxygen pressure abnormal point according to the area characteristics of the point clusters, the density characteristics and distribution characteristics of the residual points in the clusters, obtain an oxygen pressure abnormal time according to the second oxygen pressure abnormality degree, and obtain an oxygen pressure abnormal trend value according to the distribution characteristics and abnormality degree of the oxygen pressure abnormal time;

[0007] Adjust and warn the oxygen pressure according to the oxygen pressure abnormal trend.

[0008] Further, the step of obtaining an outlier residual point according to the dispersion characteristics of the residual points in the residual curve comprises:

[0009] constructing a residual mean line according to the average value of all residual points in the residual curve; calculating the average value of the distance of all residual points from the residual mean line and the distance standard deviation regarding the distance of the residual points in the residual curve from the residual mean line as the outlier residual point whose distance is not within .

[0010] Further, the step of obtaining a slow fluctuation value according to the distance characteristics of the outlier residual point and the adjacent residual points, and the data change characteristics of the neighborhood residual points of the outlier residual point comprises:

[0011] calculating the sum of the Euclidean distances of the outlier residual point and the two adjacent residual points before and after it, to obtain a comprehensive distance; calculating the slope of the connecting line between any residual point and the previous adjacent residual point to obtain the change degree of the any residual point; calculating the average value of the absolute value of the difference between the change degrees of all adjacent residual points within a preset previous adjacent range of the outlier residual point, to obtain a first change difference value; calculating the average value of the absolute value of the difference between the change degrees of all adjacent residual points within a preset subsequent adjacent range of the outlier residual point, to obtain a second change difference value; calculating the reciprocal of the sum of the first change difference value and the second change difference value, to obtain a third value; calculating the product of the reciprocal of the comprehensive distance and the third value, to obtain the slow fluctuation value of the outlier residual point.

[0012] Further, the step of obtaining a change correlation degree according to the adjacent data change difference characteristics of the residual points of other target products at the same time as the outlier residual point comprises:

[0013] normalizing the comprehensive distance of the residual points of other target products at the time of the outlier residual point to obtain a fourth value; normalizing the comprehensive distance of the outlier residual point and calculating the reciprocal of the absolute value of the difference between the fourth value to obtain the change correlation degree of the outlier residual point.

[0014] Further, the step of obtaining a first oxygen pressure abnormality degree of the outlier residual point according to the slow fluctuation value and the change correlation degree comprises:

[0015] normalizing the slow fluctuation value and the change correlation degree respectively and calculating the average value of the normalized values to obtain the first oxygen pressure abnormality degree of the outlier residual point.

[0016] Further, the step of obtaining a suspected oxygen pressure abnormal point according to the first oxygen pressure abnormal degree comprises:

[0017] The outlier residual point with the first oxygen pressure abnormal degree exceeding a preset first threshold value is taken as the suspected oxygen pressure abnormal point.

[0018] Further, the step of obtaining a second oxygen pressure abnormal degree of a suspected oxygen pressure abnormal point according to the area feature, the density feature and the distribution feature of the point cluster comprises:

[0019] The area of the minimum circumscribed circle of the point cluster is calculated and normalized to obtain an area feature value; the average value of the nearest neighbor distance of all suspected oxygen pressure abnormal points in the point cluster is calculated and normalized to obtain a spacing feature value; the standard deviation of the nearest neighbor distance of all suspected oxygen pressure abnormal points in the point cluster is calculated and normalized to obtain a distribution dispersion feature value; and the average value of the area feature value, the spacing feature value and the distribution dispersion feature value is calculated to obtain the second oxygen pressure abnormal degree of all suspected oxygen pressure abnormal points in the point cluster.

[0020] Further, the step of obtaining an oxygen pressure abnormal time according to the second oxygen pressure abnormal degree comprises:

[0021] The time point of the suspected oxygen pressure abnormal point with the second oxygen pressure abnormal degree exceeding a preset second threshold value is taken as the oxygen pressure abnormal time.

[0022] Further, the step of obtaining an oxygen pressure abnormal trend value according to the distribution feature and the abnormal degree of the oxygen pressure abnormal time comprises:

[0023] The number ratio of the oxygen pressure abnormal time after the middle time point to the oxygen pressure abnormal time before the middle time point of the concentration time sequence curve is calculated to obtain a number ratio; the average value of the distance of the residual point of the oxygen pressure abnormal time after the middle time point to the residual mean line is calculated to obtain a first distance; the average value of the distance of the residual point of the oxygen pressure abnormal time before the middle time point to the residual mean line is calculated to obtain a second distance; the ratio of the first distance to the second distance is calculated to obtain an abnormality ratio; and the product of the number ratio and the abnormality ratio is calculated to obtain the oxygen pressure abnormal trend value.

[0024] Further, the step of adjusting and warning the oxygen pressure according to the oxygen pressure abnormal trend comprises:

[0025] The sum value of the oxygen pressure abnormal trend value and a constant 1 is calculated to obtain an adjustment coefficient; the product of the adjustment coefficient and a preset basic oxygen pressure is calculated to obtain an adjusted oxygen pressure value; and when the oxygen pressure abnormal trend value exceeds a preset abnormal trend threshold value, the oxygen pressure alkali leaching process is warned.

[0026] The present application has the following advantages:

[0027] In the present application, the acquisition of the residual curve can determine the degree of each time in the concentration time curve not conforming to the periodic characteristics and trend characteristics, the acquisition of the outlier residual point can determine the time obviously not conforming to the periodic characteristics and trend characteristics, and then the oxygen pressure abnormality in the oxygen pressure alkaline leaching process is judged through the outlier residual point. Since there is noise data in the concentration collection process of the target product, the outlier residual points caused by the noise data and the outlier residual points caused by the reaction abnormality have obvious feature differences in the local area, so the acquisition of the slow fluctuation value and the change correlation degree can represent whether the outlier residual point is caused by the noise data; the acquisition of the first oxygen pressure abnormality degree can represent the possibility that the outlier residual point is caused by the noise data; the acquisition of the suspected oxygen pressure abnormality point can determine the outlier residual point caused by the non-noise data, and the accuracy of the reaction monitoring is initially improved. The acquisition of different point clusters can further judge the cause of the suspected oxygen pressure abnormality point; the acquisition of the second oxygen pressure abnormality degree can represent the possibility that the suspected oxygen pressure abnormality point is caused by the oxygen pressure abnormality, and the acquisition of the oxygen pressure abnormality time can improve the accuracy of the oxygen pressure monitoring. The acquisition of the oxygen pressure abnormality trend value can judge whether the oxygen pressure abnormality has a gradually serious trend, and finally the oxygen pressure is adjusted and warned according to the oxygen pressure abnormality trend; compared with the threshold monitoring, the timeliness and accuracy of the reaction monitoring are improved, and the preparation effect of the refined tellurium is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without any creative effort.

[0029] Figure 1 A flow chart of a method for preparing refined tellurium from copper telluride slag is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following will combine the drawings and the preferred embodiments to specifically describe the method for preparing refined tellurium from copper telluride slag according to the present application, its specific implementation, structure, features and effects in detail. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0031] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0032] The application provides a method for preparing refined tellurium from copper telluride slag.

[0033] Please refer to Figure 1 , which shows a flow chart of a method for preparing refined tellurium from copper telluride slag according to an embodiment of the application, which comprises the following steps:

[0034] Step S1, obtaining a concentration time sequence curve of a target product in an oxygen pressure alkaline leaching process.

[0035] First, a concentration time sequence curve of a target product in an oxygen pressure alkaline leaching process is obtained. The reaction process of the copper telluride slag and the oxidizing agent in the oxygen pressure alkaline leaching is as follows: 、 During the process from the start to the end of the oxygen pressure alkaline leaching reaction, the concentration of the target product tellurate ion and copper oxide is monitored by using spectrophotometry, the monitoring and collecting frequency is 10 Hz, and the concentration time sequence curve of the target product is updated in real time according to the time sequence of collection. In the embodiment of the application, the concentration time sequence segment of the previous 60 seconds at the current time in the reaction process is taken as the latest concentration time sequence curve, and the implementer can determine it according to the implementation scene.

[0036] Step S2, time sequence decomposition is performed on the concentration time sequence curve to obtain a residual curve; according to the discrete characteristics of residual points in the residual curve, outlying residual points are obtained; according to the distance characteristics of the outlying residual points and adjacent residual points and the data change characteristics of the neighborhood residual points of the outlying residual points, a slow fluctuation value is obtained; according to the adjacent data change difference characteristics of the outlying residual points and the residual points of the same time of other target products, a change correlation degree is obtained; according to the slow fluctuation value and the change correlation degree, a first oxygen pressure abnormality degree of the outlying residual points is obtained; and according to the first oxygen pressure abnormality degree, a suspected oxygen pressure abnormal point is obtained.

[0037] In the oxygen-pressure alkaline leaching process, oxygen pressure simultaneously affects the concentration changes of the target product in both reaction processes. Therefore, the concentration-time series curve of any target product can be analyzed to determine whether the oxygen pressure is abnormal. As the oxygen-pressure alkaline leaching proceeds, the concentration of the target product gradually increases, exhibiting a slow, trending change. Furthermore, the reaction rate is not constant during the reaction process. Due to the automatic alkali replenishment or intermittent heating operations typically used in the oxygen-pressure alkaline leaching process, fluctuations in liquid phase conditions can cause periodic fluctuations in the concentration of the target product. Therefore, the concentration-time series curve of the target product will exhibit a certain trend and periodicity. The periodic and trending concentration-time series curve can then be decomposed to obtain the residual curve. It should be noted that time-series decomposition is an existing technology that can obtain the trend term, periodic term, and residual term of the original data. The specific decomposition steps will not be elaborated here. The residual curve represents the residual term corresponding to the concentration-time series curve. Since the residual term is obtained by subtracting the periodic term and trend term from the original curve, the residual points in the residual curve can reflect the degree of deviation of the concentration-time series curve from the periodicity and trend at any given time. The more discrete the residuals are at the residual points in the residual curve, the more the concentration at that moment deviates from the normal periodic and trend characteristics, and the more likely there is an anomaly in the oxygen pressure. Therefore, outlier residuals can be obtained based on the discrete characteristics of the residual points in the residual curve.

[0038] Preferably, in this embodiment of the invention, the step of obtaining outlier residuals includes: constructing a residual mean line based on the average value of all residuals in the residual curve; the farther a residual point is from the residual mean line, the more discrete its distribution, and the greater its deviation from the period and trend in the concentration time series curve at that moment. The average distance of all residuals from the residual mean line is calculated. and distance standard deviation The distance between the residual points in the residual curve and the residual mean line is not... Residual points within the specified range are considered outlier residuals. The acquisition of outlier residuals is based on existing... According to the criteria, outlier residuals are more outlier than other residuals. The time of occurrence of outlier residuals indicates the periodic and trend characteristics of concentration deviation from the normal range, and may indicate the risk of abnormal reaction.

[0039] Furthermore, since numerous factors contribute to outlier residuals, it is necessary to determine the causes of these outlier residuals to improve monitoring accuracy. In actual monitoring, noise data is unavoidable during the data acquisition process, which may cause some outlier residuals. The main anomaly in the oxygen-pressure alkaline leaching reaction is oxygen pressure anomaly. Low or intermittent oxygen pressure leading to uneven oxidant supply rates can interrupt or repeat tellurium oxidation, causing concentration fluctuations. The effect of oxygen pressure on concentration is gradual, with a gentle waveform. Therefore, in the residual curve, outlier residuals caused by oxygen pressure anomalies are relatively close to adjacent residuals; and because they exhibit a gradual deviation, the change characteristics of other residuals in their neighborhood are similar. Noise data, on the other hand, is usually a highly random and sudden event, and the appearance of outlier residuals caused by noise data is more abrupt. Therefore, the slowly fluctuating values ​​can be obtained based on the distance characteristics between outlier residuals and adjacent residuals, and the data change characteristics of the neighboring residuals of the outlier residuals.

[0040] Preferably, in this embodiment of the invention, the step of obtaining the slow fluctuation value includes: calculating the sum of the Euclidean distances between the outlier residual and its two adjacent residuals to obtain the comprehensive distance; the smaller the comprehensive distance, the closer the residuals are, the more likely the outlier residual is caused by oxygen pressure anomaly, and the less likely it is caused by noise. Calculating the slope of the line connecting any residual to its previous adjacent residual to obtain the degree of change of any residual; the degree of change reflects the magnitude of the change of the residual. Calculating the average of the absolute values ​​of the differences in the degrees of change between all adjacent residuals within a preset preceding adjacent range of the outlier residual to obtain the first change difference value; in this embodiment of the invention, the preset preceding adjacent range refers to the range including the outlier residual and its three adjacent previous residuals, which can be determined by the implementer according to the implementation scenario. The smaller the first change difference value, the more similar the magnitudes of change between the adjacent residuals before the outlier residual, and the more likely it is a slow change caused by oxygen pressure anomaly; the larger the first change difference value, the greater the difference in magnitude of change, and the more likely it is a significant fluctuation caused by noise data. The average of the absolute values ​​of the differences in the degree of change between all adjacent residuals within a preset adjacent range of the outlier residual is calculated to obtain the second change difference value. In this embodiment of the invention, the preset adjacent range refers to the range including the outlier residual and the three adjacent residuals, which can be determined by the implementer according to the implementation scenario. Similarly, the larger the second change difference value, the more likely it is a slow change caused by an abnormal oxygen pressure. The reciprocal of the sum of the first and second change difference values ​​is calculated to obtain the third value. The smaller the first and second change difference values, the larger the third value, meaning that the outlier residual is less likely to be caused by noise data. The product of the reciprocal of the aggregate distance and the third value is calculated to obtain the slow fluctuation value of the outlier residual. The larger the aggregate distance and the smaller the third value, the smaller the slow fluctuation value, and the more likely it is that the outlier residual is caused by noise data. The formula for obtaining the slow fluctuation value includes:

[0041]

[0042] In the formula, R represents the slow fluctuation value, D represents the comprehensive distance, and M represents the number of residual points in the adjacent range before the preset time. This represents the degree of change at the m-th residual point. Indicates the first The degree of change of each residual point This represents the first change difference value. This indicates the number of residual points within the adjacent range after the preset value. This represents the degree of change at the nth residual point. Indicates the first The degree of change of each residual point This indicates the second variation value. This represents the third numerical value.

[0043] Furthermore, since abnormal oxygen pressure during oxygen-pressure leaching will simultaneously cause similar effects on the concentration changes of the two target products, the concentration changes are similar, and the neighborhood differences at the same residual point are similar. Therefore, the degree of correlation of changes can be obtained based on the difference characteristics of adjacent data changes of residual points of outlier residuals and other target products at the same time. Preferably, in this embodiment of the invention, the step of obtaining the degree of correlation of changes includes: normalizing the comprehensive distance of the residual points of other target products at the time where the outlier residual is located to obtain a fourth value; normalizing the comprehensive distance of the outlier residual and calculating the reciprocal of the absolute value of the difference with the fourth value to obtain the degree of correlation of changes of the outlier residual; in this embodiment of the invention, the normalization step uses a linear normalization method, and normalizing the comprehensive distances corresponding to different target products is to unify the magnitude. When the aggregate distance between two target products is close at the same time, it means that the concentration change trends are similar around that time. The greater the correlation between the changes, the greater the possibility of an oxygen pressure anomaly. The smaller the correlation between the changes, the greater the difference in aggregate distance. The concentration change trends of the two target products at the same time are quite different, and the outlier residual is more likely to be caused by noisy data.

[0044] Furthermore, the first oxygen pressure anomaly level of outlier residuals can be obtained based on the slow fluctuation value and the degree of correlation of changes. Preferably, in this embodiment of the invention, the step of obtaining the first oxygen pressure anomaly level includes: normalizing the slow fluctuation value and the degree of correlation of changes respectively and calculating the normalized average value to obtain the first oxygen pressure anomaly level of the outlier residual. The larger the first oxygen pressure anomaly level, the more likely the outlier residual is to have an oxygen pressure anomaly; the smaller the first oxygen pressure anomaly level, the more likely the outlier residual is to be caused by noise data. Therefore, obtaining suspected oxygen pressure anomaly points based on the first oxygen pressure anomaly level specifically includes: taking outlier residuals with a first oxygen pressure anomaly level exceeding a preset first threshold as suspected oxygen pressure anomaly points. In this embodiment of the invention, since the first oxygen pressure anomaly level is a normalized average value, the result corresponding to the real anomaly will tend to 1, while the result corresponding to the noise data will tend to 0. The two are distributed at opposite ends of the value range, so the preset first threshold is 0.5. By calculating the first oxygen pressure anomaly level, suspected oxygen pressure anomaly points among outlier residuals can be identified, initially improving the monitoring accuracy.

[0045] Step S3: Cluster the suspected oxygen pressure anomaly points according to their distribution characteristics to obtain different point clusters; obtain the second oxygen pressure anomaly degree of the suspected oxygen pressure anomaly points according to the area characteristics, density characteristics and distribution characteristics of residual points within the clusters; obtain the oxygen pressure anomaly time according to the second oxygen pressure anomaly degree; obtain the oxygen pressure anomaly trend value according to the distribution characteristics and anomaly degree of the oxygen pressure anomaly time.

[0046] In the oxygen-pressure alkaline leaching process, alkaline leaching is mainly a solid-liquid phase contact reaction. If the stirring is uneven, precipitation and agglomeration can easily occur during stirring, resulting in localized reactions that are initially fast and then slow down. This is typically manifested as fluctuations in the concentration curve, which are difficult to distinguish from oxygen pressure anomalies in terms of waveform. However, uneven stirring is temporary and will disappear after continuous stirring, so it should not be analyzed as abnormal reaction data. Therefore, further analysis is needed to determine whether uneven stirring is the cause of suspected oxygen pressure anomalies. The concentration anomalies caused by uneven stirring are short-lived and will disappear after further stirring. Therefore, suspected oxygen pressure anomalies caused by this reason have a concentrated distribution and similar values ​​in time. On the other hand, uneven oxidant supply rates caused by excessively low or intermittent oxygen pressure will interrupt or repeat the oxidation of tellurium and copper, resulting in an uneven distribution in time. Therefore, clustering can be performed based on the distribution characteristics of suspected oxygen pressure anomalies to obtain different clusters. In this embodiment of the invention, the existing K-means clustering algorithm is used to cluster all suspected oxygen pressure anomalies in the residual curve. The horizontal axis of the clustering space is time, and the vertical axis is the residual value. The number of clusters is obtained by the elbow method. The K-means clustering algorithm is an existing technology, and the specific clustering steps will not be described in detail. Through clustering, suspected oxygen pressure anomalies with similar time and residual values ​​can be grouped into one cluster.

[0047] Furthermore, if the cluster of points is small in size and the distribution of points within the cluster is relatively dense and uniform, then the cluster is more likely to be a cluster of suspected oxygen pressure anomaly points caused by uneven mixing; if the cluster of points is large in size and the distribution of points within the cluster is discrete and uneven, then the cluster is more likely to be a cluster of suspected oxygen pressure anomaly points caused by oxygen pressure anomaly. Therefore, the second degree of oxygen pressure anomaly of suspected oxygen pressure anomaly points is obtained based on the area characteristics of the cluster, the density characteristics and distribution characteristics of the residual points within the cluster; preferably, in this embodiment of the invention, the step of obtaining the second degree of oxygen pressure anomaly includes: calculating and normalizing the area of ​​the smallest circumcircle of the cluster to obtain an area feature value; the larger the area feature value, the more likely the cluster is to be a cluster of suspected oxygen pressure anomaly points caused by oxygen pressure anomaly. The average and normalized nearest neighbor distances of all suspected oxygen pressure anomaly points within a cluster are calculated to obtain the interval characteristic value. The nearest neighbor distance is the Euclidean distance between any suspected oxygen pressure anomaly point and the nearest other suspected oxygen pressure anomaly point within the cluster. A larger interval characteristic value indicates a more discrete distribution of suspected oxygen pressure anomaly points within the cluster, suggesting the cluster is more likely to be a cluster of suspected oxygen pressure anomaly points. The standard deviation and normalized nearest neighbor distances of all suspected oxygen pressure anomaly points within a cluster are calculated to obtain the distribution discreteness characteristic value. A larger distribution discreteness characteristic value indicates a more uneven distribution of suspected oxygen pressure anomaly points within the cluster, suggesting the cluster is more likely to be a cluster of suspected oxygen pressure anomaly points. The average of the area characteristic value, interval characteristic value, and distribution discreteness characteristic value is calculated to obtain the second oxygen pressure anomaly degree of all suspected oxygen pressure anomaly points within the cluster. A larger second oxygen pressure anomaly degree indicates that the suspected oxygen pressure anomaly points within the cluster are more likely to be caused by oxygen pressure anomalies.

[0048] Furthermore, the time of oxygen pressure anomaly can be obtained based on the degree of the second oxygen pressure anomaly. Specifically, this includes: taking the time when the suspected oxygen pressure anomaly point where the degree of the second oxygen pressure anomaly exceeds a preset second threshold as the time of oxygen pressure anomaly. In this embodiment of the invention, since the degree of the second oxygen pressure anomaly is the average value of each feature after normalization, the degree of the second oxygen pressure anomaly caused by uneven stirring tends to 0, while the degree of the second oxygen pressure anomaly caused by oxygen pressure anomaly tends to 1. The two are distributed at opposite ends of the value range, so the preset second threshold is 0.5. The time of oxygen pressure anomaly indicates that the oxygen pressure at that moment does not meet the requirements of normal oxygen pressure alkali leaching. If the number of times of oxygen pressure anomaly gradually increases and the corresponding residual values ​​are relatively discrete, it means that the oxygen pressure anomaly has a gradually worsening trend. Therefore, the oxygen pressure anomaly trend value is obtained based on the distribution characteristics and degree of anomaly of the times of oxygen pressure anomaly.

[0049] Preferably, in this embodiment of the invention, the step of obtaining the oxygen pressure anomaly trend value includes: calculating the ratio of the number of oxygen pressure anomaly moments after the midpoint of the concentration time series curve to the number before the midpoint, obtaining a quantity ratio; the larger the quantity ratio, the closer to the current moment, the more oxygen pressure anomaly moments there are, and the more severe the oxygen pressure anomaly tends to be. Calculating the average distance of the residual points of oxygen pressure anomaly moments after the midpoint from the residual mean line, obtaining a first distance; calculating the average distance of the residual points of oxygen pressure anomaly moments before the midpoint from the residual mean line, obtaining a second distance; calculating the ratio of the first distance to the second distance, obtaining an anomaly ratio; the larger the anomaly ratio, the closer to the current moment, the more discrete the residual values ​​corresponding to the oxygen pressure anomaly moments are, and the more obvious the deviation of the concentration time series curve from the normal cycle and trend characteristics, indicating a more severe oxygen pressure anomaly. Calculating the product of the quantity ratio and the anomaly ratio, obtaining the oxygen pressure anomaly trend value; the larger the oxygen pressure anomaly trend value, the closer to the current moment, the more obvious and severe the oxygen pressure anomaly. By analyzing the changing trends of abnormal oxygen pressure, abnormal oxygen pressure conditions can be quickly detected, thus improving the accuracy of oxygen pressure anomaly monitoring.

[0050] Step S4: Adjust and issue early warnings for oxygen pressure based on abnormal oxygen pressure trends.

[0051] After obtaining the current oxygen pressure anomaly trend, the oxygen pressure can be adjusted and warnings issued based on this trend. Specifically, this includes: calculating the sum of the oxygen pressure anomaly trend value and a constant 1 to obtain an adjustment coefficient; the larger the oxygen pressure anomaly trend value, the greater the need to increase the oxygen pressure, hence the larger the adjustment coefficient. The product of the adjustment coefficient and the preset baseline oxygen pressure is then calculated to obtain the adjusted oxygen pressure value; the larger the oxygen pressure anomaly trend value, the larger the adjusted oxygen pressure value, thus ensuring the normal operation of the oxygen-pressure alkaline leaching process and improving the tellurium refining effect. When the oxygen pressure anomaly trend value exceeds a preset anomaly trend threshold, a warning is issued for the oxygen-pressure alkaline leaching process, indicating that the oxygen pressure anomaly trend value is too high and requires manual inspection. In this embodiment, the preset anomaly trend threshold is 0.6, which can be determined by the implementer according to the implementation scenario. It should be noted that the oxygen pressure is adjusted and warnings are issued based on the maximum value of the oxygen pressure anomaly trend corresponding to all target products. Therefore, monitoring the oxygen pressure of the oxygen-pressure alkaline leaching process through the oxygen pressure anomaly trend value is more accurate and timely than monitoring through the oxygen pressure threshold.

[0052] In summary, this invention provides a method for preparing refined tellurium from copper telluride slag. The method involves time-series decomposition of the concentration time-series curve, obtaining outlier residuals based on the discrete characteristics of the residuals in the residual curve; obtaining slow fluctuation values ​​based on the distance characteristics between outlier residuals and adjacent residuals, and the data change characteristics of neighboring residuals; determining the degree of correlation between outlier residuals and residuals of other target products; identifying suspected oxygen pressure anomalies based on the slow fluctuation values ​​and the degree of correlation; and clustering these suspected oxygen pressure anomalies based on their distribution characteristics, obtaining the oxygen pressure anomaly time based on the area characteristics of the clusters, the density characteristics of the residuals within the clusters, and their distribution characteristics. This invention obtains oxygen pressure anomaly trend values ​​based on the distribution characteristics and anomaly degree of the oxygen pressure anomaly time, and adjusts and provides early warnings for oxygen pressure, improving the timeliness and accuracy of monitoring and ensuring the effective preparation of tellurium.

[0053] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0054] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for preparing refined tellurium from copper telluride slag, characterized in that, The method includes the following steps: Obtain the concentration-time curve of the target product during oxygen-pressure alkaline leaching; The concentration time-series curve is decomposed to obtain a residual curve; outlier residuals are obtained based on the discrete characteristics of the residuals in the residual curve; slow fluctuation values ​​are obtained based on the distance characteristics between the outlier residuals and their neighboring residuals, and the data change characteristics of the neighboring residuals of the outlier residuals; the degree of correlation of changes is obtained based on the difference in adjacent data changes between the outlier residuals and the residuals of other target products at the same time; the first oxygen pressure anomaly degree of the outlier residuals is obtained based on the slow fluctuation value and the degree of correlation of changes; and suspected oxygen pressure anomaly points are obtained based on the first oxygen pressure anomaly degree. Clustering is performed on the suspected oxygen pressure anomaly points to obtain different point clusters; the second oxygen pressure anomaly degree of the suspected oxygen pressure anomaly points is obtained based on the area characteristics, density characteristics, and distribution characteristics of the residual points within the clusters; the oxygen pressure anomaly time is obtained based on the second oxygen pressure anomaly degree; and the oxygen pressure anomaly trend value is obtained based on the distribution characteristics and anomaly degree of the oxygen pressure anomaly time. Adjustments and early warnings for oxygen pressure are issued based on the aforementioned abnormal oxygen pressure trends; The step of obtaining outlier residuals based on the discrete characteristics of residuals in the residual curve includes: Construct a residual mean line based on the average value of all residual points in the residual curve; calculate the average distance of all residual points from the residual mean line. and distance standard deviation The distance between the residual points in the residual curve and the residual mean line is not... Residual points within the range are considered outlier residual points; The step of obtaining the slow fluctuation value based on the distance characteristics between the outlier residual and its neighboring residuals, and the data change characteristics of the neighboring residuals of the outlier residual includes: The sum of the Euclidean distances between the outlier residual and its two preceding and following residuals is calculated to obtain the composite distance. The slope of the line connecting any residual to its preceding adjacent residual is calculated to obtain the degree of change of that residual. The average of the absolute values ​​of the differences in the degrees of change between all adjacent residuals within a preset preceding adjacent range of the outlier residual is calculated to obtain a first degree of change difference. The average of the absolute values ​​of the differences in the degrees of change between all adjacent residuals within a preset following adjacent range of the outlier residual is calculated to obtain a second degree of change difference. The reciprocal of the sum of the first degree of change difference and the second degree of change difference is calculated to obtain a third value. The product of the reciprocal of the composite distance and the third value is calculated to obtain the slow fluctuation value of the outlier residual. The step of clustering the suspected oxygen pressure anomaly points to obtain different point clusters includes: The K-means clustering algorithm is used to cluster all suspected oxygen pressure anomalies in the residual curve. The horizontal axis of the cluster space is time, and the vertical axis is the residual value of the suspected oxygen pressure anomaly.

2. The method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The step of obtaining the degree of correlation of changes based on the difference characteristics of adjacent data changes of the outlier residual points and the residual points of other target generators at the same time includes: The combined distance of the residual points of other target generators at the time of the outlier residual point is normalized to obtain a fourth value; the combined distance of the outlier residual point is normalized and the reciprocal of the absolute value of the difference with the fourth value is calculated to obtain the degree of correlation of the change of the outlier residual point.

3. The method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The step of obtaining the first oxygen pressure anomaly degree of outlier residuals based on the slow fluctuation value and the degree of correlation of the change includes: The slow fluctuation value and the degree of correlation of the change are normalized respectively, and the average value after normalization is calculated to obtain the first oxygen pressure anomaly degree of the outlier residual.

4. The method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The step of obtaining suspected oxygen pressure anomaly points based on the first degree of oxygen pressure anomaly includes: Outlier residuals whose oxygen pressure anomaly exceeds a preset first threshold are designated as suspected oxygen pressure anomaly points.

5. A method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The step of obtaining the second oxygen pressure anomaly degree of suspected oxygen pressure anomaly points based on the area characteristics of the point cluster, the density characteristics and distribution characteristics of the residual points within the cluster includes: The area of ​​the smallest circumcircle of the point cluster is calculated and normalized to obtain the area feature value; the average of the nearest neighbor distances of all suspected oxygen pressure anomaly points in the point cluster is calculated and normalized to obtain the interval feature value; the standard deviation of the nearest neighbor distances of all suspected oxygen pressure anomaly points in the point cluster is calculated and normalized to obtain the distribution discrete feature value; the average of the area feature value, the interval feature value, and the distribution discrete feature value is calculated to obtain the second oxygen pressure anomaly degree of all suspected oxygen pressure anomaly points in the point cluster.

6. The method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The step of obtaining the moment of oxygen pressure anomaly based on the second degree of oxygen pressure anomaly includes: The moment when the suspected oxygen pressure anomaly point exceeds the preset second threshold is defined as the oxygen pressure anomaly moment.

7. A method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The step of obtaining the oxygen pressure anomaly trend value based on the distribution characteristics and anomaly severity at the time of the oxygen pressure anomaly includes: Calculate the ratio of the number of abnormal oxygen pressure moments after the midpoint of the concentration time series curve to the number of abnormal oxygen pressure moments before the midpoint, to obtain the quantity ratio; calculate the average distance of the residual points of abnormal oxygen pressure moments after the midpoint from the residual mean line, to obtain the first distance; calculate the average distance of the residual points of abnormal oxygen pressure moments before the midpoint from the residual mean line, to obtain the second distance; calculate the ratio of the first distance to the second distance, to obtain the anomaly ratio; calculate the product of the quantity ratio and the anomaly ratio, to obtain the oxygen pressure anomaly trend value.

8. A method for preparing refined tellurium from copper telluride slag according to claim 1, characterized in that, The steps of adjusting and issuing early warnings for oxygen pressure based on the abnormal oxygen pressure trend include: The sum of the abnormal oxygen pressure trend value and constant 1 is calculated to obtain the adjustment coefficient; the product of the adjustment coefficient and the preset baseline oxygen pressure is calculated to obtain the adjusted oxygen pressure value; when the abnormal oxygen pressure trend value exceeds the preset abnormal trend threshold, an early warning is issued for the oxygen pressure alkaline leaching process.

Images