Multi-source data integration management method and system for pollution source monitoring
By identifying anomalous mutation points and growth trends in electrical data and pollutant concentration data, and adjusting the smoothing coefficient of the triple exponential smoothing algorithm, the problem of inaccurate pollutant concentration prediction caused by malfunctions in desulfurization wastewater purification equipment was solved, achieving more accurate pollutant concentration prediction and effective treatment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG HUANMAO AUTO-CONTROL TECH CO LTD
- Filing Date
- 2026-03-09
- Publication Date
- 2026-06-02
AI Technical Summary
The existing pollutant concentration prediction accuracy is low, making it impossible to achieve reliable and accurate treatment of industrial desulfurization wastewater. This is mainly due to the decrease in oxidation efficiency and the increase in pollutant concentration caused by abnormal operation of desulfurization wastewater purification equipment.
By identifying anomalous abrupt changes in the electrical data sequence, significant characteristics of power supply anomalies and trends in pollutant concentration growth are obtained. The smoothing coefficient in the triple exponential smoothing algorithm is adjusted, and the smoothing coefficient is adaptively adjusted to predict pollutant concentration based on the correlation between changes in electrical data and pollutant concentration data.
It improves the accuracy of pollutant concentration prediction, ensures reliable treatment of industrial desulfurization wastewater, enables timely detection and response to purification equipment malfunctions, and enhances purification efficiency.
Smart Images

Figure CN121786714B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology, specifically to a method and system for integrating and managing multi-source data for pollution source monitoring. Background Technology
[0002] When monitoring pollution discharge from factories, desulfurization wastewater treatment ponds are typically equipped with desulfurization wastewater purification equipment to treat industrial desulfurization wastewater and reduce pollutant concentrations. This equipment can be a combination of a heavy metal scavenging agent dosing system and a chemical oxidation system. The heavy metal scavenging agent dosing system adds specialized organic sulfide scavengers such as DTCRs to form more stable chelate precipitates with heavy metals, thereby reducing their concentration. The chemical oxidation system removes COD through oxidation, reducing its concentration. Alternatively, the desulfurization wastewater purification equipment can be a reverse osmosis or nanofiltration system to simultaneously remove heavy metals and COD. Pollutant detection equipment (such as a COD analyzer) is also installed at the industrial desulfurization wastewater effluent outlet to collect pollutant concentrations (e.g., COD concentration) in the purified wastewater. Based on the collected historical pollutant concentration data over a period of time, a cubic exponential smoothing algorithm is used to predict pollutant concentrations, and pollution control measures are implemented based on the predicted concentrations. However, the accuracy of this method heavily depends on the normal operation of the desulfurization wastewater purification equipment. If the equipment malfunctions, such as a sudden power shortage or a sudden drop in current, oxidation efficiency will decrease, leading to an increase in pollutant concentration at the industrial desulfurization wastewater effluent outlet. If the smoothing coefficient in the cubic exponential smoothing algorithm under normal operating conditions is used for prediction in this situation, the accuracy of pollutant concentration prediction will be affected, causing the prediction results to deviate from reality, thus failing to achieve reliable and accurate treatment of industrial desulfurization wastewater. Summary of the Invention
[0003] To address the technical problem of low accuracy in predicting pollutant concentrations, which hinders reliable and accurate treatment of industrial desulfurization wastewater, this invention aims to provide a multi-source data integration and treatment method and system for pollution source monitoring. The specific technical solution adopted is as follows:
[0004] In a first aspect of the present invention, a method for multi-source data integration and management for pollution source monitoring is provided, comprising:
[0005] Identify anomalous abrupt changes in the electrical data sequence of the desulfurization wastewater purification equipment within the current time period;
[0006] Based on the differences in electrical data before and after the abnormal mutation point, the significant characteristics of power supply anomalies in the desulfurization wastewater purification equipment are obtained.
[0007] Obtain the starting point of the upward trend in the pollutant concentration data sequence within the current time period, and determine the pollutant concentration growth trend of the pollutant concentration data after the starting point of the upward trend.
[0008] Based on the significant characteristics of the power supply anomaly, the trend of pollutant concentration growth, and the correlation between the changes in the electrical data sequence and the pollutant concentration data sequence, the smoothing coefficient in the triple exponential smoothing algorithm is adjusted to obtain the smoothing coefficient.
[0009] Based on the smoothing coefficient, a cubic exponential smoothing algorithm is used to predict pollutant concentrations, and the predicted pollutant concentrations are used to indicate pollution control measures.
[0010] In an exemplary embodiment, the process of obtaining the significant characteristics of the power supply anomaly includes:
[0011] Determine the number of electrical data points following the anomalous mutation point;
[0012] Determine the extent to which the mean electrical data before the anomalous abrupt change point exceeds the mean electrical data afterward;
[0013] The power supply anomaly significance characteristics are obtained based on the number of electrical data points and the degree of exceedance; the power supply anomaly significance characteristics are inversely correlated with the number of electrical data points and positively correlated with the degree of exceedance.
[0014] In one exemplary embodiment, the process of obtaining the pollutant concentration growth trend includes:
[0015] A straight line is fitted to the pollutant concentration data after the starting point of the upward trend to obtain a fitted straight line;
[0016] The slope of the fitted straight line is obtained to determine the increasing trend of the pollutant concentration.
[0017] In one exemplary embodiment, the degree of correlation of the changes is the goodness of fit between the electrical data sequence and the pollutant concentration data sequence.
[0018] In an exemplary embodiment, the process of obtaining the smoothing coefficient includes:
[0019] Based on the significant characteristics of the power supply anomaly, the trend of pollutant concentration growth, and the degree of correlation with the change, a smoothing adjustment coefficient is obtained; the smoothing adjustment coefficient is positively correlated with the significant characteristics of the power supply anomaly and the trend of pollutant concentration growth, and negatively correlated with the degree of correlation with the change.
[0020] The smoothing coefficient is obtained by adjusting the smoothing adjustment coefficient; the smoothing coefficient is positively correlated with the smoothing adjustment coefficient.
[0021] In an exemplary embodiment, adjusting the smoothing coefficient according to the smoothing adjustment coefficient includes:
[0022] Determine the smoothing coefficient adjustment range value, wherein the smoothing coefficient adjustment range value is the difference between the maximum preset smoothing coefficient and the minimum preset smoothing coefficient;
[0023] The smoothing adjustment amount is obtained by multiplying the smoothing adjustment coefficient by the smoothing coefficient adjustment range value.
[0024] The smoothing coefficient is obtained by calculating the sum of the smoothing coefficient adjustment amount and the minimum preset smoothing coefficient.
[0025] In an exemplary embodiment, after obtaining the predicted pollutant concentration, the multi-source data integration and governance method for pollution source monitoring further includes:
[0026] Determine whether the predicted pollutant concentration exceeds the preset emission standard concentration. If it does, output a pollution control early warning signal.
[0027] In an exemplary embodiment, the process of obtaining the starting point of the upward trend includes:
[0028] Determine the first-order difference sequence of the pollutant concentration data sequence;
[0029] Determine candidate times for the pollutant concentration data sequence, wherein the candidate times are the times in the first-order difference sequence that are greater than a preset pollutant concentration change threshold;
[0030] Obtain the number of candidate moments within a local time interval after each candidate moment;
[0031] The target time is taken as the starting point of the upward trend. The target time is the first candidate time in the time sequence when the number of candidate times is greater than a preset threshold.
[0032] In one exemplary embodiment, the anomalous mutation point is obtained by processing the electrical data sequence using the Pettitt mutation point detection algorithm.
[0033] In a second aspect of the present invention, a multi-source data integration and management system for pollution source monitoring is provided, comprising: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; the processor is used to implement the above-described multi-source data integration and management method for pollution source monitoring when the program instructions are executed.
[0034] This invention offers the following advantages: By analyzing the differences in electrical data before and after the anomalous abrupt change, the significant characteristics of power supply anomalies in the desulfurization wastewater purification equipment are obtained. These significant characteristics quantify the degree of operational anomalies in the equipment, providing a basis for subsequent assessment of the impact of the anomalies on purification efficiency. The invention also determines the pollutant concentration growth trend after the starting point of the upward trend, capturing the initial moment and trend of the increase in pollutant concentration caused by the anomaly in the desulfurization wastewater purification equipment, directly reflecting the actual impact of the anomaly on the pollutant concentration in the desulfurization wastewater. Furthermore, by combining the significant characteristics of the power supply anomalies (cause) with the pollutant growth trend (result), and analyzing the correlation between electrical data and pollutant concentration changes, the causal relationship between power supply anomalies and pollutant concentration is confirmed, thereby improving the accuracy of the smoothing coefficient. Finally, the invention outputs a predicted pollutant concentration that more closely reflects actual changes, improving the accuracy of pollutant concentration prediction and ultimately achieving reliable and accurate treatment of industrial desulfurization wastewater. Attached Figure Description
[0035] Figure 1 This is a flowchart of a multi-source data integration and management method for pollution source monitoring provided in one embodiment of the present invention;
[0036] Figure 2 This is a flowchart illustrating the acquisition of significant power supply anomaly features according to an embodiment of the present invention;
[0037] Figure 3 This is a flowchart illustrating the process of obtaining the starting point of an upward trend according to an embodiment of the present invention;
[0038] Figure 4 This is a flowchart illustrating the process of obtaining the smoothing coefficient according to an embodiment of the present invention. Detailed Implementation
[0039] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0040] 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 this invention pertains. All data and information collected in this application have been obtained with full consent.
[0041] This embodiment provides a multi-source data integration and governance method for pollution source monitoring. The applicable application scenario is as follows: In a factory's wastewater discharge system, desulfurization wastewater treatment equipment is installed in the factory's desulfurization wastewater treatment pond to treat industrial desulfurization wastewater and reduce its pollutant concentration. Pollutant detection equipment is installed at the industrial desulfurization wastewater outlet of the factory's desulfurization wastewater treatment pond to collect the pollutant concentration in the purified wastewater. This embodiment provides a multi-source data integration and governance method for pollution source monitoring, which uses a cubic exponential smoothing algorithm to predict pollutant concentrations based on collected historical pollutant concentration data over a period of time, and then implements pollution control measures based on the predicted pollutant concentrations.
[0042] The hardware system configured for a multi-source data integration and governance method for pollution source monitoring provided in this embodiment includes: an electrical parameter detection device, a pollutant detection device, and a data processing device. The data processing device is signal-connected to the electrical parameter detection device and the pollutant detection device. The signal connection method can be a wired connection or a wireless connection such as 5G or WiFi, without any special restrictions.
[0043] Electrical parameter detection equipment is used to monitor the electrical data of the desulfurization wastewater purification equipment, thereby determining whether the equipment is operating abnormally, such as whether there is an abnormal drop in current. The sampling frequency of the electrical parameter detection equipment and the pollutant detection equipment is set according to actual needs, such as once per second. Furthermore, the electrical parameter detection equipment and the pollutant detection equipment sample synchronously.
[0044] In one exemplary embodiment, the electrical parameter detection device is specifically a current sensor, used to detect the current of the desulfurization wastewater purification equipment by being installed at the power supply terminal. There can be only one type of pollutant detection device, or multiple types can be used. When multiple types are used, the subsequent prediction processes for different pollutant concentrations are independent and do not interfere with each other. The pollutant detection device can be a COD detector, a heavy metal concentration detector, etc. The COD detector is used to collect the COD concentration in the purified wastewater, and the heavy metal concentration detector is used to collect the heavy metal concentration in the purified wastewater. This embodiment uses a COD detector as an example of pollutant detection device, and the pollutant concentration is taken as the COD concentration.
[0045] It should be understood that if there are missing values in the data collected during the process of collecting the current and pollutant concentration of the desulfurization wastewater purification equipment, linear interpolation can be used to fill in the missing values in the collected data before subsequent data processing.
[0046] like Figure 1 As shown in the figure, the multi-source data integration and governance method for pollution source monitoring provided in this embodiment includes the following steps:
[0047] Step S1: Identify abnormal abrupt changes in the electrical data sequence of the desulfurization wastewater purification equipment within the current time period;
[0048] Step S2: Based on the differences in electrical data before and after the abnormal mutation point, obtain the significant characteristics of power supply anomalies in the desulfurization wastewater purification equipment;
[0049] Step S3: Obtain the starting point of the upward trend in the pollutant concentration data sequence within the current time period, and determine the pollutant concentration growth trend after the starting point of the upward trend.
[0050] Step S4: Based on the significant characteristics of power supply anomalies, the growth trend of pollutant concentrations, and the correlation between changes in electrical data sequences and pollutant concentration data sequences, adjust the smoothing coefficients in the triple exponential smoothing algorithm.
[0051] Step S5: Based on the smoothing coefficient, a cubic exponential smoothing algorithm is used to predict the pollutant concentration, and the predicted pollutant concentration is used to indicate pollution control measures.
[0052] The following explanation, in conjunction with the accompanying drawings, details each step.
[0053] Step S1: Identify abnormal mutation points in the electrical data sequence of the desulfurization wastewater purification equipment within the current time period.
[0054] This embodiment predefines a current time period, and the end time of the current time period can be the current moment. Therefore, the predicted time can be the next moment after the current moment. The duration of the current time period, i.e., the number of moments it contains, is set according to actual needs. For example, while ensuring the feasibility of the prediction, the duration of the current time period can be set to 10 minutes, then the number of moments it contains is 600.
[0055] The electrical data of the desulfurization wastewater purification equipment at various moments within the current time period is obtained, and then arranged chronologically to obtain the electrical data sequence of the desulfurization wastewater purification equipment within the current time period. Since the electrical data is specifically current, the current of the desulfurization wastewater purification equipment at various moments within the current time period is obtained, and then arranged chronologically to obtain the current sequence of the desulfurization wastewater purification equipment within the current time period.
[0056] It should be understood that the current sequence of the desulfurization wastewater purification equipment in the current time period is the basis for subsequent data processing. Therefore, the desulfurization wastewater purification equipment should be in a non-shutdown state in the current time period. If there is a period in the current time period in which the desulfurization wastewater purification equipment is shut down, then no further data processing will be performed on the current time period.
[0057] Under normal operating conditions, the COD and heavy metal concentrations in the discharged desulfurization wastewater tend to stabilize and meet relevant emission standards. However, equipment malfunctions are often sudden. If the equipment experiences a sudden power shortage, leading to a decrease in oxidation efficiency, the pollutant removal effect will decline, and the COD and heavy metal concentrations detected at the outlet will increase significantly (i.e., a sudden change in the monitored concentration). Therefore, it is necessary to first identify the abnormal mutation points in the current sequence of the desulfurization wastewater purification equipment within the current time period. Abnormal mutation points characterize the moments when the current abnormally decreases. In an exemplary embodiment, abnormal mutation points are obtained by processing the current sequence using the Pettitt mutation point detection algorithm. The detection principle of the Pettitt mutation point detection algorithm is as follows: for each possible segmentation point in the current sequence, it is determined whether the preceding and following subsequences come from the same distribution. If an abnormal mutation point exists, the difference in distribution between the two subsequences is maximized. Since obtaining abnormal mutation points in the current sequence, i.e., data points with abnormally decreasing current, using the Pettitt mutation point detection algorithm is a conventional technique, the implementation process is briefly described below:
[0058] Sort the current values in the current sequence from smallest to largest, and assign a rank to each current value. The number of current values in the current sequence is set to n. If some current values are repeated, their average rank is used. This eliminates the influence of the original dimensions and makes the algorithm robust to outliers.
[0059] For each possible split point t, calculate the Mann-Whitney statistic. , The formula reflects the trend that the value of the preceding subsequence is less than the value of the following subsequence:
[0060] ;
[0061] in, Let represent the Mann-Whitney statistic for the split point at time t.
[0062] here, This represents the cumulative sum of the ranks of the first t data points. The formula is derived from the Mann-Whitney statistic and is computationally efficient.
[0063] The positive and negative signs indicate the direction of change: if A negative value indicates that the preceding subsequence value tends to be greater than the following subsequence value, corresponding to a decrease in current; a positive value corresponds to an increase. Abnormal current decreases should be noted in negative values, and calculations should be performed accordingly. The minimum value (i.e., the most negative value) is because the mean of the preceding subsequence is higher than that of the following subsequence when the current decreases, making... Negative. Location of the mutation point. satisfy:
[0064] ;
[0065] in, express The minimum value. Recording makes Minimum index , as an abnormal mutation point.
[0066] It should be understood that, in addition to using the Pettitt mutation detection algorithm, this embodiment may also select other existing mutation detection algorithms as needed.
[0067] Step S2: Based on the differences between the electrical data before and after the abnormal mutation point, obtain the significant characteristics of the power supply anomaly of the desulfurization wastewater purification equipment.
[0068] After obtaining the abnormal abrupt change points in the current sequence of the desulfurization wastewater purification equipment within the current time period, the significant characteristics of the power supply anomalies of the desulfurization wastewater purification equipment are obtained based on the difference between the currents before and after the abnormal abrupt change points. In an exemplary embodiment, such as... Figure 2 As shown, the following is a specific process for obtaining significant features of power supply anomalies:
[0069] Step S21: Determine the abnormal mutation point and the number of electrical data points thereafter.
[0070] The position of the abnormal mutation point in the current sequence can reflect the impact of insufficient power supply to the desulfurization wastewater purification equipment on the prediction time. The closer the abnormal mutation point is to the prediction time, that is, the fewer the time intervals between it and the prediction time, the closer the time of insufficient power supply to the desulfurization wastewater purification equipment is to the prediction time.
[0071] The period before the anomalous abrupt change represents the stable operation phase of the desulfurization wastewater purification equipment, while the period after the anomalous abrupt change represents the phase of insufficient power supply to the equipment. Therefore, by obtaining the number of times following the anomalous abrupt change in the current sequence (i.e., the number of subsequent electrical data points) and adding it to the anomalous abrupt change itself, we obtain the number of electrical data points following the anomalous abrupt change. The number of electrical data points following the anomalous abrupt change reflects the proximity of the insufficient power supply to the predicted time. Fewer electrical data points following the anomalous abrupt change indicate that the insufficient power supply is closer to the predicted time, resulting in a higher impact on the predicted pollutant concentration and a stronger significance of the power supply anomaly. These two factors are inversely correlated. Furthermore, the fewer electrical data points following the anomalous abrupt change, the more attention should be paid to the monitoring data in the current time period when using the triple exponential smoothing algorithm for prediction.
[0072] To facilitate subsequent data calculations, this embodiment obtains the total number of current data points in the current sequence, and then calculates the ratio of the number of electrical data points at and after the abnormal abrupt change point to the total number of current data points. This ratio is used as the proportion of electrical data points at and after the abnormal abrupt change point. The proportion of electrical data points at and after the abnormal abrupt change point characterizes the proximity of the abnormal power supply change moment of the desulfurization wastewater purification equipment to the predicted moment. Furthermore, by calculating the proportion, the number of electrical data points at and after the abnormal abrupt change point can be normalized to a value range of 0-1, eliminating the influence of dimensions and facilitating subsequent data processing.
[0073] Step S22: Determine the extent to which the mean electrical data before the anomalous mutation point exceeds the mean electrical data afterward.
[0074] The current applied to desulfurization wastewater purification equipment reflects the severity of power supply insufficiency; the lower the current, the more severe the power supply insufficiency. A higher degree of power supply insufficiency has a greater impact on the purification effect of desulfurization wastewater.
[0075] In calculating the current sequence, the average current value at and after the abnormal abrupt change point is calculated as the first current mean, which represents the overall current level after the abnormal current drop. Simultaneously, the average current value before the abnormal abrupt change point is calculated as the second current mean, which represents the overall current level before the current anomaly, i.e., under normal conditions.
[0076] The degree to which the average second current exceeds the average first current is determined. A higher degree of exceedance indicates a more severe power supply insufficiency, a greater impact on the purification effect of desulfurization wastewater, and a stronger indication of power supply anomalies in the desulfurization wastewater purification equipment; the two are positively correlated. In an exemplary embodiment, a specific quantification method for the degree of exceedance is given below:
[0077] ;
[0078] in, Indicates exceeding a certain degree. This represents the average value of the first current. This represents the average value of the second current.
[0079] Since the first current mean represents the overall current level after an abnormal current drop, then ,therefore, Greater than 0. In extreme cases, if If there is no abnormal drop in current, then let = Additionally, if A value of 0 indicates that the average current of the desulfurization wastewater purification equipment before the abnormal change point in the current time period is 0, and the equipment is in a shutdown state. As analyzed above, if subsequent predictions are made based on the current sequence in the current time period, the prediction accuracy will be severely affected. Therefore, no further prediction processing is performed. It should be noted that, in order to avoid the problem of the formula being meaningless due to a denominator of 0, when a denominator of 0 appears in any formula involved in this embodiment of the invention, a preset minimum positive number is used to replace the corresponding denominator in the calculation. In this embodiment of the invention, the preset minimum positive number is set to 0.01, which can be adjusted according to the specific implementation environment, and will not be further elaborated here.
[0080] The degree of deviation represents the difference rate between abnormal data and normal data. In this embodiment, the current value at and after the abnormal abrupt change point is taken as the abnormal current, and the current value before the abnormal abrupt change point is taken as the normal current value. The degree of deviation is obtained using the principle of rate of change to characterize the degree of power supply insufficiency of the desulfurization wastewater purification equipment. Moreover, through the above processing method, the numerical range of the degree of deviation can be kept within 0-1, and the influence of dimensions is eliminated, which facilitates subsequent data processing.
[0081] Step S23: Based on the number of electrical data points and the degree of exceedance, obtain the significant characteristics of power supply anomalies.
[0082] Based on the proportion of abnormal abrupt change points and subsequent electrical data points, as well as the degree of exceedance, the significant characteristics of power supply anomalies in desulfurization wastewater purification equipment are obtained, which are used to characterize the significance of power supply anomalies in desulfurization wastewater purification equipment. Based on the above logical analysis, a specific calculation method for the significant characteristics of power supply anomalies is given below:
[0083] ;
[0084] in, This indicates a significant characteristic of abnormal power supply to the desulfurization wastewater purification equipment. Indicates exceeding a certain degree. This indicates the percentage of electrical data points that occur at or after an abnormal abrupt change.
[0085] In the calculation formula for the significant characteristics of power supply anomalies, the data from two aspects are integrated by averaging. The smaller the proportion of the number of abnormal change points and subsequent electrical data points, the closer the time of the current power supply anomaly is to the predicted time, the greater the power supply abruptness, the higher the degree of excess, the more abnormal the power supply of the desulfurization wastewater purification equipment, and the stronger the significant characteristics of the power supply anomaly of the desulfurization wastewater purification equipment.
[0086] It should be understood that if there are no abnormal abrupt changes, it indicates that the desulfurization wastewater purification equipment is operating normally within the current time period, and the power supply anomaly significance characteristic of the desulfurization wastewater purification equipment is set to 0. Moreover, if the power supply anomaly significance characteristic of the desulfurization wastewater purification equipment is 0, it indicates that the desulfurization wastewater purification equipment is powered normally within the current time period, and therefore there is no need to adjust the smoothing coefficient; the smoothing adjustment coefficient can be directly set to 0.
[0087] Step S3: Obtain the starting point of the upward trend in the pollutant concentration data sequence within the current time period, and determine the pollutant concentration growth trend after the starting point of the upward trend.
[0088] The COD concentration at the outlet of industrial desulfurization wastewater is obtained at various times within the current time period, and then arranged in chronological order to obtain the COD concentration sequence within the current time period.
[0089] When the power supply to the desulfurization wastewater purification equipment is abnormal, it usually leads to insufficient COD purification, resulting in a higher COD concentration in subsequent monitoring. Therefore, upon detecting a power supply abnormality, the smoothing coefficient of the cubic exponential smoothing algorithm needs to be adjusted adaptively in a timely manner to ensure that the prediction focuses more on the COD concentration in the current time period. Simply adjusting the smoothing coefficient based on the significance of the power supply abnormality ignores actual conditions, such as the transmission delay of wastewater in the pipeline. That is, the high-COD-concentration wastewater output after a power supply abnormality usually takes a certain amount of time to reach the industrial desulfurization wastewater outlet. At this time, the COD detector has not yet collected the high-concentration COD. If the smoothing coefficient is directly amplified, it will lead to inaccurate COD concentration prediction. Therefore, this embodiment requires adjusting the smoothing coefficient only after an upward trend in COD concentration is detected at the industrial desulfurization wastewater outlet.
[0090] Abnormal power supply to desulfurization wastewater purification equipment often leads to incomplete decomposition of large amounts of organic flocculents, which gradually accumulate at the effluent outlet of industrial desulfurization wastewater, causing a gradual increase in the monitored COD concentration. Therefore, the moment when the COD concentration begins to show an upward trend is taken as the starting point of the upward trend in the COD concentration sequence for the current time period. After the starting point of the upward trend, it indicates that the COD concentration is affected by the abnormal power supply to the desulfurization wastewater purification equipment, and the growth trend is relatively rapid. In an exemplary embodiment, such as... Figure 3 As shown, the following is a specific process for obtaining the starting point of an upward trend:
[0091] Step S31: Determine the first-order difference sequence of the pollutant concentration data sequence.
[0092] Obtain the first-order difference sequence of COD concentration within the current time period. Each element in the first-order difference sequence is the difference between the next and previous COD concentrations. For any element in the first-order difference sequence, a positive value indicates that the corresponding two adjacent COD concentrations are increasing; a negative value indicates that the corresponding two adjacent COD concentrations are decreasing; and a value of 0 indicates that the corresponding two adjacent COD concentrations remain unchanged.
[0093] Step S32: Determine candidate times for the pollutant concentration data sequence.
[0094] This embodiment presets a pollutant concentration change threshold. This threshold is used to compare with each element in the first-order difference sequence to determine whether the value of each element in the first-order difference sequence is large. A large value indicates that the increase in COD concentration between two adjacent elements is large. This pollutant concentration change threshold can be a preset value (specifically a positive value) with the dimension of COD concentration. It is directly compared with each element in the first-order difference sequence to determine the elements in the first-order difference sequence that are greater than the pollutant concentration change threshold. For any element that is greater than the pollutant concentration change threshold, the second of the two adjacent time points is taken as a candidate time point, thus obtaining several candidate time points in the COD concentration sequence. In this embodiment, the preset pollutant concentration change threshold is set to 10 mg / L, which can be adjusted according to the specific implementation environment.
[0095] It should be understood that, based on the above analysis, if there are no candidate moments in the COD concentration sequence, it means that there is no starting point of the upward trend in the COD concentration sequence, that is, there is no obvious upward trend in the COD concentration. Therefore, there is no abnormality in the COD concentration in the current time period, and no further processing is required. The smoothing adjustment coefficient in the subsequent process is directly set to 0.
[0096] Step S33: Obtain the number of candidate moments within the local time interval after each candidate moment.
[0097] It should be understood that since the starting point of the upward trend indicates the beginning of a relatively obvious upward trend in COD concentration, and power supply abnormalities in desulfurization wastewater purification equipment usually lead to a continuous increase in COD concentration, the upward trend in COD concentration is relatively obvious after the starting point of the upward trend within the current time period. Therefore, for example, multiple candidate moments will appear consecutively within a period of time after the starting point of the upward trend. Thus, for any candidate moment, this embodiment presets a local time interval, that is, a local time interval with the candidate moment as the starting point and a preset number of moments as the duration. The preset number of moments is set according to actual needs, such as 20. The local time interval of the candidate moment is determined and referred to as the local time interval after the candidate moment. The number of candidate moments within the local time interval of the candidate moment is obtained, thereby obtaining the number of candidate moments within the local time interval of each candidate moment.
[0098] Step S34: Take the target time as the starting point of the upward trend.
[0099] For any given candidate moment, the higher the number of candidate moments within its local time interval, the more likely it is to be the starting point of an upward trend. Therefore, this embodiment presets a quantity threshold, which is set by the implementer based on the duration of the local time interval and the judgment requirements. For example, if the duration of the local time interval includes 20 moments, the quantity threshold can be set to 15. The number of candidate moments within the local time interval of each candidate moment is compared with the preset quantity threshold to obtain each candidate moment that exceeds the preset quantity threshold. The first candidate moment in time sequence among all candidate moments exceeding the preset quantity threshold is taken as the target moment. The target moment is then taken as the starting point of the upward trend.
[0100] After obtaining the starting point of the upward trend in the COD concentration sequence within the current time period, the pollutant concentration growth trend of COD concentration after the starting point of the upward trend is determined. In an exemplary embodiment, the COD concentration at each time point after the starting point of the upward trend in the COD concentration sequence is obtained, and then mapped onto a two-dimensional coordinate system with time series as the horizontal axis and COD concentration as the vertical axis. A straight line is then fitted to the COD concentration at each time point after the starting point of the upward trend to obtain a fitted straight line. The slope of this fitted straight line is then obtained, and the pollutant concentration growth trend of COD concentration after the starting point of the upward trend is obtained based on the slope. If the slope is positive, it indicates that the COD concentration at each time point after the starting point of the upward trend shows an increasing trend, and the larger the slope, the more obvious the increasing trend. If the slope is negative, it indicates that the COD concentration at each time point after the starting point of the upward trend shows a decreasing trend. If the slope is 0, it indicates that the COD concentration at each time point after the starting point of the upward trend shows a stable changing trend. Therefore, in this embodiment, if the slope is negative or 0, the pollutant concentration growth trend is directly set to 0. When the slope is positive, the arctangent operation is performed on the slope to obtain the angle value corresponding to the slope. It should be understood that the numerical range of the angle value corresponding to the slope is 0-π / 2. Then, the ratio of the angle value corresponding to the slope to π / 2 is calculated to normalize the angle value corresponding to the slope and eliminate the influence of dimensions. The normalized result of the angle value corresponding to the slope is used as the pollutant concentration growth trend of COD concentration after the starting point of the upward trend.
[0101] Step S4: Based on the significant characteristics of power supply anomalies, the growth trend of pollutant concentrations, and the correlation between changes in electrical data sequences and pollutant concentration data sequences, adjust the smoothing coefficients in the triple exponential smoothing algorithm.
[0102] It should be understood that during stable operation of the desulfurization wastewater purification equipment, the COD removal rate is high, and the COD concentration at the industrial desulfurization wastewater effluent outlet is relatively stable, within the relevant emission standards. The current of the desulfurization wastewater purification equipment is also relatively stable. Therefore, there is a clear correlation between current changes and COD concentration changes. However, when the power supply to the desulfurization wastewater purification equipment is abnormal, the current drops significantly, the COD removal rate in the desulfurization wastewater becomes lower, and a large amount of organic flocculent matter accumulates at the industrial desulfurization wastewater effluent outlet, causing the COD concentration to gradually increase. At this time, the correlation between current changes and COD concentration changes is poor. Therefore, it is necessary to obtain the degree of correlation between the current sequence and the COD concentration sequence within the current time period. In an exemplary embodiment, the degree of correlation specifically refers to the goodness of fit between the current sequence and the COD concentration sequence within the current time period. Goodness of fit reflects the linear correlation between the current data of the oxidation equipment and the COD concentration within the current time period. The value of goodness of fit ranges from 0 to 1 and is dimensionless. A higher goodness of fit indicates a stronger correlation between the current and COD concentration of the desulfurization wastewater purification equipment during the change process, suggesting that the equipment is in a stable operating state. Conversely, a lower goodness of fit indicates a weaker correlation, suggesting that the equipment is in a power supply abnormality state. The calculation method for goodness of fit is a conventional technique, and the specific implementation will not be elaborated further. As another implementation method, the degree of correlation can also be obtained by calculating the Pearson correlation coefficient, etc. Taking the Pearson correlation coefficient as an example, the Pearson correlation coefficient is normalized, and the normalized result is the degree of correlation. The normalization method can be (Pearson correlation coefficient + 1) / 2.
[0103] The lower the goodness of fit between the current sequence and the COD concentration sequence within the current time period, the more attention should be paid to the COD concentration data of the current time period when using the triple exponential smoothing algorithm for COD concentration prediction. Therefore, the smoothing coefficient should be appropriately increased to improve prediction accuracy. Similarly, the higher the significance of power supply anomalies in the desulfurization wastewater purification equipment, the more attention should be paid to the COD concentration data of the current time period when using the triple exponential smoothing algorithm for COD concentration prediction. Therefore, the smoothing coefficient should be appropriately increased to improve prediction accuracy. Furthermore, the higher the pollutant concentration growth trend of COD concentration after the starting point of the upward trend (i.e., the faster the growth rate of COD concentration after the starting point of the upward trend), the more attention should be paid to the COD concentration data of the current time period when using the triple exponential smoothing algorithm for COD concentration prediction. Therefore, the smoothing coefficient should be appropriately increased to improve prediction accuracy.
[0104] In one exemplary embodiment, such as Figure 4 As shown, the following is a specific process for obtaining the smoothing coefficient:
[0105] Step S41: Obtain the smoothing adjustment coefficient based on the characteristics of power supply anomalies, the growth trend of pollutant concentrations, and the degree of correlation between changes.
[0106] The smoothing adjustment coefficient characterizes the adjustment magnitude of the smoothing coefficient. A larger smoothing adjustment coefficient indicates a larger adjustment magnitude, and thus, the smoothing coefficient and the smoothing adjustment coefficient are positively correlated. Therefore, based on the above logical analysis, the lower the goodness of fit between the current sequence and the COD concentration sequence within the current time period, the larger the smoothing adjustment coefficient, indicating an inverse correlation between the two; the higher the significance of power supply anomalies in the desulfurization wastewater purification equipment, the larger the smoothing adjustment coefficient, indicating a positive correlation; and the higher the pollutant concentration growth trend of COD concentration after the starting point of the upward trend, the larger the smoothing adjustment coefficient, indicating a positive correlation. Based on this, a specific calculation method for the smoothing adjustment coefficient is given below:
[0107] ;
[0108] in, Indicates the smoothing adjustment coefficient. This indicates the goodness of fit between the current sequence and the COD concentration sequence within the current time period. This indicates the pollutant concentration growth trend of COD concentration after the starting point of the upward trend.
[0109] The three parameters are combined and calculated by averaging to obtain the smoothing adjustment coefficient. The larger the smoothing adjustment coefficient, the more obvious the sudden fluctuation of COD concentration in the current time period. Therefore, the triple exponential smoothing algorithm should pay more attention to the COD concentration in the current time period and pay less attention to the COD concentration in the historical time periods before the current time period. In order to ensure that the predicted COD concentration in the next moment is closer to the true value, a larger smoothing adjustment coefficient should be set adaptively.
[0110] Step S42: Adjust the smoothing coefficient according to the smoothing adjustment coefficient to obtain the smoothing coefficient.
[0111] When a power supply failure occurs in the desulfurization wastewater purification equipment, the COD concentration of the discharged wastewater will be too high. If the original smoothing coefficient is used, the predicted COD concentration by the cubic exponential smoothing algorithm will be too low, making it difficult to detect the power supply failure in a timely manner and affecting the purification efficiency. Therefore, an adaptive smoothing coefficient is obtained by adjusting the smoothing coefficient based on the original smoothing coefficient. The smoothing coefficient can characterize the dependence of COD concentration prediction on the COD concentration of the current time period and the COD concentration of previous historical time periods.
[0112] In one exemplary embodiment, this embodiment determines a maximum preset smoothing coefficient. With minimum preset smoothing coefficient Maximum preset smoothness coefficient Greater than the minimum preset smoothing coefficient Maximum preset smoothness coefficient With minimum preset smoothing coefficient The specific value is set according to actual needs. In an exemplary embodiment, the maximum preset smoothness coefficient is... Set to 0.3, minimum preset smoothing coefficient Set to 0.1. Calculate the maximum preset smoothing coefficient. With minimum preset smoothing coefficient The difference is used as the smoothing coefficient adjustment range value. Then, the product of the smoothing adjustment coefficient and the smoothing coefficient adjustment range value is calculated, and the result is used as the smoothing coefficient adjustment amount. Finally, the smoothing coefficient adjustment amount and the minimum preset smoothing coefficient are calculated. The sum of the values is used as the smoothing coefficient.
[0113] It should be understood that, based on prior knowledge of interval mapping, in order to map a parameter to a fixed interval, the interval mapping of the parameter is achieved by multiplying a mapping coefficient by the size of the fixed interval (the maximum value of the fixed interval minus the minimum value) and adding the minimum value of the interval. In this embodiment, the smoothing adjustment coefficient is used as the mapping coefficient, multiplied by the size of the smoothing coefficient adjustment range value, and added to the minimum preset smoothing coefficient to achieve adaptive adjustment of the smoothing coefficient of the triple exponential smoothing algorithm.
[0114] In addition, in the early stages of prediction, if the actual duration of the current time period does not meet the duration requirement of the current time period mentioned above, this embodiment can default the smoothing coefficient to the average of the maximum preset smoothing coefficient and the minimum preset smoothing coefficient.
[0115] The above are the smoothing coefficients corresponding to COD concentrations. For other types of pollutant concentrations, such as heavy metal concentrations, the smoothing coefficients are obtained in the same way.
[0116] Step S5: Based on the smoothing coefficient, a cubic exponential smoothing algorithm is used to predict the pollutant concentration, and the predicted pollutant concentration is used to indicate pollution control measures.
[0117] The smoothing coefficient corresponding to the COD concentration in the current time period is configured into the triple exponential smoothing algorithm, and the COD concentration sequence in the current time period is input into the triple exponential smoothing algorithm to predict the COD concentration at the prediction time, thus obtaining the predicted COD concentration at the prediction time.
[0118] Then, pollution control is carried out based on the predicted COD concentration at the predicted time. In an exemplary embodiment, this embodiment presets an emission standard concentration for COD. This emission standard concentration can be determined by relevant standards or based on the overall level of historical COD concentrations. It is then determined whether the predicted COD concentration at the predicted time exceeds the preset emission standard concentration. If it does, a pollution control early warning signal is output to remind staff to check for abnormalities in a timely manner, thus achieving multi-source data integration and control for pollution source monitoring. In addition, while outputting the pollution control early warning signal, this embodiment can also output a significant power supply anomaly characteristic of the desulfurization wastewater purification equipment, which is used to simultaneously prompt staff to pay attention to the operating status of the desulfurization wastewater purification equipment and check its power supply status in a timely manner.
[0119] Similarly, the smoothing coefficient corresponding to the heavy metal concentration in the current time period is configured into the cubic exponential smoothing algorithm, and the heavy metal concentration sequence in the current time period is input into the cubic exponential smoothing algorithm to predict the heavy metal concentration at the prediction time, thus obtaining the predicted heavy metal concentration at the prediction time. Then, pollution control is carried out based on the predicted heavy metal concentration at the prediction time. In an exemplary embodiment, this embodiment focuses on heavy metals and presets an emission standard concentration, which can be determined by relevant standards. It determines whether the predicted heavy metal concentration at the prediction time exceeds the preset emission standard concentration. If it does, a pollution control early warning signal is output to remind staff to check for abnormalities in a timely manner, thereby realizing multi-source data integration and control for pollution source monitoring.
[0120] This embodiment also provides a multi-source data integration and governance system for pollution source monitoring, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above-described embodiment of the multi-source data integration and governance method for pollution source monitoring when the program instructions are executed.
[0121] In one exemplary embodiment, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the above embodiments of the multi-source data integration and governance method for pollution source monitoring.
[0122] 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.
[0123] 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 multi-source data integration and governance method for pollution source monitoring, characterized in that, include: Identify anomalous abrupt changes in the electrical data sequence of the desulfurization wastewater purification equipment within the current time period; Based on the differences in electrical data before and after the abnormal mutation point, the significant characteristics of power supply anomalies in the desulfurization wastewater purification equipment are obtained. Obtain the starting point of the upward trend in the pollutant concentration data sequence within the current time period, and determine the pollutant concentration growth trend of the pollutant concentration data after the starting point of the upward trend. Based on the significant characteristics of the power supply anomaly, the trend of pollutant concentration growth, and the correlation between the changes in the electrical data sequence and the pollutant concentration data sequence, the smoothing coefficient in the triple exponential smoothing algorithm is adjusted to obtain the smoothing coefficient. Based on the smoothing coefficient, a triple exponential smoothing algorithm is used to predict pollutant concentrations, and the predicted pollutant concentrations are used to indicate pollution control measures. The process of obtaining the smoothing coefficient includes: Based on the significant characteristics of the power supply anomaly, the trend of pollutant concentration growth, and the degree of correlation with the change, a smoothing adjustment coefficient is obtained; the smoothing adjustment coefficient is positively correlated with the significant characteristics of the power supply anomaly and the trend of pollutant concentration growth, and negatively correlated with the degree of correlation with the change. The smoothing coefficient is obtained by adjusting the smoothing adjustment coefficient; the smoothing coefficient is positively correlated with the smoothing adjustment coefficient. The degree of correlation of the changes refers to the goodness of fit between the electrical data sequence and the pollutant concentration data sequence. The goodness of fit reflects the degree of linear correlation between the electrical data sequence and the pollutant concentration data sequence within the current time period.
2. The multi-source data integration and governance method for pollution source monitoring as described in claim 1, characterized in that, The process of obtaining the significant features of the power supply anomaly includes: Determine the number of electrical data points following the anomalous mutation point; Determine the extent to which the mean electrical data before the anomalous abrupt change point exceeds the mean electrical data afterward; The power supply anomaly significance characteristics are obtained based on the number of electrical data points and the degree of exceedance; the power supply anomaly significance characteristics are inversely correlated with the number of electrical data points and positively correlated with the degree of exceedance.
3. The multi-source data integration and governance method for pollution source monitoring as described in claim 1, characterized in that, The process of obtaining the pollutant concentration growth trend includes: A straight line is fitted to the pollutant concentration data after the starting point of the upward trend to obtain a fitted straight line; The slope of the fitted straight line is obtained to determine the increasing trend of the pollutant concentration.
4. The multi-source data integration and governance method for pollution source monitoring as described in claim 1, characterized in that, The step of adjusting the smoothing coefficient according to the smoothing adjustment coefficient includes: Determine the smoothing coefficient adjustment range value, wherein the smoothing coefficient adjustment range value is the difference between the maximum preset smoothing coefficient and the minimum preset smoothing coefficient; The smoothing adjustment amount is obtained by multiplying the smoothing adjustment coefficient by the smoothing coefficient adjustment range value. The smoothing coefficient is obtained by calculating the sum of the smoothing coefficient adjustment amount and the minimum preset smoothing coefficient.
5. The multi-source data integration and governance method for pollution source monitoring as described in claim 1, characterized in that, After obtaining the predicted pollutant concentration, the multi-source data integration and governance method for pollution source monitoring further includes: Determine whether the predicted pollutant concentration exceeds the preset emission standard concentration. If it does, output a pollution control early warning signal.
6. The multi-source data integration and governance method for pollution source monitoring as described in claim 1, characterized in that, The process of obtaining the starting point of the upward trend includes: Determine the first-order difference sequence of the pollutant concentration data sequence; Determine candidate times for the pollutant concentration data sequence, wherein the candidate times are the times in the first-order difference sequence that are greater than a preset pollutant concentration change threshold; Obtain the number of candidate moments within a local time interval after each candidate moment; The target time is taken as the starting point of the upward trend. The target time is the first candidate time in the time sequence when the number of candidate times is greater than a preset threshold.
7. The multi-source data integration and governance method for pollution source monitoring as described in claim 1, characterized in that, The anomalous mutation points are obtained by processing the electrical data sequence using the Pettitt mutation point detection algorithm.
8. A multi-source data integration and management system for pollution source monitoring, characterized in that, include: Memory and processor; The memory is connected to the processor; The memory is used to store program instructions; The processor is configured to implement the multi-source data integration and governance method for pollution source monitoring as described in any one of claims 1-7 when program instructions are executed.