Intelligent monitoring method and system for operation state of pollution abatement equipment

Abstract

The invention relates to the field of data processing, and relates to an intelligent monitoring method and system for the operation state of pollution control equipment, and the method comprises the steps: collecting pressure data in the pollution control equipment at a plurality of sampling moments, and constructing a first data set; taking the target data as other data except the noise data in the first data set; correcting the pressure values of all the target data to obtain corrected normal data; based on all the corrected normal data, constructing a box plot on the second data set, and performing anomaly detection on the second data set by using the box plot; and when at least one abnormal data is detected in the second data set, judging that the pollution treatment equipment is abnormal. According to the method, the new data set is created by correcting the pressure value error caused by equipment deviation or environmental factors, and an accurate and effective data source is provided for subsequent pressure data anomaly detection by using the box plot, so that the anomaly detection accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing and to a method and system for intelligently monitoring the operating status of pollution control equipment. Background Art

[0002] Soil is the loose material layer on the earth's surface, mainly composed of minerals, organic matter, water and air. The interaction of these components forms the unique structure and function of soil.

[0003] To ensure the effectiveness of soil pollution control, we must first ensure the stable operation of the pollution control equipment. We can do this by performing abnormal detection on the pressure data of the internal pipelines of the pollution control equipment. If the detected pressure data does not match the normal pressure data, it may indicate that the equipment has an abnormality or failure.

[0004] The box plot algorithm is an anomaly detection algorithm that can be used to detect whether different categories of data have anomalies. For a data set, draw a box plot, such as Figure 1 As shown, Q1 is the lower quartile, Q2 is the median, Q3 is the upper quartile, IQR is the interquartile range, Q3+1.5IQR is the upper limit, Q1-1.5IQR is the lower limit, and the hollow circles represent outliers.

[0005] The drawn box plot provides us with a standard for identifying outliers: outliers are defined as values ​​less than Q1-1.5IQR or greater than Q3+1.5IQR, thereby identifying outliers in the data set. However, when constructing a box plot based on a data set, some data in the data set may have numerical deviations due to external interference during collection. When performing box plot anomaly detection, some data cannot meet the standard for identifying outliers, resulting in reduced accuracy of anomaly detection. Summary of the invention

[0006] In order to solve the technical problem that when constructing a box plot based on a data set, some data in the data set may have numerical deviations due to external interference during collection, and when performing box plot anomaly detection, some data cannot meet the standard for identifying outliers, resulting in reduced accuracy of anomaly detection, the present invention provides solutions in the following multiple aspects.

[0007] In a first aspect, a method for intelligently monitoring the operating status of pollution control equipment includes: Collect pressure data in the pollution control equipment at multiple sampling times to construct a first data set; divide the pressure data in the first data set into noise data and target data; the target data is other data in the first data set except the noise data; Correcting the pressure values ​​of all the target data to obtain corrected normal data; Based on all the corrected normal data, construct a second data set, construct a box plot on the second data set, and perform anomaly detection on the second data set using the box plot; When at least one abnormal data is detected in the second data set, it is determined that the pollution control equipment is abnormal; Among them, the pressure value of the corrected normal data is obtained by subtracting the deviation value from the pressure value of the target data, the deviation value is the product of the difference between the pressure value of the target data and the average pressure value in the target data segment and the correction coefficient, the target data segment is centered on the corresponding target data, and the corresponding data segment is intercepted in the first data set; the correction coefficient is used to adjust the pressure value corresponding to the target data.

[0008] Beneficial effects: First, all pressure data in the first data set are denoised to eliminate noise components that may affect the accuracy of the detection results. Then, all the denoised pressure data are further processed, and the correction coefficient is applied to adjust the pressure value of the pressure data to correct the pressure value deviation caused by equipment deviation or environmental factors. A new data set is created, which only contains corrected normal data, providing an accurate and effective data source for subsequent anomaly detection using box plots, thereby improving the accuracy of anomaly detection.

[0009] In one embodiment, the step of acquiring the target data is: The steps for obtaining the target data are as follows: Collect flow data of the eluent in the pollution control equipment at multiple sampling times to construct a flow data set; The noise coefficient of any pressure data in the first data set is calculated. When the noise coefficient of the pressure data is greater than a preset threshold, the pressure data is determined to be noise data, and then the other data in the first data set except the noise data is marked as target data.

[0010] Beneficial effects: By calculating the noise coefficient of pressure data and setting a threshold, noise data and valid data can be effectively distinguished, thereby improving the overall quality of the data set; marking other data except noise data as target data ensures the accuracy of subsequent processing and analysis, laying the foundation for the correct assessment of equipment status.

[0011] In one embodiment, the noise coefficient is calculated as follows: At the same sampling time, a first data segment is obtained with the corresponding pressure data as the center, and a second data segment of the same length as the first data segment is obtained with the corresponding flow data as the center; Calculating a first volatility of the pressure data in the first data segment, and calculating a second volatility between all the pressure data and the corresponding flow data in the first data segment; The product of the first volatility and the second volatility is taken as the noise coefficient.

[0012] Beneficial effects: Calculating the first volatility of pressure data and the second volatility between pressure and flow helps to comprehensively evaluate the variability and correlation of data. At the same time, taking the product of the first volatility and the second volatility as the noise coefficient can amplify the impact of noise, making the identification of noise data more sensitive and accurate.

[0013] In one embodiment, the second volatility is: ; In the formula, is the second volatility, The first data segment Percentile of pressure data, The second data segment The percentile of the flow data, Represents the mean of the percentile differences between all pressure data and corresponding flow data in the first data segment and the second data segment; the percentile is used to characterize the degree of abnormality of the data.

[0014] Beneficial effects: By calculating the percentiles of pressure data and flow data, the position of the data in the overall distribution can be accurately characterized, thereby identifying abnormal values ​​that deviate from the normal range; calculating the absolute value of the difference between the percentiles of pressure data and flow data helps capture inconsistencies and potential abnormal fluctuations between the two.

[0015] In one embodiment, the first volatility is The difference between the pressure value of the pressure data and the maximum pressure value in the first data segment and the minimum pressure value.

[0016] Beneficial Effects: By calculating the difference between the pressure value and the difference between the maximum and minimum pressure values ​​in the data segment, abnormal fluctuations in the pressure data can be effectively identified, thereby improving the overall quality of the first data set.

[0017] In one embodiment, the correction factor is: ; In the formula, represents the correction coefficient, Indicates The minimum sampling time interval between the target data and the surrounding noise data, Indicates The pressure value of the target data, Indicates The pressure value of the noise data with the smallest sampling time interval of the target data, Indicates the number of noise data in the target data segment.

[0018] In one embodiment, the step of performing anomaly detection on the second data set using a box plot is: Segmenting the time series of the pressure data in the second data set to obtain a plurality of sub-data segments; The pressure data in any sub-data segment are sorted according to the pressure value to obtain the upper quartile, lower quartile, median, interquartile range, upper limit and lower limit; Determine the abnormal range and treat data that is less than the lower limit or greater than the upper limit as abnormal data.

[0019] In a second aspect, a pollution control equipment operating status intelligent monitoring system includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned pollution control equipment operating status intelligent monitoring method is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 It is a schematic diagram of a box plot.

[0021] Figure 2 It is a method flow chart of steps S1-S4 in the method for intelligently monitoring the operating status of pollution control equipment in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0023] The embodiment of the present invention discloses a method for intelligently monitoring the operating status of pollution control equipment, referring to Figure 2 , including steps S1-S4, as follows S1: Collect pressure data in the pollution control equipment at multiple sampling times to construct a first data set; divide the pressure data in the first data set into noise data and target data; the target data is other data in the first data set except the noise data.

[0024] In the embodiment of the present invention, a pressure sensor is used to collect pressure data in the soil pollution control equipment, and a flow meter is used to measure the flow data of the eluent at the outlet of the above equipment. The specific collection time is thirty minutes, and the collection frequency is once per second. Then, an analog-to-digital conversion device is used to digitize the pressure data and the flow data to obtain digital representations of the pressure data and the flow data, thereby constructing the corresponding first data set and second data set.

[0025] When collecting pressure data, noise data appears in the first data set due to the influence of the external environment or the collection equipment. The existence of noise data will cause the pressure values ​​corresponding to the surrounding pressure data to be abnormal. Therefore, the pressure data in the first data set is divided into noise data and other data except the noise data. The other data except the noise data is marked as the target data.

[0026] The steps for obtaining the above target data are as follows: At any sampling time, the first data set corresponding to The pressure data is taken as the center, and 50 pressure data are selected on both sides as the first data segment. If one side of the pressure data is less than 50 pressure data, it can be supplemented on the other side. Similarly, the second data set can be obtained at the same sampling time. The second data segment of the traffic data.

[0027] According to Bernoulli's principle, when a fluid passes through a pipe or equipment, there is a balance between the kinetic energy and pressure energy of the fluid, and an increase in pressure will increase the kinetic energy of the fluid.

[0028] First, calculate the The sum of the difference between the pressure value of the first pressure data and the maximum pressure value in the first data segment and the The first volatility of the pressure data in the first data segment is obtained by calculating the difference between the pressure value of the pressure data and the minimum pressure value in the first data segment.

[0029] Then the expression of the first volatility is as follows:

[0030] In the formula, For the The first volatility of the pressure data, For the The pressure value of the pressure data, , They are the maximum pressure value and the minimum pressure value in the first data segment respectively.

[0031] The first volatility obtained by the above calculation reflects the volatility of the corresponding pressure data in its first data segment. If the value of the first volatility is larger, it means that the pressure data is closer to a certain extreme value and shows a larger volatility, then the pressure data will be more abnormal.

[0032] Secondly, calculate the first data segment The percentile of the pressure data and the corresponding The percentile of each flow data is used to obtain the second volatility between the pressure data and the corresponding flow data.

[0033] Then the expression of the second volatility is as follows:

[0034] In the formula, is the second volatility, The first data segment Percentile of pressure data, The second data segment The percentile of the flow data, Represents the mean of the percentile differences between all pressure data and corresponding flow data in the first data segment and the second data segment.

[0035] It should be noted that percentile is a very key concept in statistics, which is used to characterize the position of a certain percentage in a data set. In the implementation of the present invention, by comparing the percentiles of the data, outliers in the data set can be found. For example, in the first data set, the percentiles of If the value of the pressure data is far below the 25th percentile or above the 75th percentile, then The pressure data may be considered as outliers.

[0036] The product of the first volatility and the second volatility is taken as The noise coefficient of each pressure data is obtained by the same method, and the noise coefficient of all pressure data in the first data set is obtained, and a threshold is set, and the pressure data with a noise coefficient greater than or equal to the threshold is marked as noise data, and vice versa, it is marked as target data. In the implementation of the present invention, the threshold is set to 0.85, and in other embodiments, it can be set according to specific circumstances.

[0037] The above The expression of the noise coefficient of pressure data is as follows:

[0038] In the formula, For the The noise coefficient of the pressure data is is the normalization function, is the second volatility, The first volatility.

[0039] S2: Correcting the pressure values ​​of all the target data to obtain corrected normal data.

[0040] After all the target data are acquired in the previous step, the numerical performance of the surrounding data of the target data is analyzed. For a target data, the more noise data contained in its data segment, the closer the sampling time interval from the noise data, and the smaller the numerical difference with the noise data, the greater the impact of noise on the target data, and the more the pressure value of the target data needs to be corrected.

[0041] In an embodiment of the present invention, a pressure value correction coefficient of the target data is calculated according to the actual difference between the target data and the surrounding noise data, and the pressure value of the target data is adjusted by the correction coefficient to correct the pressure value error caused by equipment deviation or environmental factors.

[0042] Specifically, for all target data, select the first The target data segment corresponding to each target data is intercepted. The target data segment is obtained in the same way as the first data segment described above, and no further description is given here.

[0043] The first The pressure value correction coefficient of the target data is:

[0044] In the formula, Indicates The pressure value correction coefficient of the target data, Indicates The minimum sampling time interval between the target data and the surrounding noise data, Indicates The pressure value of the target data, Indicates The pressure value of the noise data with the smallest sampling time interval of the target data, Indicates the number of noise data in the target data segment.

[0045] No. The larger the pressure value correction coefficient of a target data is, the more the target data needs to be corrected.

[0046] It should be noted that, in other embodiments, the method for correcting the pressure value of the target data may also adopt: mechanical correction method: calibrating the pressure sensor using a standard calibration device; static pressure calibration method: comparing the sensor with a standard of known accuracy to determine the deviation of the sensor and perform correction, etc.

[0047] According to the above The pressure value correction coefficient of each target data can be obtained by the pressure value correction coefficient of all target data. The pressure value of the corresponding target data is corrected by the correction coefficient, and then all corrected normal data can be obtained.

[0048] Specifically, the Take target data as an example, obtain the average pressure value in the target data segment, and calculate the deviation value of the target data. The expression of the pressure value of the corrected normal data is as follows:

[0049] In the formula, is the pressure value of the normal data after correction, For the The pressure value of the target data, is the average pressure value in the target data segment, For the The pressure value correction coefficient of the target data. Indicates The deviation value of the target data.

[0050] The value of 0.5 is set to round off the pressure value of the pressure correction.

[0051] S3: constructing a second data set based on all the corrected normal data, constructing a box plot on the second data set, and performing anomaly detection on the second data set using the box plot.

[0052] A second data set is constructed using all corrected normal data. At this time, all data in the second data set are accurate and valid data. A box plot is constructed on the second data set, and anomaly detection is performed on the second data set using the box plot.

[0053] The steps for anomaly detection on the second data set using the box plot are as follows: The time series of the pressure data in the second data set is segmented and 200 data are used as a sub-data segment for anomaly detection; The pressure data in any sub-data segment are sorted according to the pressure value to obtain the upper quartile, lower quartile, median, interquartile range, upper limit and lower limit; Determine the abnormal range and treat data that is less than the lower limit or greater than the upper limit as abnormal data.

[0054] S4: When at least one abnormal data is detected in the second data set, it is determined that the pollution control equipment is abnormal.

[0055] The abnormal data detected in step S3 are summarized. When there is at least one abnormal data, the soil pollution control equipment is abnormal. An early warning can be issued later to enable relevant technical personnel to discover and deal with equipment failure problems in time, thereby ensuring the efficiency and safety of soil pollution control.

[0056] The present invention distinguishes noise data from other data, eliminates noise components that may affect the accuracy of detection results, applies correction coefficients to adjust pressure values ​​of other data to correct pressure value errors caused by equipment deviations or environmental factors, and creates a new data set that only contains corrected normal data, thereby providing an accurate and effective data source for subsequent anomaly detection using box plots, thereby improving the accuracy of anomaly detection.

[0057] The embodiment of the present invention also discloses an intelligent monitoring system for the operation status of pollution control equipment, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the intelligent monitoring method for the operation status of pollution control equipment according to the present invention is implemented.

[0058] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0059] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0060] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0061] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. An intelligent monitoring method for the operation status of pollution control equipment, characterized in that: include: Collect pressure data in the pollution control equipment at multiple sampling times to construct a first data set; dividing the pressure data in the first data set into noise data and target data; The target data is other data in the first data set except the noise data; Correcting the pressure values ​​of all the target data to obtain corrected normal data; Based on all the corrected normal data, construct a second data set, construct a box plot on the second data set, and perform anomaly detection on the second data set using the box plot; When at least one abnormal data is detected in the second data set, it is determined that the pollution control equipment is abnormal; Among them, the pressure value of the corrected normal data is obtained by subtracting the deviation value from the pressure value of the target data, the deviation value is the product of the difference between the pressure value of the target data and the average pressure value in the target data segment and the correction coefficient, the target data segment is centered on the corresponding target data, and the corresponding data segment is intercepted in the first data set; the correction coefficient is used to adjust the pressure value corresponding to the target data.

2. The method for intelligently monitoring the operating status of pollution control equipment according to claim 1, characterized in that: The steps for obtaining the target data are as follows: Collect flow data of the eluent in the pollution control equipment at multiple sampling times to construct a flow data set; The noise coefficient of any pressure data in the first data set is calculated. When the noise coefficient of the pressure data is greater than a preset threshold, the pressure data is determined to be noise data, and then the other data in the first data set except the noise data is marked as target data.

3. The method for intelligently monitoring the operating status of pollution control equipment according to claim 2, characterized in that: The calculation process of the noise coefficient is: At the same sampling time, a first data segment is obtained with the corresponding pressure data as the center, and a second data segment of the same length as the first data segment is obtained with the corresponding flow data as the center; Calculating a first volatility of the pressure data in the first data segment, and calculating a second volatility between all the pressure data and the corresponding flow data in the first data segment; The product of the first volatility and the second volatility is taken as the noise coefficient.

4. The method for intelligently monitoring the operating status of pollution control equipment according to claim 3 is characterized in that: The second volatility is: ; In the formula, is the second volatility, The first data segment Percentile of pressure data, The second data segment The percentile of the flow data, Represents the mean of the percentile differences between all pressure data and corresponding flow data in the first data segment and the second data segment; the percentile is used to characterize the degree of abnormality of the data.

5. The method for intelligently monitoring the operating status of pollution control equipment according to claim 4, characterized in that: The first volatility is The difference between the pressure value of the pressure data and the maximum pressure value in the first data segment and the minimum pressure value.

6. The method for intelligently monitoring the operating status of pollution control equipment according to claim 5, characterized in that: The correction factor is: ; In the formula, represents the correction coefficient, Indicates The minimum sampling time interval between the target data and the surrounding noise data, Indicates The pressure value of the target data, Indicates The pressure value of the noise data with the smallest sampling time interval of the target data, Indicates the number of noise data in the target data segment.

7. The method for intelligently monitoring the operating status of pollution control equipment according to claim 6, characterized in that: The step of using the box plot to perform anomaly detection on the second data set is: Segmenting the time series of the pressure data in the second data set to obtain a plurality of sub-data segments; The pressure data in any sub-data segment are sorted according to the pressure value to obtain the upper quartile, lower quartile, median, interquartile range, upper limit and lower limit; Determine the abnormal range and treat data that is less than the lower limit or greater than the upper limit as abnormal data.

8. Intelligent monitoring system for operation status of pollution control equipment, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for intelligently monitoring the operating status of pollution control equipment according to any one of claims 1 to 7 is implemented.

Images