A method and system for intelligent data processing of artificial intelligence

Abstract

The present invention relates to the technical field of data processing, and in particular to an artificial intelligence data intelligent processing method and system, the method comprising: sampling based on a smoke exhaust port to obtain PM2.5 content data and wind speed data; confirming a missing part of the PM2.5 content data, generating a missing data segment, and calculating the degree of irrelevance of the PM2.5 content data change amount and the wind speed data change amount in combination with the wind speed data; obtaining a variation coefficient affected by the wind speed data according to the PM2.5 content data and the wind speed data, restoring the PM2.5 content data at a previous sampling moment of the missing data segment through the variation coefficient, and restoring the data in combination with the degree of irrelevance of the PM2.5 content data change amount and the wind speed data change amount; performing abnormal monitoring according to the restored PM2.5 content data; finding and restoring the missing data segment through the degree of irrelevance between the PM2.5 content data change and the wind speed change at the same moment, so as to accurately monitor the abnormality.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an artificial intelligence data intelligent processing method and system. Background Art

[0002] When monitoring the PM2.5 content in the exhaust gas emitted from the smoke outlet of a factory chimney, traditional monitoring methods often use a particle sampler to collect particulate matter in the exhaust gas. The particle sampler is usually placed in a specific position in order to accurately capture the tiny particles in the emitted exhaust gas; however, traditional detection methods are not only inefficient, but also affected by many external factors, such as ambient temperature, humidity, wind speed, etc., which may interfere with the sampling results and affect the accuracy of the data.

[0003] In the prior art, PM2.5 sensors can be used to replace traditional particle samplers. They can monitor the concentration of particulate matter in real time. By continuously collecting data, PM2.5 sensors can provide more accurate and timely monitoring results, and perform abnormal analysis on these real-time data. Abnormal fluctuations in the data can be discovered in time, and whether the data exceeds the standard can be determined by its changes over time. However, the wind speed around the smoke exhaust port is not controlled by humans, and it will change with changes in weather and environment. Changes in wind speed will affect the monitoring process of PM2.5 content data to a certain extent. For example, when the wind speed is high, PM2.5 particles in the exhaust gas may be blown away quickly, which makes the PM2.5 content data that exceeds the standard at certain times not be monitored due to the influence of wind speed. The abnormal analysis of PM2.5 content data in this scenario will produce large errors, affecting subsequent processing. Summary of the invention

[0004] In order to solve the technical problem that the existing method cannot combine the measurement of PM2.5 content data with wind speed analysis, resulting in large errors, the purpose of the present invention is to provide an artificial intelligence data intelligent processing method, and the technical solution adopted is as follows:

[0005] Sampling is performed based on the smoke exhaust port to obtain PM2.5 content data and wind speed data;

[0006] Analyze the PM2.5 content data, identify the missing parts of the PM2.5 content data, generate missing data segments, and combine with the wind speed data to calculate the degree of irrelevance between the change in PM2.5 content data and the change in wind speed data;

[0007] The variation coefficient of PM2.5 content data affected by wind speed data is obtained according to PM2.5 content data and wind speed data, and the PM2.5 content data of the previous sampling time of the missing data segment is restored by the variation coefficient. The PM2.5 content data of the missing data segment is restored by combining the irrelevance between the variation of PM2.5 content data and the variation of wind speed data;

[0008] Abnormal monitoring is carried out based on the restored PM2.5 content data.

[0009] Preferably, analyzing the PM2.5 content data, confirming the missing parts of the PM2.5 content data, generating missing data segments, and calculating the degree of irrelevance between the PM2.5 content data change and the wind speed data change in combination with the wind speed data, including:

[0010] Calculate the abnormality of any data point based on the PM2.5 content data, set a screening threshold, and define the data point with an abnormality greater than the screening threshold as a missing data point. Use the missing data points to confirm the missing part of the PM2.5 content data and generate a missing data segment.

[0011] Based on any missing data segment, the degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling moment and the degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment are calculated respectively.

[0012] Preferably, the abnormality degree of any data point is calculated based on the PM2.5 content data, and the corresponding calculation formula is:

[0013]

[0014] in, Indicates the PM2.5 content data The degree of abnormality of each data point; Indicates the mean value of PM2.5 content data; Indicates the PM2.5 content data The data value of the data point; Indicates The range of neighboring data points of a data point; Indicates Neighborhood data points of data points The data value of represents an exponential function with a natural constant as base; represents the linear normalization function.

[0015] Preferably, data points with an abnormality greater than the screening threshold are missing data points, and missing data points are used to confirm missing parts of PM2.5 content data and generate missing data segments, including:

[0016] Calculate the abnormality of all PM2.5 content data. When the abnormality of any data point in the PM2.5 content data is greater than the screening threshold, the current data point is confirmed as a missing data point.

[0017] Set judgment conditions on the time series, integrate all missing data points that continuously meet the judgment conditions, and generate missing data segments.

[0018] Preferably, the degree of uncorrelation between the PM2.5 content data change and the wind speed data change at the previous sampling moment is calculated based on any missing data segment, and the corresponding calculation formula is:

[0019]

[0020] in, Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling time of the missing data segment; Indicates The change in PM2.5 content data actually sampled at the previous sampling time of the missing data segment; Indicates The change in wind speed data actually sampled at the previous sampling moment of the missing data segment; Indicates the range of missing data segments; and is calculated in the same way according to the calculation formula , Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment after the missing data segment.

[0021] Preferably, the coefficient of variation of the PM2.5 content data affected by the wind speed data is obtained based on the PM2.5 content data and the wind speed data, and the corresponding calculation formula is:

[0022]

[0023] in, Indicates the coefficient of variation of PM2.5 content data affected by wind speed data; and Respectively represent Any two PM2.5 content data at adjacent sampling times before the previous sampling time of the missing data segment; and Respectively represent Any two wind speed data at adjacent sampling times before the previous sampling time of the missing data segment.

[0024] Preferably, the PM2.5 content data at the previous sampling time of the missing data segment is restored by the variation coefficient, and the corresponding calculation formula is:

[0025]

[0026] in, Indicates The PM2.5 content data restored from the previous sampling time of the missing data segment; Indicates The PM2.5 content data actually sampled at the previous sampling time for the missing data segment; Indicates the coefficient of variation of PM2.5 content data affected by wind speed data; Indicates The wind speed data actually sampled at the previous sampling moment of the missing data segment.

[0027] Preferably, the PM2.5 content data of the missing data segment is restored in combination with the irrelevance between the PM2.5 content data change and the wind speed data change, and the corresponding calculation formula is:

[0028]

[0029] in, Expressing the PM2.5 content data after restoration of missing data segments; Indicates The PM2.5 content data restored from the previous sampling time of the missing data segment; Indicates the change in PM2.5 content data at two sampling moments when the degree of irrelevance is not 0; Indicates the change in wind speed data at two sampling moments corresponding to the time when the degree of irrelevance is not 0; Indicates the coefficient of variation of PM2.5 content data affected by wind speed data; Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling time of the missing data segment; Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment after the missing data segment.

[0030] Preferably, abnormality monitoring is performed based on the restored PM2.5 content data, including: calculating the local reachable density based on the restored PM2.5 content data, using the LOF algorithm to obtain the LOF value of the restored PM2.5 content data, comparing the local reachable density with the LOF value, and realizing abnormality monitoring of the restored PM2.5 content data.

[0031] In order to solve the above problems, the present application also provides an artificial intelligence data intelligent processing system, wherein the system stores program data, and when the program data is executed, an artificial intelligence data intelligent processing method as described in any of the above items is implemented.

[0032] The present invention has the following beneficial effects:

[0033] 1. Comprehensive consideration is made based on the smoke exhaust outlet wind speed data and PM2.5 content data to take into account the impact of changes in wind speed in actual scenarios on PM2.5 content data detection. The missing PM2.5 content data is found through the degree of irrelevance between the PM2.5 content data change and the wind speed change at the same moment, and then the abnormal PM2.5 content data segment is restored, so that subsequent abnormal monitoring of PM2.5 content data is more accurate and more in line with actual scenarios, making this application more reliable and robust.

[0034] 2. The artificial intelligence data intelligent processing system provided by the present invention has the same beneficial effects as the artificial intelligence data intelligent processing method provided by the present invention, which will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 A flowchart of the steps of an artificial intelligence data intelligent processing method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the data intelligent processing method and system of artificial intelligence proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0038] Unless defined otherwise, 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 belongs.

[0039] It should be noted that in order to ensure that the calculation results are meaningful, when performing fractional operations in the embodiments of the present invention, when the denominator is 0, it is necessary to add a parameter adjustment factor greater than 0 to the denominator to prevent the denominator from being 0. The value of the parameter adjustment factor is set by the implementer according to actual conditions, and this application does not impose any special restrictions.

[0040] The specific scheme of an artificial intelligence data intelligent processing method and system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0041] The wind speed around the smoke exhaust port is not controlled by humans, resulting in large fluctuations in the PM2.5 content data monitored using the existing technology, and thus large data errors, which affects the subsequent processing of the data. In one embodiment of the present invention, an artificial intelligence data intelligent processing method is provided, which is based on a comprehensive consideration of the smoke exhaust port wind speed data and the PM2.5 content data, and the missing PM2.5 content data is found through the degree of irrelevance between the PM2.5 content data change and the wind speed change at the same moment, and then the abnormal PM2.5 content data segment is restored, so that the subsequent abnormal monitoring of the PM2.5 content data is more accurate. In order to realize an artificial intelligence data intelligent processing method, an artificial intelligence data intelligent processing system is provided, which is essentially a software system, which is composed of modules that realize corresponding functions. The specific steps in the method are now introduced in detail.

[0042] See also Figure 1 , which shows a flowchart of the steps of an artificial intelligence data intelligent processing method provided by an embodiment of the present invention, the method comprising:

[0043] Step S1: Sampling based on the smoke exhaust port to obtain PM2.5 content data and wind speed data;

[0044] Step S2: Analyze the PM2.5 content data, identify the missing parts of the PM2.5 content data, generate missing data segments, and calculate the degree of irrelevance between the PM2.5 content data change and the wind speed data change in combination with the wind speed data;

[0045] Step S3: according to the PM2.5 content data and the wind speed data, a variation coefficient of the PM2.5 content data affected only by the wind speed data is obtained, and the PM2.5 content data of the previous sampling time of the missing data segment is restored by the variation coefficient, and the PM2.5 content data of the missing data segment is restored in combination with the irrelevance degree of the PM2.5 content data variation and the wind speed data variation;

[0046] Step S4: Perform abnormal monitoring based on the restored PM2.5 content data.

[0047] To better explain, PM2.5 refers to fine particulate matter, that is, particles with a diameter of less than or equal to 2.5 microns, which can remain suspended in the air for a long time and have a profound impact on the environment and human health; the exhaust gas emitted by factory chimneys includes PM2.5, which not only reduces the air quality and increases the probability of smog weather, but also causes damage to the respiratory system of surrounding residents and increases the risk of respiratory diseases; and PM2.5 particles may also carry various harmful substances, such as heavy metals and polycyclic aromatic hydrocarbons, which have potential carcinogenic risks to human health.

[0048] In real life, when detecting the PM2.5 content in exhaust gas emitted from factory chimneys, fluctuations in the wind speed at the exhaust outlet may cause abnormal fluctuations in the PM2.5 content data, resulting in data discontinuity and even data loss, which brings great troubles and challenges to the subsequent abnormal analysis of PM2.5 content data. The accuracy and completeness of the PM2.5 content data cannot be guaranteed, that is, the basis for data analysis cannot be provided, making it impossible for regulators to obtain accurate information to take corresponding emission reduction measures, and thus unable to effectively control and reduce the impact of PM2.5 on the environment and human health.

[0049] As an optional implementation, the present application is not limited to factory chimneys, but also includes any scenario where the monitoring of PM2.5 content is affected by wind speed, such as urban environments or relatively remote areas.

[0050] It can be explained that step S1: sampling is carried out based on the smoke exhaust port to obtain PM2.5 content data and wind speed data; specifically, a PM2.5 sensor is arranged based on the smoke exhaust port of the factory chimney to monitor the concentration of PM2.5 particles in real time, and an anemometer is used to monitor the wind speed data in real time during the same time period; wherein, the PM2.5 sensor is a device for detecting the concentration of fine particles in the air, and can accurately measure particles with a diameter less than or equal to 2.5 microns; an anemometer is an instrument for measuring wind speed, which can accurately detect and record the speed of wind movement.

[0051] It is understandable that in the process of collecting pollutant data such as PM2.5 content data around factory chimney exhaust ports, the collected monitoring data will often be abnormal due to the influence of environmental wind. In order to ensure accurate monitoring of PM2.5 content data, these abnormal data need to be corrected and restored. In order to complete this process, it is first necessary to determine the specific range of these abnormal data.

[0052] Further, step S2 includes:

[0053] Step S21: Calculate the abnormality of any data point based on the PM2.5 content data, set a screening threshold, and define the data point with an abnormality greater than the screening threshold as a missing data point. Use the missing data points to confirm the missing part of the PM2.5 content data and generate a missing data segment.

[0054] It is clarified that calculating the degree of abnormality of any data point based on PM2.5 content data refers to an in-depth analysis of the PM2.5 content data in order to accurately locate and identify abnormal data in the PM2.5 content data set, which is conducive to the subsequent confirmation of missing data segments.

[0055] Furthermore, in step S21, the abnormality degree of any data point is calculated based on the PM2.5 content data, and the corresponding calculation formula is:

[0056]

[0057] in, Indicates the PM2.5 content data The degree of abnormality of each data point; Indicates the mean value of PM2.5 content data; Indicates the PM2.5 content data The data value of the data point; Indicates The range of neighboring data points of a data point; Indicates Neighborhood data points of data points The data value of represents an exponential function with a natural constant as base; represents the linear normalization function.

[0058] To explain, Indicates the current data point in the PM2.5 content data, i.e. The difference between the data point and the overall mean of the PM2.5 content data. The greater the difference between the two, the greater the impact of the wind speed data on the current PM2.5 content data, so the PM2.5 concentration shown by the PM2.5 content data is reduced; if the difference between the two is a negative value, it means that the currently monitored PM2.5 content data is abnormal and the PM2.5 concentration is high. In this case, use Indicates the abnormality of PM2.5 content data; Indicates the current data point in the PM2.5 content data, i.e. The relationship between the data points and the neighboring data points. When the smoke outlet is affected by the wind speed data, then within a period of time, that is, The PM2.5 content data monitored in the area are all abnormal. The difference between the PM2.5 content data and the neighboring data points is calculated based on the PM2.5 content data. The smaller the difference, the more abnormal the data point of the PM2.5 content data currently analyzed is. In this embodiment, when When the difference is less than 2, it is considered is the data point in the PM2.5 content data currently being analyzed, i.e. The neighborhood data points of a data point.

[0059] Further, in step S21, the data points whose abnormality is greater than the screening threshold are missing data points, and the missing data points are used to confirm the missing parts of the PM2.5 content data, and generate missing data segments, including:

[0060] Step S211: Calculate the abnormality level of all PM2.5 content data. When the abnormality level of any data point in the PM2.5 content data is greater than the screening threshold, confirm the current data point as a missing data point.

[0061] As an optional implementation, in this embodiment, the screening threshold .

[0062] To explain, , then the data point in the PM2.5 content data currently being analyzed is considered to be a missing data point; that is, data points are missing data points; otherwise, , it indicates that the data points in the current PM2.5 content data are normal and are not affected by the wind speed data.

[0063] Step S212: setting a judgment condition in time series, integrating all missing data points that continuously meet the judgment condition, and generating a missing data segment.

[0064] Optionally, in this embodiment, the judgment condition is equal to 5.

[0065] Specifically, when the missing data points of the determined PM2.5 content data exceed 5 consecutive data points in time series, this series of consecutive missing data points is defined as a missing data segment; for example, assuming that 7 consecutive data points are currently determined to be missing data points, these 7 data points are connected and confirmed as a complete missing data segment; thereby effectively identifying all missing data segments in the PM2.5 content data and ensuring the integrity of the PM2.5 content data processing.

[0066] To restore the PM2.5 content data of the missing data segment, it is also necessary to calculate the change in the smoke exhaust volume of the smoke exhaust outlet at the time corresponding to the missing data segment, so the degree of irrelevance between the PM2.5 content data and the wind speed data at the sampling moment before and after the missing data segment is analyzed; wherein, the degree of irrelevance refers to the fact that at a specific time point, there is no significant correlation between the wind speed data and the PM2.5 content data, indicating that the change in wind speed data has no direct impact on the change in PM2.5 content data, or the impact is weak or even undetectable; it can be explained that the degree of irrelevance indicates that the smoke exhaust volume of the smoke exhaust outlet at the time corresponding to the missing data segment is generated by itself, that is, the change in the smoke exhaust volume of the smoke exhaust outlet at this moment has a direct impact on the PM2.5 content data.

[0067] Step S22: Based on any missing data segment, calculate the degree of irrelevance between the PM2.5 content data change and the wind speed data change at the previous sampling moment, and the degree of irrelevance between the PM2.5 content data change and the wind speed data change at the next sampling moment.

[0068] To explain, taking the previous sampling moment of the missing data segment as an example, several sampling moments before this moment in which no PM2.5 content data is missing are selected, and the ratio of the absolute value of the change in PM2.5 content data to the absolute value of the change in wind speed data between each adjacent sampling moment is calculated, and the mean of these ratios is obtained; then the ratio of the absolute value of the change in PM2.5 content data to the absolute value of the change in wind speed data at the previous sampling moment of the missing data segment is subtracted from the previously calculated mean to indicate the degree of irrelevance of the previous sampling moment of the missing data segment; and then the specific impact of the change in smoke exhaust volume at the chimney exhaust outlet at a specific moment on the PM2.5 content data is quantitatively evaluated, thereby providing a scientific basis for the restoration of PM2.5 content data.

[0069] Further, in step S22, the uncorrelated degree of the PM2.5 content data change and the wind speed data change at the previous sampling moment is calculated based on any missing data segment, and the corresponding calculation formula is:

[0070]

[0071] in, Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling time of the missing data segment; Indicates The change in PM2.5 content data actually sampled at the previous sampling time of the missing data segment; Indicates The change in wind speed data actually sampled at the previous sampling moment of the missing data segment; Indicates the range of missing data segments; and is calculated in the same way according to the calculation formula , Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment after the missing data segment.

[0072] As an optional implementation, Indicates the range of missing data segments. In this embodiment , indicating the There are a total of The adjacent sampling moments participate in the calculation of the mean of the ratio of the absolute value of the PM2.5 content data change to the wind speed data change. It represents the mean.

[0073] Explanation, The ratio of the PM2.5 content data change to the wind speed data change at the previous sampling time of the missing data segment to the previous The larger the difference between the average ratio of PM2.5 content data change to wind speed data change at adjacent sampling times, the greater the difference in the average ratio of PM2.5 content data change to wind speed data change. The greater the degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the sampling moment before the missing data segment, it means that at this sampling moment, the change in PM2.5 content data is caused by the smoke exhaust volume of the smoke exhaust port itself, and has no significant correlation with the change in wind speed data; conversely, the smaller the degree of irrelevance, the stronger the correlation between the change in PM2.5 content data and the change in wind speed data at this sampling moment.

[0074] It is understandable that the missing PM2.5 content data in the time series does not occur instantaneously, that is, the change in wind speed data will not increase suddenly without any warning. Therefore, at the sampling time before the missing data segment, the PM2.5 content data should show a gradual downward trend for a period of time. In order to restore these missing data segments, it is necessary to deeply analyze the changing relationship between PM2.5 content data and wind speed data. Based on this analysis trend, the PM2.5 content data at the sampling time before the missing data segment is first reconstructed and inferred.

[0075] It can be explained that, in order to achieve this process, it is first necessary to quantify the changes in PM2.5 content data to explore the relationship between the changes in PM2.5 content data and the changes in wind speed data, because the changes in PM2.5 content data are not only affected by the smoke exhaust volume of the chimney exhaust port itself, but also by the changes in wind speed data; in this embodiment, based on the Before the last sampling time of the missing data segment The PM2.5 content data and wind speed data corresponding to two sampling moments with an irrelevant degree of 0 are selected from two adjacent sampling moments to analyze the influence of wind speed data on the change of PM2.5 content data; then, the PM2.5 content data actually sampled is added to the missing data segment to restore the PM2.5 content data of the previous sampling moment of the missing data segment, so as to improve the integrity of the PM2.5 content data.

[0076] Further, in step S3, the variation coefficient of the PM2.5 content data affected only by the wind speed data is obtained according to the PM2.5 content data and the wind speed data, and the corresponding calculation formula is:

[0077]

[0078] in, The coefficient of variation indicating that PM2.5 content data is only affected by wind speed data; and Respectively represent Any two PM2.5 content data at adjacent sampling times before the previous sampling time of the missing data segment; and Respectively represent Any two wind speed data at adjacent sampling times before the previous sampling time of the missing data segment.

[0079] To explain, The coefficient of variation that indicates that the PM2.5 content data is only affected by the wind speed data, that is, how the PM2.5 content data changes under the change of specific wind speed data; and , and Respectively represent Before the last sampling time of the missing data segment The PM2.5 content data and wind speed data corresponding to two adjacent sampling moments with an irrelevant degree of 0; and the irrelevant degree is 0, which means that at this sampling moment, the PM2.5 content data is only affected by the wind speed data.

[0080] Furthermore, in step S3, the PM2.5 content data at the previous sampling time of the missing data segment is restored by the variation coefficient, and the corresponding calculation formula is:

[0081]

[0082] in, Indicates The PM2.5 content data restored from the previous sampling time of the missing data segment; Indicates The PM2.5 content data actually sampled at the previous sampling time for the missing data segment; The coefficient of variation indicating that PM2.5 content data is only affected by wind speed data; Indicates The wind speed data actually sampled at the previous sampling moment of the missing data segment.

[0083] To explain, Indicates The PM2.5 content data at the previous sampling time of the missing data segment is reduced due to the wind speed data; The larger the The greater the PM2.5 content data restored at the previous sampling time of the missing data segment, the greater the PM2.5 content data actually collected at the first sampling time. The larger the wind speed data at the previous sampling moment of the missing data segment, the greater the The greater the PM2.5 content data restored from the previous sampling moment of the missing data segment.

[0084] It can be understood that the above-obtained degree of irrelevance can reflect the influence of the chimney exhaust outlet's own exhaust volume at the corresponding sampling time. First, the influence of the chimney exhaust outlet's own exhaust volume on the PM2.5 content data change is calculated, and the influence of the wind speed data obtained above on the PM2.5 content data change is calculated. The influence coefficient of the chimney exhaust outlet's own exhaust volume on the PM2.5 content data change is, that is, The PM2.5 content data of the missing data segment at the previous moment after the restoration is combined to restore the PM2.5 content data of the missing data segment.

[0085] Specifically, based on the aforementioned selection Before the last sampling time of the missing data segment The total amount of change in PM2.5 content data corresponding to two sampling moments with an irrelevant degree of 0 among the adjacent sampling moments is subtracted from the amount of change in PM2.5 content data affected only by the change in wind speed data, so as to obtain the partial amount of PM2.5 content data affected by the change in the smoke exhaust volume of the chimney exhaust port itself at these two moments; then, based on the ratio of this partial amount and the difference between the sampling moments with an irrelevant degree of 0, the coefficient of change of the smoke exhaust volume of the chimney exhaust port itself and the change in PM2.5 content data can be obtained, and the PM2.5 content data caused by the change in the smoke exhaust volume of the chimney exhaust port itself at the sampling moment corresponding to the missing data segment is calculated in combination with the irrelevant degree of the two sampling moments before and after the missing data segment, and finally, the PM2.5 content data restored at the previous sampling moment of the missing data segment is added based on this data to obtain the missing data of the restored overall PM2.5 content data.

[0086] Furthermore, in step S3, the PM2.5 content data of the missing data segment is restored by combining the irrelevance between the PM2.5 content data change and the wind speed data change, and the corresponding calculation formula is:

[0087]

[0088] in, Expressing the PM2.5 content data after restoration of missing data segments; Indicates The PM2.5 content data restored from the previous sampling time of the missing data segment; Indicates the change in PM2.5 content data at two sampling moments when the degree of irrelevance is not 0; Indicates the change in wind speed data at two sampling moments corresponding to the time when the degree of irrelevance is not 0; The coefficient of variation indicating that PM2.5 content data is only affected by wind speed data; Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling time of the missing data segment; Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment after the missing data segment.

[0089] To explain, It indicates the amount of change in PM2.5 content data at two sampling moments with a non-zero degree of irrelevance due to the change in the smoke exhaust volume of the chimney exhaust port itself; It represents the coefficient of variation between the amount of PM2.5 content data at two sampling moments with non-zero uncorrelation due to the change in the smoke exhaust volume of the chimney exhaust port itself and the amount of uncorrelation; It indicates the degree of irrelevance between the potential PM2.5 content data change and the wind speed data change in the missing data segment, that is, the change in the chimney exhaust outlet itself is The influence of the two sampling moments before and after the missing data segment on the change of PM2.5 content data.

[0090] It can be explained that by analyzing and modeling the changes in PM2.5 content data and wind speed data at the two sampling moments before and after any missing data segment, the missing data segment can be restored, thereby restoring all missing data segments and obtaining the entire restored PM2.5 content data, so that subsequent analysis operations on abnormal data in the PM2.5 content data are more accurate, and the present application has higher reliability and robustness.

[0091] Further, step S4 includes:

[0092] The local reachable density is calculated based on the restored PM2.5 content data, and the LOF algorithm is used to obtain the LOF value of the restored PM2.5 content data. The local reachable density is compared with the LOF value to realize abnormal monitoring of the restored PM2.5 content data.

[0093] To better explain, local reachable density refers to the density of data points in a specific area in multidimensional space, which reflects the density of data in the area; LOF (Local Outlier Factor) algorithm, that is, local outlier factor algorithm, is an unsupervised learning method for anomaly detection, which identifies outliers by comparing the density difference between data points and points in their neighborhood. Specifically, the LOF algorithm calculates the local density of each data point and compares it with the local density of neighboring points to determine whether the point is significantly lower than the density of its neighborhood. If the local density of a point is much lower than the density of its neighborhood, then this point may be an outlier.

[0094] Specifically, an anomaly analysis is performed on the restored PM2.5 content data. First, the PM2.5 content data in the entire monitoring period are clustered using the LOF algorithm, and the local reachable density is calculated to measure the sample density around the data points in the PM2.5 content data. For each data point in the PM2.5 content data, the average distance to any of its nearest neighbors is calculated, and its reciprocal is used as the estimate of the local reachable density, which represents the sample density around any data point in the PM2.5 content data, that is, it reflects the sample distribution. Then, the local anomaly factor is calculated by comparing the local reachable density of each data point with the average local reachable density of all data points in its neighborhood. If there is a local anomaly factor in the neighborhood, The higher the local reachable density, the smaller the LOF value of the data point, and vice versa, so as to identify the abnormal values ​​that deviate significantly from the data points in other PM2.5 content data in the local area; finally, the LOF values ​​corresponding to all PM2.5 content data are obtained, and the obtained LOF values ​​are sorted to the data points in the PM2.5 content data, and the larger the value, the more abnormal it is; that is, the data points of the PM2.5 content data corresponding to these values ​​are abnormal, and then the abnormal situations that may occur during the monitoring period are effectively identified and accurately judged, which is conducive to the subsequent processing of PM2.5 content data, which not only helps to improve the efficiency of PM2.5 content data processing, but also ensures the accuracy and reliability of PM2.5 content data.

[0095] It can be understood that a comprehensive consideration is made based on the smoke exhaust outlet wind speed data and PM2.5 content data to take into account the impact of changes in wind speed in actual scenarios on PM2.5 content data detection. The missing PM2.5 content data is found through the degree of irrelevance between the PM2.5 content data change and the wind speed change at the same moment, and then the abnormal PM2.5 content data segment is restored, so that subsequent abnormal monitoring of PM2.5 content data is more accurate and more in line with actual scenarios, making the present application more reliable and robust.

[0096] An embodiment of the present invention proposes an artificial intelligence data intelligent processing system, which stores program data. When the program data is executed, an artificial intelligence data intelligent processing method as described in the aforementioned embodiment is implemented; the system has the same beneficial effects as the artificial intelligence data intelligent processing method provided above, which will not be repeated here.

[0097] It can be understood that when a module of an artificial intelligence data intelligent processing system is in operation, it is necessary to utilize an artificial intelligence data intelligent processing method provided by the aforementioned embodiment. Therefore, whether the method is integrated with program data or different hardware is configured to produce functions similar to the effects achieved by the present invention, it falls within the scope of protection of the present invention.

[0098] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0099] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. An artificial intelligence data intelligent processing method, characterized in that: The method comprises: Sampling is performed based on the smoke exhaust port to obtain PM2.5 content data and wind speed data; Analyze the PM2.5 content data, identify the missing parts of the PM2.5 content data, generate missing data segments, and combine with the wind speed data to calculate the degree of irrelevance between the change in PM2.5 content data and the change in wind speed data; The variation coefficient of PM2.5 content data affected by wind speed data is obtained according to PM2.5 content data and wind speed data, and the PM2.5 content data of the previous sampling time of the missing data segment is restored by the variation coefficient. The PM2.5 content data of the missing data segment is restored by combining the irrelevance between the variation of PM2.5 content data and the variation of wind speed data; Conduct abnormal monitoring based on the restored PM2.5 content data; The method for generating a missing data segment includes: calculating the abnormality of any data point based on the PM2.5 content data, setting a screening threshold, and defining data points with an abnormality greater than the screening threshold as missing data points, and confirming the missing part of the PM2.5 content data through the missing data points to generate a missing data segment; The method for calculating the degree of irrelevance includes: calculating the degree of irrelevance of the change amount of PM2.5 content data and the change amount of wind speed data at the previous sampling moment and the degree of irrelevance of the change amount of PM2.5 content data and the change amount of wind speed data at the next sampling moment respectively based on any missing data segment; Based on any missing data segment, the irrelevance of the PM2.5 content data change and wind speed data change at the previous sampling time is calculated respectively. The corresponding calculation formula is: in, Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling time of the missing data segment; Indicates The change in PM2.5 content data actually sampled at the previous sampling time of the missing data segment; Indicates The change in wind speed data actually sampled at the previous sampling moment of the missing data segment; Indicates the range of missing data segments; and is calculated in the same way according to the calculation formula , Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment after the missing data segment; According to the PM2.5 content data and wind speed data, the variation coefficient of PM2.5 content data affected by wind speed data is obtained, and the corresponding calculation formula is: in, Indicates the coefficient of variation of PM2.5 content data affected by wind speed data; and Respectively represent Any two PM2.5 content data at adjacent sampling times before the previous sampling time of the missing data segment; and Respectively represent Any two wind speed data at adjacent sampling times before the previous sampling time of the missing data segment; The PM2.5 content data of the previous sampling time of the missing data segment is restored by the variation coefficient. The corresponding calculation formula is: in, Indicates The PM2.5 content data restored from the previous sampling time of the missing data segment; Indicates The PM2.5 content data actually sampled at the previous sampling time for the missing data segment; Indicates the coefficient of variation of PM2.5 content data affected by wind speed data; Indicates The wind speed data actually sampled at the previous sampling moment of the missing data segment.

2. The method for intelligent data processing of artificial intelligence according to claim 1, characterized in that: The abnormal degree of any data point is calculated based on the PM2.5 content data. The corresponding calculation formula is: in, Indicates the PM2.5 content data The degree of abnormality of each data point; Indicates the mean value of PM2.5 content data; Indicates the PM2.5 content data The data value of the data point; Indicates The range of neighboring data points of a data point; Indicates Neighborhood data points of data points The data value of represents an exponential function with a natural constant as base; represents the linear normalization function.

3. The method for intelligent data processing of artificial intelligence according to claim 2, characterized in that: Data points with an abnormality greater than the screening threshold are considered missing data points. The missing data points are used to confirm the missing parts of the PM2.5 content data and generate missing data segments, including: Calculate the abnormality of all PM2.5 content data. When the abnormality of any data point in the PM2.5 content data is greater than the screening threshold, the current data point is confirmed as a missing data point. Set judgment conditions on the time series, integrate all missing data points that continuously meet the judgment conditions, and generate missing data segments.

4. The method for intelligent data processing of artificial intelligence according to claim 1, characterized in that: The PM2.5 content data of the missing data segment is restored by combining the irrelevance between the PM2.5 content data change and the wind speed data change. The corresponding calculation formula is: in, Indicates PM2.5 content data after restoration of missing data segments; Indicates The PM2.5 content data restored from the previous sampling time of the missing data segment; Indicates the change in PM2.5 content data at two sampling moments when the degree of irrelevance is not 0; Indicates the change in wind speed data at two sampling moments corresponding to the time when the degree of irrelevance is not 0; Indicates the coefficient of variation of PM2.5 content data affected by wind speed data; Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the previous sampling time of the missing data segment; Indicates The degree of irrelevance between the change in PM2.5 content data and the change in wind speed data at the next sampling moment after the missing data segment is It indicates the change in the degree of uncorrelation between two sampling moments when the degree of uncorrelation is not 0.

5. The method for intelligent data processing of artificial intelligence according to claim 1, characterized in that: Abnormal monitoring is performed based on the restored PM2.5 content data, including: calculating the local reachable density based on the restored PM2.5 content data, using the LOF algorithm to obtain the LOF value of the restored PM2.5 content data, comparing the local reachable density with the LOF value, and realizing abnormal monitoring of the restored PM2.5 content data.

6. An artificial intelligence data intelligent processing system, characterized in that: The system stores program data, and when the program data is executed, an artificial intelligence data intelligent processing method according to any one of claims 1 to 5 is implemented.

Images