CN116955938B - Dry-type waste gas treatment equipment monitoring method and system based on data analysis - Google Patents
Dry-type waste gas treatment equipment monitoring method and system based on data analysis Download PDFInfo
- Publication number
- CN116955938B CN116955938B CN202311212448.5A CN202311212448A CN116955938B CN 116955938 B CN116955938 B CN 116955938B CN 202311212448 A CN202311212448 A CN 202311212448A CN 116955938 B CN116955938 B CN 116955938B
- Authority
- CN
- China
- Prior art keywords
- extremum
- value
- data
- window
- obtaining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012544 monitoring process Methods 0.000 title claims abstract description 41
- 238000007405 data analysis Methods 0.000 title claims abstract description 30
- 239000002912 waste gas Substances 0.000 title claims abstract description 29
- 238000012937 correction Methods 0.000 claims abstract description 67
- 238000012512 characterization method Methods 0.000 claims abstract description 43
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 36
- 239000007789 gas Substances 0.000 claims description 19
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 description 10
- 230000008569 process Effects 0.000 description 10
- 230000006870 function Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 3
- 238000001179 sorption measurement Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
- G06F18/15—Statistical pre-processing, e.g. techniques for normalisation or restoring missing data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
Abstract
The invention relates to the technical field of data denoising, in particular to a method and a system for monitoring dry-type waste gas treatment equipment based on data analysis, wherein the method comprises the following steps: obtaining an extremum of pressure data in a pipeline of the waste gas treatment equipment, and obtaining a density representation value of the extremum according to extremum distribution conditions in a preset window corresponding to the extremum; obtaining correction data corresponding to each extremum according to the fitting data and the density characterization value corresponding to each extremum; obtaining a noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum; obtaining the effective degree of each extremum according to the density representation value and the noise representation value of each extremum; EMD (empirical mode decomposition) is carried out on all pressure data in a set time period by utilizing the effective degree, and optimal pressure data are obtained according to decomposition results; and monitoring the dry exhaust treatment device according to the preferred pressure data. The invention can obtain more accurate equipment monitoring results.
Description
Technical Field
The invention relates to the technical field of data denoising treatment, in particular to a dry type waste gas treatment equipment monitoring method and system based on data analysis.
Background
The dry type waste gas treatment equipment is an environment-friendly equipment which can protect the environment and purify the air by recycling or removing and reducing harmful components of the discharged gas, and is particularly important to monitor, diagnose faults and the like in order to ensure that the equipment can safely run. In monitoring the dry exhaust gas treatment device, the line pressure data of the dry exhaust gas treatment device is often monitored, but due to the influence of objective factors such as errors of the sensor itself, environmental interference, transmission interference and the like, noise exists in the collected line pressure data of the dry exhaust gas treatment device, and thus deviation exists in the monitoring result of the dry exhaust gas treatment device.
The existing method utilizes an EMD decomposition algorithm to decompose the acquired pipeline pressure data of the dry-type waste gas treatment equipment, and reconstructs the acquired data according to the decomposed data to obtain new data, so that noise data in the original data can be effectively removed. However, the EMD decomposition algorithm decomposes through a local extremum, and due to noise in the original data, the extremum is obtained inaccurately, so that the EMD decomposition result is inaccurate, and further, the monitoring result of the dry exhaust gas treatment equipment is inaccurate.
Disclosure of Invention
In order to solve the technical problem that the monitoring result of the existing method on the dry type waste gas treatment equipment is inaccurate, the invention aims to provide a monitoring method of the dry type waste gas treatment equipment based on data analysis, and the adopted technical scheme is as follows:
respectively collecting pressure data in the pipeline of the waste gas treatment equipment at different moments in a set time period; obtaining extreme values of all pressure data in a set time period, and obtaining a density representation value of each extreme value according to extreme value distribution conditions in a preset window corresponding to each extreme value; the extremum includes a maximum and a minimum;
performing curve fitting on all maximum values and all minimum values in a preset window corresponding to each extremum respectively, and obtaining correction data corresponding to each extremum according to fitting data and density characterization values corresponding to each extremum; obtaining a noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum;
obtaining the effective degree of each extremum according to the density representation value and the noise representation value of each extremum; EMD (empirical mode decomposition) is carried out on all pressure data in a set time period by utilizing the effective degree, and optimal pressure data are obtained according to decomposition results;
and monitoring the dry exhaust treatment device according to the preferred pressure data.
Preferably, the obtaining the density representation value of each extremum according to the extremum distribution condition in the preset window corresponding to each extremum specifically includes:
and in a preset window corresponding to any extremum, acquiring a time interval between the moments corresponding to every two adjacent extremums in the window, taking the average value of the time intervals between the moments corresponding to every two adjacent extremums in the window as a time average value, and obtaining a density representation value of the extremum according to the time average value, wherein the time average value and the density representation value are in a negative correlation.
Preferably, the curve fitting is performed on all the maxima in the preset window corresponding to each extremum, and the curve fitting is performed on all the minima, so as to obtain correction data corresponding to each extremum according to the fitting data and the density characterization value corresponding to each extremum, which specifically includes:
marking any extremum as a selected extremum, in a preset window corresponding to the selected extremum, performing curve fitting on all maximum values in the window as a maximum value curve corresponding to the selected extremum, and performing curve fitting on all minimum values in the window as a minimum value curve corresponding to the selected extremum;
the time corresponding to the selected extremum is marked as the selected time;
if the selected extreme value is the maximum value, fitting data of a maximum value curve corresponding to each maximum value in the window at the selected moment is obtained, and a correction coefficient of each maximum value in the window is obtained according to the density characterization value of each maximum value in the window; taking the average value of products between correction coefficients of all maxima in the window and corresponding fitting data as correction data corresponding to the selected extremum;
if the selected extreme value is the minimum value, fitting data of a minimum value curve corresponding to each minimum value in the window at the selected moment is obtained, and a correction coefficient of each minimum value in the window is obtained according to the density characterization value of each minimum value in the window; and taking the average value of products between the correction coefficients of all the minima in the window and the corresponding fitting data as correction data corresponding to the selected extremum.
Preferably, the obtaining the correction coefficient of each maximum value in the window according to the density representation value of each maximum value in the window specifically includes:
respectively carrying out normalization processing on the density characterization value of each maximum value in the window corresponding to the selected extremum to obtain a correction coefficient of each maximum value in the window;
the density characterization value of each minimum value in the window obtains a correction coefficient of each minimum value in the window, and specifically comprises the following steps:
and respectively carrying out normalization processing on the density characterization value of each minimum value in the window corresponding to the selected extremum to obtain the correction coefficient of each minimum value in the window.
Preferably, the obtaining the noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum specifically includes:
for any extremum, acquiring a difference value between pressure data and correction data corresponding to each extremum in a window corresponding to the extremum, and taking the average value of the square sums of the difference values corresponding to all the extremums as a first error coefficient; obtaining a difference value between pressure data and correction data corresponding to each minimum value, and taking the average value of the square sums of the difference values corresponding to all the minimum values as a second error coefficient;
and taking the sum of the first error coefficient and the second error coefficient as a noise characterization value of the extreme value.
Preferably, the obtaining the validity degree of each extremum according to the density representation value and the noise representation value of each extremum specifically includes:
and for any extremum, calculating the product between the noise characterization value and the density characterization value of the extremum, and obtaining the effective degree of the extremum according to the product, wherein the product and the effective degree have a negative correlation.
Preferably, the method for acquiring the preset window specifically includes:
taking any extremum as a target extremum, and taking the moment corresponding to the target extremum as a center moment to acquire all extremum in a preset time length to form a characteristic extremum sequence corresponding to the target extremum;
clustering the extremum in the characteristic extremum sequence by using a density clustering algorithm, obtaining a cluster in which the target extremum is located in a clustering result, and taking the maximum value of the time interval between every two extremum corresponding moments in the cluster as the length of a preset window corresponding to the target extremum.
Preferably, the obtaining the preferable pressure data according to the decomposition result specifically includes:
the decomposition result comprises at least two IMF components, each IMF component is subjected to denoising treatment by utilizing wavelet threshold denoising, and the IMF components subjected to denoising treatment are reconstructed to obtain the optimal pressure data corresponding to each moment.
Preferably, the monitoring of the dry exhaust treatment device according to the preferred pressure data specifically includes:
calculating the average value of the preferable pressure data at all the moments in the set time period, marking the dry type waste gas treatment equipment when the average value of the pressure is larger than a preset pressure threshold value.
The invention also provides a dry exhaust treatment device monitoring system based on data analysis, comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the computer program realizes the steps of a dry exhaust treatment device monitoring method based on data analysis when being executed by the processor.
The embodiment of the invention has at least the following beneficial effects:
the method comprises the steps of firstly obtaining extreme values of pressure data in a pipeline of the waste gas treatment equipment, analyzing extreme value distribution conditions in a preset window corresponding to each extreme value to obtain density representation values of the extreme values, and reflecting the density degree of extreme value distribution of each extreme value in a corresponding local range by using the density representation values of each extreme value. And then, fitting the maximum value and the minimum value respectively in a preset window of each extremum, simulating the fitting process of the maximum value and the minimum value respectively in the EMD decomposition process, analyzing the noise condition of correction data corresponding to the extremum in a local range, and correcting the fitting data of the extremum by using the density characterization value to obtain correction data. Further, the difference condition between the correction data and the pressure data corresponding to each extremum is compared, a noise characterization value of the extremum is obtained, and the noise distribution condition of each extremum in the corresponding local range is reflected by the noise characterization value. Finally, the effective degree of each extremum is obtained by combining the distribution characteristics of the two aspects, the effect degree of envelope curve fitting by using the extremum is reflected, the result of EMD decomposition on all pressure data by using the effective degree is more accurate, and further, preferable pressure data with better denoising effect is obtained, so that more accurate monitoring results can be obtained by using the preferable pressure data.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a method flow diagram of a method of monitoring a dry exhaust treatment device based on data analysis in accordance with the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purposes, the following description refers to the specific implementation, structure, characteristics and effects of the method and system for monitoring the dry exhaust gas treatment device based on data analysis according to the present invention, with reference to the accompanying drawings and the preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
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.
The following specifically describes a specific scheme of a method and a system for monitoring a dry exhaust gas treatment device based on data analysis.
An embodiment of a method for monitoring a dry exhaust treatment device based on data analysis:
the main purpose of the embodiment of the invention is as follows: in the process of carrying out EMD decomposition on the acquired pipeline pressure data of the dry-type waste gas treatment equipment, the extremum in the process of routing the package is screened by using the effective degree of the extremum as the weight of the extremum, so that the IMF component obtained through EMD decomposition is more effective and accurate, the denoising effect with better management pressure data is obtained, and the more accurate equipment monitoring result is obtained.
Referring to FIG. 1, a flow chart of a method for monitoring a dry exhaust treatment device based on data analysis according to one embodiment of the present invention is shown, the method comprising the steps of:
step one, respectively collecting pressure data in pipelines of the waste gas treatment equipment at different moments in a set time period; obtaining extreme values of all pressure data in a set time period, and obtaining a density representation value of each extreme value according to extreme value distribution conditions in a preset window corresponding to each extreme value; the extremum includes a maximum and a minimum.
Firstly, in the operation process of the dry type waste gas treatment equipment, pressure data in the pipeline of the equipment needs to be monitored and collected at any time, so that when the pipeline pressure is too high, the internal emergency cylinder can be opened in time, and the equipment can be ensured to operate normally. Namely, a gas pressure gauge is arranged at the position of the main pipeline before the branch of the main adsorption cylinder pipeline and the internal emergency adsorption cylinder pipeline of the dry type waste gas treatment equipment and is used for detecting pressure data of the internal pipeline of the dry type waste gas treatment equipment.
In this embodiment, pressure data in the exhaust gas treatment device pipeline is collected at different times in the set time period, the time length of the set time period is set to 2s, the time interval between two adjacent times is set to 2ms, and the implementer can set according to the specific implementation scenario.
The empirical mode decomposition (Empirical Mode Decomposition, EMD) algorithm is to find the extreme points in the data to obtain the envelope of the local oscillation mode function, i.e. the intrinsic mode function (Intrinsic Mode Function, IMF), so the accuracy of the extreme point acquisition in the collected raw data is particularly important.
Based on this, it is necessary to first acquire all the extremum values, including the maximum value and the minimum value, in the pressure data at each time point in the set period. In this embodiment, curve fitting is performed by using pressure data at each moment in a set time period, and a maximum value point and a minimum value point on the curve are obtained by calculating a curve obtained by fitting, so that the pressure data at the moment corresponding to the maximum value point and the minimum value point on the curve are used as corresponding extremum data. In other embodiments, the practitioner may also obtain extremum data of the pressure data at all times during the set time period by using the method of obtaining extremum data in the discrete data.
In the operation process of the dry-type waste gas treatment equipment, in order to ensure that the waste gas treatment equipment can efficiently operate and avoid generating force impact on the equipment, the pressure change of waste gas is slower and gentle under the normal condition, the existence of noise can lead to the abnormal severe change condition of the collected pressure data, the time interval between adjacent extreme values in the data is shorter, the extreme value distribution is denser, and based on the situation, the extreme value can be screened by analyzing the distribution condition of the extreme value.
And obtaining the density distribution in the local range of each extremum by analyzing the extremum distribution condition in the local range of each extremum. In order to obtain the extremum density distribution condition of each extremum in the local range more accurately, the length of the preset window corresponding to each extremum is obtained in a self-adaptive mode according to the extremum distribution in the local range.
Taking any extremum as a target extremum, and taking the moment corresponding to the target extremum as a center moment to acquire all extremum in a preset time length to form a characteristic extremum sequence corresponding to the target extremum; in this embodiment, the preset time length is set to 50ms, and the practitioner can set according to the specific implementation scenario.
Clustering the extremum in the characteristic extremum sequence by using a density clustering algorithm, obtaining a cluster in which the target extremum is located in a clustering result, and taking the maximum value of the time interval between every two extremum corresponding moments in the cluster as the length of a preset window corresponding to the target extremum. In this embodiment, the extremum in the characteristic extremum sequence is clustered by using a DBSCAN clustering algorithm, which is a well-known technique and will not be described here too much.
The clustering result obtained after density clustering may have more clusters and may have fewer clusters, but the extreme values in different clusters have different densities and are distributed densely in the same cluster, so in this embodiment, the preset window length is obtained in a self-adaptive manner based on the time length between the extreme values in the cluster where the target extreme value is located. In the cluster where the target extremum is located, a corresponding time interval can be obtained between every two extremums, and in order to obtain the distribution of all the extremums in the cluster, the maximum value of the time interval between the moments corresponding to every two extremums in the cluster is further used as the length of a preset window corresponding to the target extremum.
Further, according to the extremum distribution condition in the preset window corresponding to each extremum, the density representation value of each extremum is obtained. Specifically, in a preset window corresponding to any extremum, acquiring a time interval between the moments corresponding to every two adjacent extremums in the window, taking the average value of the time intervals between the moments corresponding to every two adjacent extremums in the window as a time average value, and obtaining a density representation value of the extremum according to the time average value, wherein the time average value and the density representation value are in a negative correlation.
In this embodiment, the nth extremum is taken as an example, and in a preset window taking the time corresponding to the nth extremum as the center time, the time interval between the times corresponding to the two adjacent extremums reflects the distance between the two adjacent extremums in time sequence, and the smaller the time interval, the closer the distance between the two extremums in time sequence is.
The time mean value reflects the overall distribution condition of the time interval corresponding to each two adjacent extremum in the window, and the larger the value is, the larger the time interval between each two adjacent extremum in the window is, and further the extremum in the window is more discrete in distribution, and further the smaller the corresponding density representation value is, and the smaller the density of the extremum in the window is. The smaller the value of the time mean value is, the smaller the time interval between every two adjacent extremum values in the window is, the extremum values in the window are more densely distributed, the corresponding density representation value is larger, and the density of the extremum values in the window is larger. In this embodiment, the inverse of the time average is taken as the density representation value of the extremum, and the implementer may select other suitable methods to perform negative correlation mapping on the time average according to the specific implementation scenario.
Performing curve fitting on all maximum values in a preset window corresponding to each extremum, performing curve fitting on all minimum values, and obtaining correction data corresponding to each extremum according to fitting data and density characterization values corresponding to each extremum; and obtaining the noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum.
In the process of performing EMD decomposition on data, the envelopes of the maximum value and the minimum value need to be fitted respectively, namely, all the maximum values are fitted respectively to obtain an upper envelope, all the minimum values are fitted to obtain a lower envelope, and then the IMF component is acquired based on the upper envelope and the lower envelope. Because the obtained extremum is inaccurate due to the existence of noise, and errors can occur in fitting of the envelope curve, the fitting error corresponding to each extremum can be represented by analyzing the difference between the fitting data and the real data on the local envelope curve in the local range of each extremum, and the noise condition of each extremum in the local range can be reflected.
In the time sequence, the degree of density of the extreme points can reflect the accuracy of fitting errors, when the degree of density of the extreme points is higher, the fitting model has more data points which can be used for capturing and modeling the change characteristics in the data, so that a more accurate model is provided, and the difference caused by overlarge intervals among the data points can be reduced, so that in the embodiment, the error condition of the fitting result is analyzed, and meanwhile, the distribution condition of the degree of density corresponding to the extreme points is considered.
And marking any extremum as a selected extremum, in a preset window corresponding to the selected extremum, performing curve fitting on all maximum values in the window as a maximum value curve corresponding to the selected extremum, and performing curve fitting on all minimum values in the window as a minimum value curve corresponding to the selected extremum. According to the same method, in the preset window of the selected extremum, each extremum has a corresponding preset window and a corresponding density representation value, and then each extremum has a corresponding maximum value curve and minimum value curve.
The time corresponding to the selected extremum is marked as the selected time; the selected extremum and each extremum in the preset window corresponding to the selected extremum can acquire corresponding fitting data at the selected moment, and then the fitting data at the same moment on different fitting curves in the local range can be weighted and summed to correct the fitting data. Meanwhile, when the fitting data is corrected, the extreme value needs to be considered as the maximum value or the minimum value, and when the fitting data is corrected, the maximum value or the minimum value needs to be ensured.
If the selected extremum is the maximum value, fitting data of a maximum value curve corresponding to each maximum value in the window at the selected moment is obtained, and a correction coefficient of each maximum value in the window is obtained according to the density representation value of each maximum value in the window, namely, the density representation value of each maximum value in the window corresponding to the selected extremum is respectively normalized to obtain the correction coefficient of each maximum value in the window; and taking the average value of products between the correction coefficients of all the maxima in the window and the corresponding fitting data as correction data corresponding to the selected extremum.
If the selected extremum is a minimum value, fitting data of a minimum value curve corresponding to each minimum value in the window at the selected moment is obtained, a correction coefficient of each minimum value in the window is obtained according to a density representation value of each minimum value in the window, and normalization processing is carried out on the density representation value of each minimum value in the window corresponding to the selected extremum to obtain the correction coefficient of each minimum value in the window; and taking the average value of products between the correction coefficients of all the minima in the window and the corresponding fitting data as correction data corresponding to the selected extremum.
In this embodiment, taking the u-th maximum value as the selected extremum, the calculation formula of the correction data corresponding to the u-th maximum value may be expressed as:
wherein,represents the correction data corresponding to the u-th maximum value,represents the number of maximum points contained in the window corresponding to the u-th maximum,fitting data representing the ith maximum value at the selected moment in the window corresponding to the ith maximum value,indicating the time corresponding to the u-th maximum, i.e. the selected time,and the correction coefficient of the ith maximum value in the window corresponding to the ith maximum value is represented.
In this embodiment, the ratio between the density representation value of the ith maximum value in the window corresponding to the ith maximum value and the density representation values and values of all the maximum values in the window is used as the correction coefficient of the ith maximum value. And taking the correction coefficient as the weight of the fitting data corresponding to each maximum value in the window, namely, taking the sum of the correction coefficients of all the maximum values in the window corresponding to the ith maximum value as 1, and selecting other proper methods by an implementer for processing.
The smaller the intensity of the maximum value, the greater the possibility of error when fitting the maximum value in the window of the maximum value, so the smaller the weight occupied by the corresponding fitting data, and the smaller the value of the corresponding correction coefficient. The higher the intensity of the maxima, the less likely there is an error in fitting all maxima within the window of maxima, so the corresponding fitting data occupies a larger weight and the corresponding correction coefficient takes a larger value.
According to the method, the fitting data corresponding to each extremum are simply corrected, the obtained correction data avoid the influence of discrete noise points on the adjacent data to a great extent, the envelope curve fitting is carried out based on the corrected correction data, a better fitting result can be obtained, and the fitting result is compared with the original data to obtain the error condition between the original data and the fitting result.
And obtaining the noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum. Specifically, for any extremum, acquiring a difference value between pressure data and correction data corresponding to each extremum in a window corresponding to the extremum, and taking a mean value of square sums of the difference values corresponding to all the extremums as a first error coefficient; obtaining a difference value between pressure data and correction data corresponding to each minimum value, and taking the average value of the square sums of the difference values corresponding to all the minimum values as a second error coefficient; and taking the sum of the first error coefficient and the second error coefficient as a noise characterization value of the extreme value.
In this embodiment, the mean square error between the pressure data corresponding to each maximum value and the correction data and the mean square error between the pressure data corresponding to each minimum value and the correction data are calculated respectively, and the noise condition of the extremum in the local range is obtained by combining the error conditions of the two aspects. The noise characterization value reflects the noise distribution degree of the extremum in the corresponding preset window range, and the larger the value is, the larger the noise degree of the extremum in the local range is.
Thirdly, obtaining the effective degree of each extremum according to the density representation value and the noise representation value of each extremum; and carrying out EMD decomposition on all the pressure data in the set time period by using the effective degree, and obtaining preferable pressure data according to the decomposition result.
The density representation value of each extremum represents the density degree of extremum distribution in a local range, the noise representation value of each extremum represents the noise distribution degree of each extremum in the local range, and the effective condition of each extremum is analyzed by combining the characteristics of two aspects. And obtaining the effective degree of each extremum according to the density representation value and the noise representation value of each extremum.
Specifically, for any extremum, calculating the product between the noise characterization value and the density characterization value of the extremum, and obtaining the effective degree of the extremum according to the product, wherein the product and the effective degree have a negative correlation. In this embodiment, taking the nth extremum as an example for explanation, the calculation formula of the effective degree of the nth extremum can be expressed as:
wherein,indicating the degree of validity of the nth extremum,a density characterization value representing the nth extremum,the noise characterization value representing the nth extremum, exp () represents an exponential function based on the natural constant e.
The larger the value of the density representation value of the nth extremum is, the larger the distribution density degree of the extremum in the local range of the nth extremum is, the larger the corresponding possibility of noise exists, the larger the value of the noise representation value of the nth extremum is, the larger the noise degree in the local range of the nth extremum is, the smaller the effective degree of the corresponding extremum is, namely the extremum is more likely to be affected by noise, and the worse the effect of carrying out envelope curve fitting as real data is.
Further, EMD decomposition is performed on all pressure data within a set time period by using the effective degree. In the EMD decomposition process, the weight may be used to determine the position of the envelope, which describes the low frequency component of the data by connecting the extreme points in the data. And once the weight of the extreme point is confirmed, an envelope corresponding to the extreme point can be obtained, which is a known technology and will not be described in detail. The extreme points of higher weight have a greater influence on the shape and position of the envelope, corresponding to the dominant fluctuation mode in the data. In this embodiment, the effective degree of each extremum is used as the weight corresponding to the extremum, and EMD decomposition is performed.
Specifically, when the extremum and the effective degree of the extremum are utilized to perform envelope curve fitting, if one extremum point is needed to be screened out from a plurality of extremum points in a local range to be used as the vertex of the envelope curve, the extremum point corresponding to the maximum value of the effective degree is used as the vertex of the envelope curve. The extreme point selection method based on the weight can improve the fitting precision of the envelope curve and reduce errors caused by noise and details.
The preferred pressure data is obtained from the decomposition results. Specifically, the decomposition result includes at least two IMF components, denoising is performed on each IMF component by using wavelet threshold denoising, and the IMF components after denoising are reconstructed to obtain preferable pressure data corresponding to each moment.
In this embodiment, after EMD decomposition is performed on the data, a plurality of IMF components may be obtained, and a wavelet threshold is used to denoise each IMF component, and each IMF component after denoise is reconstructed to obtain denoised pipeline pressure data, and the denoised pipeline pressure data is recorded as preferred pressure data, where the wavelet basis function adopts a coidlet wavelet basis function, and the threshold selection adopts a soft threshold, which is a known technique and will not be described herein too much. The EMD decomposition can keep local features and details of the data, so that the denoised data can still accurately reflect the local changes and features of the original data.
And step four, monitoring the dry type waste gas treatment equipment according to the preferable pressure data.
In this embodiment, the average value of the preferred pressure data at all times in the set time period is calculated and recorded as a pressure average value, when the pressure average value is greater than a preset pressure threshold value, the dry-type exhaust gas treatment device is marked, and the relevant staff can continuously observe the marked device to perform further relevant treatment operation. The value of the pressure threshold value can be 80% of the maximum allowable pressure of the dry type waste gas treatment equipment pipeline, and an implementer can set according to a specific implementation scene.
In other embodiments, when the pressure average value is greater than a preset pressure threshold value, the emergency adsorption pipeline channel inside the dry type waste gas treatment equipment is opened, so that accidents caused by overlarge internal pressure management are avoided.
An embodiment of a dry exhaust treatment device monitoring system based on data analysis:
the embodiment provides a dry exhaust treatment device monitoring system based on data analysis, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the computer program is executed by the processor to realize the steps of a dry exhaust treatment device monitoring method based on data analysis. Since one embodiment of a method for monitoring a dry exhaust gas treatment device based on data analysis has been described in detail, it will not be described in detail.
A data processing method embodiment of a dry exhaust gas treatment device based on data analysis:
the existing data denoising method utilizes an EMD decomposition algorithm to decompose the acquired pipeline pressure data of the dry-type waste gas treatment equipment, and reconstructs the acquired data according to the decomposed data to obtain new data, so that noise data in the original data can be effectively removed. However, the EMD decomposition algorithm decomposes through a local extremum, and due to noise in the original data, the extremum is obtained inaccurately, so that the EMD decomposition result is inaccurate, and the data denoising effect is poor.
In order to solve the technical problem of poor data denoising effect in the existing method, the embodiment provides a data processing method of dry-type waste gas treatment equipment based on data analysis, which comprises the following steps:
step one, respectively collecting pressure data in pipelines of the waste gas treatment equipment at different moments in a set time period; obtaining extreme values of all pressure data in a set time period, and obtaining a density representation value of each extreme value according to extreme value distribution conditions in a preset window corresponding to each extreme value; the extremum includes a maximum and a minimum.
Performing curve fitting on all maximum values in a preset window corresponding to each extremum, performing curve fitting on all minimum values, and obtaining correction data corresponding to each extremum according to fitting data and density characterization values corresponding to each extremum; and obtaining the noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum.
Thirdly, obtaining the effective degree of each extremum according to the density representation value and the noise representation value of each extremum; and carrying out EMD decomposition on all the pressure data in the set time period by using the effective degree, and obtaining preferable pressure data according to the decomposition result.
The first to third steps are described in detail in the embodiment of the method for monitoring a dry exhaust gas treatment device based on data analysis, and are not described herein.
The embodiment has at least the following advantages:
the method comprises the steps of firstly obtaining extreme values of pressure data in a pipeline of the waste gas treatment equipment, analyzing extreme value distribution conditions in a preset window corresponding to each extreme value to obtain density representation values of the extreme values, and reflecting the density degree of extreme value distribution of each extreme value in a corresponding local range by using the density representation values of each extreme value. And then, fitting the maximum value and the minimum value respectively in a preset window of each extremum, simulating the fitting process of the maximum value and the minimum value respectively in the EMD decomposition process, analyzing the noise condition of correction data corresponding to the extremum in a local range, and correcting the fitting data of the extremum by using the density characterization value to obtain correction data. Further, the difference condition between the correction data and the pressure data corresponding to each extremum is compared, a noise characterization value of the extremum is obtained, and the noise distribution condition of each extremum in the corresponding local range is reflected by the noise characterization value. Finally, the effective degree of each extremum is obtained by combining the distribution characteristics of the two aspects, the effect degree of envelope curve fitting by using the extremum is reflected, the result of EMD decomposition on all pressure data by using the effective degree is more accurate, and further, the preferable pressure data with better denoising effect is obtained.
The above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the scope of the embodiments of the present application, and are intended to be included within the scope of the present application.
Claims (10)
1. A method of monitoring a dry exhaust treatment device based on data analysis, the method comprising the steps of:
respectively collecting pressure data in the pipeline of the waste gas treatment equipment at different moments in a set time period; obtaining extreme values of all pressure data in a set time period, and obtaining a density representation value of each extreme value according to extreme value distribution conditions in a preset window corresponding to each extreme value; the extremum includes a maximum and a minimum;
performing curve fitting on all maximum values and all minimum values in a preset window corresponding to each extremum respectively, and obtaining correction data corresponding to each extremum according to fitting data and density characterization values corresponding to each extremum; obtaining a noise characterization value of each extremum according to the difference condition between the pressure data and the correction data corresponding to each extremum;
obtaining the effective degree of each extremum according to the density representation value and the noise representation value of each extremum; EMD (empirical mode decomposition) is carried out on all pressure data in a set time period by utilizing the effective degree, and optimal pressure data are obtained according to decomposition results;
and monitoring the dry exhaust treatment device according to the preferred pressure data.
2. The method for monitoring a dry exhaust gas treatment device based on data analysis according to claim 1, wherein the obtaining the density representation value of each extremum according to the extremum distribution condition in the preset window corresponding to each extremum specifically comprises:
and in a preset window corresponding to any extremum, acquiring a time interval between the moments corresponding to every two adjacent extremums in the window, taking the average value of the time intervals between the moments corresponding to every two adjacent extremums in the window as a time average value, and obtaining a density representation value of the extremum according to the time average value, wherein the time average value and the density representation value are in a negative correlation.
3. The method for monitoring a dry exhaust gas treatment device based on data analysis according to claim 1, wherein the performing curve fitting on all maximum values and performing curve fitting on all minimum values in a preset window corresponding to each extremum respectively, and obtaining correction data corresponding to each extremum according to the fitting data and the density characterization value corresponding to each extremum specifically comprises:
marking any extremum as a selected extremum, in a preset window corresponding to the selected extremum, performing curve fitting on all maximum values in the window as a maximum value curve corresponding to the selected extremum, and performing curve fitting on all minimum values in the window as a minimum value curve corresponding to the selected extremum;
the time corresponding to the selected extremum is marked as the selected time;
if the selected extreme value is the maximum value, fitting data of a maximum value curve corresponding to each maximum value in the window at the selected moment is obtained, and a correction coefficient of each maximum value in the window is obtained according to the density characterization value of each maximum value in the window; taking the average value of products between correction coefficients of all maxima in the window and corresponding fitting data as correction data corresponding to the selected extremum;
if the selected extreme value is the minimum value, fitting data of a minimum value curve corresponding to each minimum value in the window at the selected moment is obtained, and a correction coefficient of each minimum value in the window is obtained according to the density characterization value of each minimum value in the window; and taking the average value of products between the correction coefficients of all the minima in the window and the corresponding fitting data as correction data corresponding to the selected extremum.
4. A method for monitoring a dry exhaust treatment device based on data analysis according to claim 3, wherein the obtaining the correction coefficient of each maximum value in the window according to the density representation value of each maximum value in the window specifically comprises:
respectively carrying out normalization processing on the density characterization value of each maximum value in the window corresponding to the selected extremum to obtain a correction coefficient of each maximum value in the window;
the density characterization value of each minimum value in the window obtains a correction coefficient of each minimum value in the window, and specifically comprises the following steps:
and respectively carrying out normalization processing on the density characterization value of each minimum value in the window corresponding to the selected extremum to obtain the correction coefficient of each minimum value in the window.
5. The method for monitoring a dry exhaust gas treatment device based on data analysis according to claim 1, wherein the obtaining the noise characterization value of each extremum according to the difference between the pressure data and the correction data corresponding to each extremum specifically comprises:
for any extremum, acquiring a difference value between pressure data and correction data corresponding to each extremum in a window corresponding to the extremum, and taking the average value of the square sums of the difference values corresponding to all the extremums as a first error coefficient; obtaining a difference value between pressure data and correction data corresponding to each minimum value, and taking the average value of the square sums of the difference values corresponding to all the minimum values as a second error coefficient;
and taking the sum of the first error coefficient and the second error coefficient as a noise characterization value of the extreme value.
6. The method for monitoring a dry exhaust treatment device based on data analysis according to claim 1, wherein the obtaining the validity degree of each extremum according to the density representation value and the noise representation value of each extremum specifically comprises:
and for any extremum, calculating the product between the noise characterization value and the density characterization value of the extremum, and obtaining the effective degree of the extremum according to the product, wherein the product and the effective degree have a negative correlation.
7. The method for monitoring a dry exhaust gas treatment device based on data analysis according to claim 1, wherein the method for acquiring the preset window specifically comprises:
taking any extremum as a target extremum, and taking the moment corresponding to the target extremum as a center moment to acquire all extremum in a preset time length to form a characteristic extremum sequence corresponding to the target extremum;
clustering the extremum in the characteristic extremum sequence by using a density clustering algorithm, obtaining a cluster in which the target extremum is located in a clustering result, and taking the maximum value of the time interval between every two extremum corresponding moments in the cluster as the length of a preset window corresponding to the target extremum.
8. The method for monitoring a dry exhaust gas treatment device based on data analysis according to claim 1, wherein the obtaining preferable pressure data based on the decomposition result specifically comprises:
the decomposition result comprises at least two IMF components, each IMF component is subjected to denoising treatment by utilizing wavelet threshold denoising, and the IMF components subjected to denoising treatment are reconstructed to obtain the optimal pressure data corresponding to each moment.
9. A method of monitoring a dry exhaust treatment device based on data analysis according to claim 1, wherein the monitoring of the dry exhaust treatment device based on the preferred pressure data comprises:
calculating the average value of the preferable pressure data at all the moments in the set time period, marking the dry type waste gas treatment equipment when the average value of the pressure is larger than a preset pressure threshold value.
10. A dry exhaust treatment device monitoring system based on data analysis, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the computer program when executed by the processor implements the steps of a dry exhaust treatment device monitoring method based on data analysis as claimed in any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311212448.5A CN116955938B (en) | 2023-09-20 | 2023-09-20 | Dry-type waste gas treatment equipment monitoring method and system based on data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311212448.5A CN116955938B (en) | 2023-09-20 | 2023-09-20 | Dry-type waste gas treatment equipment monitoring method and system based on data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116955938A CN116955938A (en) | 2023-10-27 |
| CN116955938B true CN116955938B (en) | 2023-12-29 |
Family
ID=88442843
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311212448.5A Active CN116955938B (en) | 2023-09-20 | 2023-09-20 | Dry-type waste gas treatment equipment monitoring method and system based on data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116955938B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117167903B (en) * | 2023-11-03 | 2024-01-30 | 江苏中安建设集团有限公司 | Artificial intelligence-based foreign matter fault detection method for heating ventilation equipment |
| CN117216489B (en) * | 2023-11-07 | 2024-01-26 | 山东正为新材料科技有限公司 | Waterproof coating quality inspection analysis method and system based on Internet |
| CN117349611B (en) * | 2023-12-06 | 2024-03-08 | 山东清控生态环境产业发展有限公司 | Water quality fluctuation instrument monitoring method based on big data analysis |
| CN117688311A (en) * | 2024-02-04 | 2024-03-12 | 深圳市纯水一号水处理科技有限公司 | Real-time monitoring method and system for advanced treatment of refractory salt-containing wastewater |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101482062A (en) * | 2008-01-08 | 2009-07-15 | 通用电气公司 | Methods and systems for providing real-time comparision with an alternate control strategy for a turbine |
| CN101568705A (en) * | 2007-06-08 | 2009-10-28 | 丰田自动车株式会社 | Exhaust gas purification system for an internal combustion engine |
| CN106096242A (en) * | 2016-06-01 | 2016-11-09 | 浙江浙能北海水力发电有限公司 | A kind of based on improving the Pressure Fluctuation in Draft Tube integrated evaluating method that EMD decomposes |
-
2023
- 2023-09-20 CN CN202311212448.5A patent/CN116955938B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101568705A (en) * | 2007-06-08 | 2009-10-28 | 丰田自动车株式会社 | Exhaust gas purification system for an internal combustion engine |
| CN101482062A (en) * | 2008-01-08 | 2009-07-15 | 通用电气公司 | Methods and systems for providing real-time comparision with an alternate control strategy for a turbine |
| CN106096242A (en) * | 2016-06-01 | 2016-11-09 | 浙江浙能北海水力发电有限公司 | A kind of based on improving the Pressure Fluctuation in Draft Tube integrated evaluating method that EMD decomposes |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116955938A (en) | 2023-10-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116955938B (en) | Dry-type waste gas treatment equipment monitoring method and system based on data analysis | |
| CN110717472A (en) | Fault diagnosis method and system based on improved wavelet threshold denoising | |
| Wróbel et al. | Improving fetal heart rate signal interpretation by application of myriad filtering | |
| CN109443752B (en) | Gear vibration signal noise reduction and fault diagnosis method based on VMD | |
| CN116304751B (en) | Operation data processing method for overhauling motor train unit components | |
| CN116451029A (en) | Dehumidifier working state early warning method | |
| CN116992393B (en) | Safety production monitoring method based on industrial Internet of things | |
| CN117131336B (en) | Data processing method for electronic equipment connector | |
| CN112945546A (en) | Accurate diagnosis method for complex fault of gear box | |
| CN112747921A (en) | Multi-sensor mechanical fault diagnosis method based on NA-MEMD | |
| CN115736944A (en) | Atrial fibrillation detection model MCNN-BLSTM based on short-time single lead electrocardiosignal | |
| CN116609440B (en) | Intelligent acceptance management method and system for building engineering quality based on cloud edge cooperation | |
| CN117251798A (en) | Meteorological equipment anomaly detection method based on two-layer progressive process | |
| CN112033656A (en) | Mechanical system fault detection method based on broadband spectrum processing | |
| CN112202630A (en) | Network quality abnormity detection method and device based on unsupervised model | |
| CN112237433B (en) | Electroencephalogram signal abnormity monitoring system and method | |
| CN113273992B (en) | Signal processing method and device | |
| CN112287835B (en) | Blade acoustic signal denoising method based on EWT-SE and wavelet threshold | |
| CN117541020B (en) | Scheduling management method and system for urban drainage pump station | |
| CN117788847A (en) | Method, system and related products for detecting respiratory rate of pigs | |
| CN117216489B (en) | Waterproof coating quality inspection analysis method and system based on Internet | |
| CN117454085B (en) | Vehicle online monitoring method and system | |
| CN113609207B (en) | Data preprocessing method for slope deformation monitoring data | |
| CN117434153B (en) | Road nondestructive testing method and system based on ultrasonic technology | |
| CN113642407B (en) | Feature extraction optimization method suitable for predicting residual service life of bearing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |