CN117517596B - Method and system for monitoring combustible and toxic harmful gases in real time based on Internet of things - Google Patents
Method and system for monitoring combustible and toxic harmful gases in real time based on Internet of things Download PDFInfo
- Publication number
- CN117517596B CN117517596B CN202410021934.7A CN202410021934A CN117517596B CN 117517596 B CN117517596 B CN 117517596B CN 202410021934 A CN202410021934 A CN 202410021934A CN 117517596 B CN117517596 B CN 117517596B
- Authority
- CN
- China
- Prior art keywords
- data
- concentration
- area
- value
- values
- 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
- 239000007789 gas Substances 0.000 title claims abstract description 46
- 238000012544 monitoring process Methods 0.000 title claims abstract description 43
- 238000000034 method Methods 0.000 title claims abstract description 37
- 231100000331 toxic Toxicity 0.000 title claims abstract description 27
- 230000002588 toxic effect Effects 0.000 title claims abstract description 27
- 238000009826 distribution Methods 0.000 claims abstract description 54
- 230000002159 abnormal effect Effects 0.000 claims abstract description 35
- 239000006185 dispersion Substances 0.000 claims description 46
- 238000005070 sampling Methods 0.000 claims description 39
- 230000000875 corresponding effect Effects 0.000 claims description 24
- 238000010586 diagram Methods 0.000 claims description 11
- 238000013480 data collection Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 230000002596 correlated effect Effects 0.000 claims description 7
- 238000003064 k means clustering Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 238000009825 accumulation Methods 0.000 claims description 2
- 238000004220 aggregation Methods 0.000 abstract description 3
- 230000002776 aggregation Effects 0.000 abstract description 3
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 10
- 230000008569 process Effects 0.000 description 7
- 230000008859 change Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 238000003912 environmental pollution Methods 0.000 description 3
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 2
- 229910002091 carbon monoxide Inorganic materials 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
- G01N33/0063—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display using a threshold to release an alarm or displaying means
-
- 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
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
- Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Food Science & Technology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Immunology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Medicinal Chemistry (AREA)
- Artificial Intelligence (AREA)
- Combustion & Propulsion (AREA)
- General Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of gas anomaly monitoring, in particular to a method and a system for monitoring combustible, toxic and harmful gases in real time based on the Internet of things. Acquiring real-time concentration data and historical concentration data of gas to be detected, dividing the historical concentration data into data categories, analyzing the distribution density condition of data values in each data category, and acquiring a data concentration area; and finally, obtaining an adjustment weight according to the distribution difference condition of the data value in the maximum data-intensive area and the data value in the historical concentration data, and monitoring the real-time concentration data by utilizing the adjustment weight to obtain an abnormal concentration threshold value. According to the method, the distribution aggregation condition of the historical concentration data is analyzed, so that an accurate abnormal concentration threshold value is obtained to monitor the real-time data, and the accuracy and the reliability of the monitoring result are improved.
Description
Technical Field
The invention relates to the technical field of gas anomaly monitoring, in particular to a method and a system for monitoring combustible, toxic and harmful gases in real time based on the Internet of things.
Background
Combustible and toxic harmful gases are one of the main causes of environmental pollution and safety accidents, and the accidents cause serious and serious damages to society and environment, thereby not only bringing great threat to life and property safety of people, but also causing environmental pollution and ecological damage. Therefore, the method for monitoring the combustible and toxic harmful gases in real time has important significance for preventing and reducing accidents.
When combustible and toxic harmful gases are monitored, concentration data are monitored, if abnormal data are monitored, early warning is carried out, the prior art generally utilizes a box diagram to monitor the gas concentration data abnormally, but the method does not consider the overall distribution condition of the data, only the upper edge value of the box diagram is obtained through an empirical coefficient, the obtained upper edge value is inaccurate, and the abnormal monitoring effect is inaccurate.
Disclosure of Invention
In order to solve the technical problems that in the prior art, the gas concentration data is normally monitored abnormally by using a box diagram, but the overall distribution condition of the data is not considered, and the abnormal monitoring effect is inaccurate due to the fact that the upper edge value of the box diagram is obtained only through an empirical coefficient, the invention aims to provide a combustible and toxic harmful gas real-time monitoring method and system based on the Internet of things, and the adopted technical scheme is as follows:
Acquiring real-time concentration data of gas to be detected and historical concentration data in a preset time period;
obtaining at least two data categories according to the similar conditions of all sampling point data values in the historical concentration data; taking any one data category as a data category to be measured, and acquiring a data concentration area according to the distribution density condition of data values in the data category to be measured;
acquiring a subsequence according to sequential continuous conditions of sampling points corresponding to data values in the data collection area; obtaining a similarity value according to the length difference and the similarity condition of adjacent subsequences; obtaining the data dispersion of the data concentration area according to the difference of the similarity values among the subsequences and the distribution condition of the subsequences on time sequence;
obtaining the relative concentration of each data concentration area according to the data quantity and the data dispersion contained in the data concentration areas in all the data categories; screening the maximum data-intensive area from all the data-intensive areas according to the relative concentration degree; obtaining adjustment weights according to the distribution difference condition of the data values in the maximum data-intensive area and the data values in the historical concentration data, the difference of the relative densities of the maximum data-intensive area and other data-intensive areas and the distribution condition of the data values in all data categories;
And obtaining an abnormal concentration threshold value according to the distribution of the data values in the historical concentration data and the adjustment weight, and monitoring the real-time concentration data.
Further, the data dispersion of the data set area is obtained according to the difference of the similarity values among the subsequences and the distribution condition of the subsequences on time sequence, and a formula model of the data dispersion is as follows:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>Representing the category +.>Data dispersion of the region in the dataset, +.>Representing the category +.>Is +.>The time value of the last sample point of the sub-sequence is equal to +.>Difference in time values of first sample point of sub-sequence, +.>Representing the category +.>Is +.>The time value of the last sample point of the sub-sequence is equal to +.>Difference in time values of first sample point of sub-sequence, +.>Representing the category +.>Is +.>Subsequence and->The similarity value between the sub-sequences,representing the category +.>Is +.>Subsequence and->Similarity between subsequences, < >>Representing the category +.>Is a total number of subsequences in the region in the dataset.
Further, the obtaining the relative density of each data set area according to the data amount and the data dispersion included in the data set areas in all the data categories includes:
taking the ratio of the data quantity contained in each data concentration area to the data quantity contained in all the data concentration areas as the data duty ratio of each data concentration area;
the value obtained by carrying out negative correlation mapping on the ratio of the data dispersion of each data set area to the data dispersion accumulation sum of all the data set areas is used as the data concentration of each data set area;
and obtaining the relative concentration of each data concentration area according to the data duty ratio and the data concentration of each data concentration area, wherein the data duty ratio and the data concentration are positively correlated with the relative concentration.
Further, the obtaining the adjustment weight according to the distribution difference between the data value in the maximum data-intensive area and the data value in the historical concentration data, the difference between the relative concentration of the maximum data-intensive area and other data-intensive areas, and the distribution of the data value in all data categories includes:
acquiring the lower quartile of all data values in the historical concentration data, and taking the difference value between the average value of all data values in the maximum data-intensive area and the lower quartile as a first important factor of the maximum data-intensive area;
Taking the average value of the difference between the relative density of the maximum data-intensive area and the relative density of all other data-intensive areas as a second important factor of the maximum data-intensive area;
obtaining an adjustment weight according to the first important factor and the second important factor of the maximum data-intensive area;
calculating the average value of all the data values in each data category to obtain an average value characteristic value of each data category, and when the data in the maximum data-intensive area belongs to the data category corresponding to the maximum average value characteristic value, positively correlating the first important factor and the second important factor with the adjustment weight; when the data in the maximum data-intensive area does not belong to the data category corresponding to the maximum mean value characteristic value, the first important factor and the second important factor are in negative correlation with the adjustment weight.
Further, the monitoring of the real-time concentration data according to the distribution of the data values in the historical concentration data and the abnormal concentration threshold obtained by the adjustment weight includes:
acquiring an upper quartile of a data value in the historical concentration data, and acquiring a quartile range according to the upper quartile and the lower quartile corresponding to the historical concentration data;
Acquiring an upper edge value in a box diagram of historical concentration data according to the upper quartile, the quartile range, the adjustment weight and the preset experience coefficient, wherein the upper quartile, the quartile range, the adjustment weight and the preset experience coefficient are positively correlated with the upper edge value; taking the upper edge value as the abnormal concentration threshold value;
and monitoring the real-time concentration data based on the abnormal concentration threshold, and performing early warning when the data value of the real-time concentration data is larger than the abnormal concentration threshold.
Further, the obtaining at least two data categories according to the similarity of all the sampling point data values in the historical concentration data includes:
clustering the data values of all sampling points in the historical concentration data based on a K-means clustering algorithm to obtain at least two clusters, wherein the data in each cluster is a data category; wherein the distance measure at the time of clustering is the difference between the data values.
Further, in the data category to be measured, the acquiring the data set area according to the distribution density condition of the data values includes:
and carrying out density analysis on the data in the data category to be detected based on a kernel density estimation algorithm, and acquiring a concentrated area of all the data in the data category to be detected as a data concentrated area.
Further, the step of obtaining the subsequence according to the sequential condition of the sampling points corresponding to the data values in the data collection area includes:
the sampling points of all the data values in the data collection area are sequentially ordered, and the data values of the sampling points with continuous moments are used as a subsequence; if the sampling point is not continuous with other sampling points, the data value of the sampling point is also used as a subsequence.
Further, the obtaining the similarity value according to the length difference and the similarity of the adjacent subsequences includes:
obtaining the similarity value of the two adjacent subsequences according to the length difference and the DTW distance of the two adjacent subsequences; the length difference and the DTW distance are both inversely related to the similarity value.
The invention also provides a combustible and toxic harmful gas real-time monitoring system based on the Internet of things, which comprises:
a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of any one of the methods when the computer program is executed.
The invention has the following beneficial effects:
firstly, acquiring real-time concentration data of gas to be detected and historical concentration data in a preset time period, and analyzing the historical concentration data to realize monitoring of the real-time concentration data; the data values of all sampling points in the historical concentration data can be initially classified according to the similarity conditions of the data values, so that at least two data categories are obtained; because the abnormal data are individual data and most of the data are normal data, the data which are distributed in a concentrated way are normal data to a great extent, so that the data intensity can be analyzed, and the final abnormal concentration threshold value can be determined; analyzing the distribution density condition of the data values in each data category, and acquiring a data set area; further, similarity and data dispersion analysis are performed on the data values in the data concentration area, so that the relative concentration of the data concentration area is obtained, the maximum data concentration area is further determined, the overall trend of the data is represented by the maximum data concentration area, and therefore, finally, the adjustment weight is obtained according to the distribution difference condition of the data values in the maximum data concentration area and the data values in the historical concentration data, the difference of the relative concentration of the maximum data concentration area and other data concentration areas, and the distribution condition of the data values in all data categories, and the abnormal concentration threshold is obtained by utilizing the adjustment weight and the distribution of the data values in the historical concentration data, so that the real-time concentration data is monitored. In summary, the invention analyzes the distribution aggregation condition of the historical concentration data to obtain the adjustment weight, so that the overall trend of the data can be more accurately mastered, the more accurate abnormal concentration threshold value is obtained to monitor the real-time data, and the accuracy and the reliability of the monitoring result are improved.
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 flowchart of a method for monitoring combustible and toxic and harmful gases in real time based on the internet of things according to an embodiment of the invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of the combustible and toxic and harmful gas real-time monitoring method and system based on the internet of things according to the invention by combining the accompanying drawings and the preferred embodiment. 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 invention provides a combustible and toxic and harmful gas real-time monitoring method and a system based on the Internet of things, which are specifically described below with reference to the accompanying drawings.
Referring to fig. 1, a method flowchart of a method for monitoring combustible and toxic and harmful gases in real time based on internet of things according to an embodiment of the invention is shown, the method comprises the following steps:
step S1: and acquiring real-time concentration data of the gas to be detected and historical concentration data in a preset time period.
In the production process of a factory, concentration monitoring of combustible and toxic and harmful gases is a very necessary step, and abnormal concentration data can be timely early-warned in the real-time monitoring process according to historical concentration data of the combustible and toxic and harmful gases; by the method, abnormal conditions can be found in time, and corresponding measures are taken to avoid environmental pollution and safety accidents.
In the process of monitoring combustible and toxic and harmful gases, data acquisition is a very important step. In order to monitor the concentration of combustible and toxic harmful gases in real time, data needs to be collected by sensors. The gas concentration of combustible gas and toxic and harmful gas is obtained by installing corresponding sensors in a factory. For example, for carbon monoxide and methane, a carbon monoxide sensor and a methane sensor may be installed correspondingly.
In the embodiment of the invention, methane gas is used as the gas to be detected for analysis, and the concentration data of the methane gas is obtained through a methane sensor. Firstly, acquiring historical concentration data and real-time concentration data in a preset time period in a database; real-time concentration data is monitored by analyzing the historical concentration data. It should be noted that, the specific gas concentration data collection device operator may adjust according to the implementation scenario, and is not limited herein; the preset time period is one week in the past, the sampling interval is acquired every 5 minutes, and the specific time period and the sampling interval implementation can be adjusted according to the implementation scene, so that the method is not limited.
Step S2: obtaining at least two data categories according to the similarity of all sampling point data values in the historical concentration data; and taking any one data category as a data category to be measured, and acquiring a data concentration area according to the distribution density condition of the data values in the data category to be measured.
The box line graph can be used for monitoring abnormal data, the box line graph regards data exceeding the upper edge value and the lower edge value as the abnormal data, but inaccurate setting of the upper edge value and the lower edge value can cause inaccurate abnormal monitoring results and deviation, so that the overall trend of the data needs to be analyzed, and more accurate edge values are obtained. Since the embodiment of the invention analyzes based on the concentration data of combustible and toxic harmful gases in a factory, the abnormal data in the data is usually high concentration data, so that only the upper edge value of the box diagram is adjusted and corrected in the embodiment of the invention.
Since the dense region of data inside the box is typically a region in the normal dataset, and when the location of the data dense region within the box is different, the subsequent determination of the upper edge value should also be different, so the dataset region needs to be acquired. All data values in the historical concentration data may be initially divided with the objective of classifying them into categories with significant differences in concentration.
Preferably, in one embodiment of the present invention, obtaining at least two data categories based on similar conditions of data values of all sampling points in the historical concentration data includes:
classifying the data values of all sampling points in the historical concentration data, and clustering the data values of all sampling points in the historical concentration data based on a K-means clustering algorithm in the embodiment of the invention, so as to obtain at least two clusters, wherein the data in each cluster is used as a data category; wherein the distance measure at the time of clustering is the difference between the data values. It should be noted that the K-means clustering algorithm is a technical means well known to those skilled in the art, and is not described herein in detail; in the embodiment of the invention, three clusters, namely three data types, are acquired, and the purpose is that the data types can be corresponding to three areas in a box diagram, wherein the cluster with the largest average value of gas concentration corresponds to the upper area of the box, the cluster with the second average value of gas concentration corresponds to the middle area of the box, the cluster with the smallest gas concentration corresponds to the lower area of the box, the concentration data in the upper area is high concentration data, the concentration data in the middle area is medium concentration data, and the concentration data in the lower area is low concentration data; the number of clusters can also be adjusted according to implementation scenes, and at least two clusters are provided, and the specific number is not limited herein.
To this end, the historical concentration data may be initially divided into three data categories, and then the data set conditions in each data category may be further analyzed. For convenience of explanation and explanation, any one data class is used as the data class to be tested herein for explaining the subsequent process.
Preferably, in one embodiment of the present invention, in the class of data to be measured, acquiring the data set area according to the distribution density of the data values includes:
the data in the data category to be tested can be subjected to density analysis based on a kernel density estimation algorithm, and a centralized area of all data in the data category to be tested is obtained and is used as a data centralized area of the data category to be tested. It should be noted that, the process of acquiring the data set area may be briefly described as: the core of the kernel density estimation is to select a proper kernel function, wherein the common kernel function comprises a Gaussian kernel, a polynomial kernel, a sigmoid kernel and the like, the Gaussian kernel can be selected in the embodiment of the invention, then the bandwidth is set, the proper bandwidth can be selected by adopting a cross verification method and the like, finally parameters such as data, the kernel function, the bandwidth and the like are input into a kernel density estimation algorithm, a kernel density estimation result can be obtained, a data concentration area can be determined according to the kernel density estimation result, and the data concentration area can be determined by adopting a method of setting density quantiles or calculating a threshold value of probability density and the like; the kernel density estimation algorithm is a technical means well known to those skilled in the art, and therefore, a detailed process is not described herein.
Thus, the data concentration area in the data category to be tested can be obtained, and the data concentration area can be further analyzed later.
Step S3: acquiring a subsequence according to sequential continuous conditions of sampling points corresponding to data values in a data collection area; obtaining a similarity value according to the length difference and the similarity condition of adjacent subsequences; and obtaining the data dispersion of the data set area according to the difference of the similarity values among the subsequences and the distribution condition of the subsequences on time sequence.
In the process, the data concentration area in the data category to be detected is obtained, and then the data concentration area can be analyzed for data dispersion, so that the maximum data concentration area in the box body can be conveniently determined, and the overall trend of the data is obtained.
When the data dispersion of the data concentration area is acquired, the data in the data concentration area is required to be analyzed, the subsequence can be acquired according to the time sequence continuity of the sampling points corresponding to the data values in the data concentration area, and the discrete points are formed into the subsequence, so that the change condition of the data can be more intuitively reflected, and the discrete condition of the data can be better analyzed.
Preferably, in one embodiment of the present invention, the sub-sequence is acquired according to sequential conditions of sampling points corresponding to data values in the data collection area, including:
Firstly, the sampling points of all data values in a data collection area can be ordered according to time sequence, and then the data values of sampling points with continuous moments are used as a subsequence; meanwhile, if a certain sampling point is not continuous with other sampling points, the data value of the sampling point is also used as a subsequence. In this way, the discrete data is incorporated into the sub-sequence, so that the overall characteristics and rules of the data, such as the discrete condition, the change trend and the like of the data, can be more comprehensively and intuitively analyzed.
After the subsequences are acquired, similarity analysis may be performed on adjacent subsequences, in preparation for subsequent discrete acquisition of the region in the dataset.
Preferably, in one embodiment of the present invention, obtaining the similarity value according to the length difference and the similarity of the adjacent sub-sequences includes:
since the DTW distance can be used to measure the similarity between two time sequences, the similarity value of two adjacent subsequences can be obtained according to the DTW distance of the two adjacent subsequences in combination with the length difference of the subsequences, and both the length difference and the DTW distance are inversely related to the similarity value. Adjacent two subsequences are in the firstSubsequence and- >For example, the formula model of the similarity value may be, for example:
wherein,indicate->Subsequence and->Similarity value of subsequence->Indicate->Subsequence and->DTW distance of subsequence, < >>Indicate->Length of subsequenceDegree (f)>Indicate->Length of subsequence, < ->Representing a preset first parameter.
In the formula model of similarity values, when the value of the DTW distance between two adjacent subsequences, i.eThe smaller the two subsequences are, the more similar the two subsequences can be said, while when the two adjacent subsequences are different in length +.>The smaller the two sub-sequences, the more consistent the data amount contained in the two sub-sequences, the greater the likelihood that the two sub-sequences will be similar, and thus the two sub-sequences are multiplied and then compared with the preset first parameter->The added value is taken as a denominator, so that logic relation correction is realized, and the similarity value of the two subsequences is obtained, wherein the smaller the difference of the lengths of the two subsequences is, and the smaller the DTW distance is, the more similar the two subsequences are. It should be noted that the first parameter is preset +.>The function of (2) is to prevent the denominator from being 0, and the value can be 0.001, and the specific value can be adjusted according to the implementation scene, and is not limited herein. It should be noted that, the DTW distance obtaining method is a technical means well known to those skilled in the art, and will not be described herein.
Thus, the similar values of the adjacent subsequences are obtained by analyzing the length differences and the similar conditions of the adjacent subsequences, the data dispersion of the region in the data set can be obtained based on the similar values of the subsequences in the region in the data set of the data category to be tested, and the distribution of the subsequences in time sequence can be used as one of indexes for evaluating the data dispersion because the subsequences are obtained according to the time sequence continuity of the sampling points.
Preferably, in one embodiment of the present invention, obtaining the data dispersion of the data set region according to the difference of the similarity values between the subsequences and the distribution of the subsequences in time sequence includes:
after the similarity values of the adjacent subsequences are obtained, the difference of the similarity values between the subsequences is further analyzed, so that the approaching degree of the subsequences in the data concentration area in the aspects of change trend, fluctuation condition and the like can be more intuitively reflected, and the approaching degree is used as one of indexes for evaluating the data dispersion; in addition, the distribution of the subsequences in time sequence, such as the subsequences are more concentrated or more scattered in time sequence, can also provide information about the data dispersion, so that the two are combined to obtain the data dispersion of the data concentration area. The formula model of the data dispersion is:
Wherein,representing the category +.>Data dispersion of the region in the dataset, +.>Representing the category +.>Is +.>The time value of the last sample point of the sub-sequence is equal to +.>Individual subsequenceDifference in time value of first sample point of column, +.>Representing the category +.>Is +.>The time value of the last sample point of the sub-sequence is equal to +.>Difference in time values of first sample point of sub-sequence, +.>Representing the category +.>Is +.>Subsequence and->Similarity between subsequences, < >>Representing the category +.>Is +.>Subsequence and->Similarity between subsequences, < >>Representing the category +.>Is a total number of subsequences in the region in the dataset.
In the formula model of data dispersion, since the similarity value is a similarity value between adjacent two subsequences, therefore,the similarity between the adjacent three subsequences can be indirectly represented, and the smaller the value is, the more similar the adjacent three subsequences can be considered, and the more concentrated the distribution of the data on the adjacent three subsequences can be explained, namely, the lower the dispersion of the data is; at the same time, the difference between the time value obtained from the last sampling point of the next sub-sequence and the time value obtained from the first sampling point of the adjacent previous sub-sequence is calculated, namely- >The difference reflects the temporal distribution of the subsequences, and therefore, the +.>The time sequence distribution situation between three adjacent subsequences can be characterized, and the smaller the value is, the more concentrated the three adjacent subsequences are, namely the stronger the periodicity between the three adjacent subsequences is, the lower the dispersion of the data is; therefore, will->And (3) withCombining and accumulating to obtain data dispersion of the data set region>。
Thus, the data dispersion of the data set areas of all data categories can be obtained based on the steps, and then the area with the most dense data can be screened out.
Step S4: obtaining the relative concentration of each data concentration area according to the data quantity and the data dispersion contained in the data concentration areas in all the data categories; screening the maximum data-intensive area from all the data-intensive areas according to the relative concentration; and obtaining adjustment weights according to the distribution difference condition of the data values in the maximum data-intensive area and the data values in the historical concentration data, the difference of the relative densities of the maximum data-intensive area and other data-intensive areas and the distribution condition of the data values in all data categories.
After the data concentration areas and the data dispersion thereof in all the data categories are obtained, the data concentration areas can be combined with the data quantity contained in the data concentration areas, the relative concentration of each data concentration area is obtained, the relative concentration can represent the relative relation between the data dispersion of each data concentration area and other data concentration areas and the data quantity, and more accurate results can be provided for the subsequent process, so that the relative concentration of each data concentration area is used as an index for screening the maximum data concentration area.
Preferably, in one embodiment of the present invention, obtaining the relative concentration of each data set region according to the data amount and the data dispersion included in the data set regions in all data categories includes:
the ratio of the data volume contained in each data concentration area to the data volume contained in all data concentration areas is firstly used as the data duty ratio of each data concentration area, and the data duty ratio represents the relative relation between the data volume of each data concentration area and the data volume of all data concentration areas. And then carrying out negative correlation mapping on the ratio of the data dispersion of each data set region to the sum of the data dispersion of all the data set regions to obtain a value which is used as the data concentration of each data set region, wherein the data concentration represents the relative relation between the data dispersion of each data set region and the data dispersion of all the data set regions.
And finally, the relative concentration of each data concentration area can be obtained according to the data occupation ratio and the data concentration of each data concentration area, and the data occupation ratio and the data concentration are positively related to the relative concentration. In the first placeFor example, the formula model of the relative concentration may be specifically, for example:
wherein,indicate->Relative concentration of the regions in the individual dataset, +.>Indicate->Data volume contained in the respective data set area, +.>Indicate->Data volume contained in the respective data set area, +.>Representing the total number of regions in the dataset, +.>Indicate->Data dispersion of the regions in the data set, < >>Indicate->Data dispersion for the regions in the data set.
Male at relative concentrationIn the formula model, calculating the ratio of the data volume of each data set area to the data volume of all data set areas to obtain the data duty ratioThe value may characterize the ratio of the data volume of a certain data set area to the data volume of all data set areas, and the larger the value is, the larger the data volume of a certain data set area is, the larger the data density of the data set area is reflected; likewise, the ratio of the data dispersion of each data set region to the cumulative sum of the dispersions of all data set regions is calculated, i.e. >The smaller the value, the smaller the value is, the larger the data density of the data concentration area is reflected, the more concentrated the data distribution is, and therefore, the negative correlation mapping is carried out, and the logic relation correction is completed, so that the data concentration is obtained>And finally multiplying the data with the data duty ratio to obtain the relative density.
Thus, the relative concentration degree of each data concentration area can be obtained, the relative concentration degree represents the concentration degree and the density degree of data distribution, and therefore the data concentration area corresponding to the maximum relative concentration degree is taken as the maximum data concentration area.
After the maximum data-intensive area is obtained, the adjustment weight can be determined according to the data category to which the maximum data-intensive area belongs, the distribution condition of the historical concentration data and the like.
Preferably, in one embodiment of the present invention, the obtaining the adjustment weight according to a distribution difference between the data value in the maximum data-intensive area and the data value in the historical concentration data, a difference between the relative densities of the maximum data-intensive area and other data-intensive areas, and a distribution of the data value in all data categories includes:
Firstly, the lower quartile of all data values in the historical concentration data is obtained, and the difference value between the average value of all data values in the maximum data-intensive area and the lower quartile is used as a first important factor of the maximum data-intensive area.
The average of the differences between the relative densities of the region of maximum data density and the relative densities of all other data set regions is then taken as a second important factor for the region of maximum data density.
Finally, obtaining an adjustment weight according to the first important factor and the second important factor of the maximum data-intensive area, calculating the average value of all data values in each data category, obtaining an average value characteristic value of each data category, and when the data in the maximum data-intensive area belongs to the data category corresponding to the maximum average value characteristic value, namely the upper area of the box body in the step S2, the first important factor and the second important factor are positively correlated with the adjustment weight; when the data in the region with the maximum data density does not belong to the data category corresponding to the maximum mean value characteristic value, that is, when the data in the region with the maximum data density belongs to the middle region or the lower region in the step S2, both the first important factor and the second important factor are inversely related to the adjustment weight. The formula model for adjusting the weight may specifically be, for example:
Wherein,representing adjustment weights, ++>Representing the mean value of all data values in the region of maximum data density, +.>Representing the lower quartile,/->Representing the relative concentration of the area of maximum data density, +.>Represents +.>Relative concentration of the regions in the individual dataset, +.>Representing a preset second parameter, ">Indicating that the adjustment parameters are to be used,representing the normalization function.
In the formula model for adjusting the weight, the difference between the average value and the lower quartile of all the data values in the maximum data-intensive area is taken as a first important factorThe lower quartile is the bottom of the box in the box diagram, and when the value of the first important factor is larger, the larger the maximum data density area is, the farther the maximum data density area is from the bottom of the box; at the same time, a second important factor is obtained according to the difference between the relative density of the maximum data-intensive region and the relative density of other data-intensive regionsThe larger the value, the higher the importance of the region of maximum data density inside the box; the first important factor and the second important factor can be combined and the adjustment parameter +.>When the maximum data-intensive area belongs to the data category corresponding to the maximum mean value characteristic value, the maximum data-intensive area is indicated to be in the upper area of the box body, namely, most data are data with larger data values, Then the adjustment weight should be made larger in order to avoid excluding normal data when subsequently determining the upper edge value of the box, so the adjustment parameter +.>It can be set to 1, that is, the first important factor and the second important factor are both positively correlated with the adjustment weight; conversely, if the maximum data-intensive region does not belong to the data category corresponding to the maximum mean characteristic value, it is indicated that the maximum data-intensive region is in the middle and lower regions of the case, that is, most of the data is medium-concentration data or low-concentration data, and in order to avoid the abnormal data from being included in the normal data when the upper edge value of the case is determined later, the adjustment weight should be reduced, so the adjustment parameter +_>It may be set to-1, i.e. both the first importance factor and the second importance factor are inversely related to the adjustment weight. It should be noted that the second parameter is preset +.>The value may be 0.001, in order to prevent the denominator from being 0, and the specific value may be adjusted according to the implementation scenario, which is not limited herein.
So far, the adjustment weight is obtained according to the position of the maximum data-intensive area inside the box body.
Step S5: and obtaining an abnormal concentration threshold value according to the distribution of the data values in the historical concentration data and the adjustment weight, and monitoring the real-time concentration data.
Because the adjustment weight integrates the distribution characteristics, the discrete conditions and the change relation of the data, the value of the upper edge line in the box diagram is calculated based on the adjustment weight, which is more in accordance with the overall change condition of the data, and more accurate monitoring results can be obtained.
Preferably, in one embodiment of the present invention, the monitoring of the real-time concentration data according to the distribution of the data values in the historical concentration data and the adjustment weight to obtain the abnormal concentration threshold includes:
first, the upper quartile of the data value in the historical concentration data is obtained, and the quartile distance is obtained according to the upper quartile and the lower quartile corresponding to the historical concentration data.
And then, obtaining an upper edge value in the box diagram of the historical concentration data according to the upper quartile, the quartile distance, the adjustment weight and the preset experience coefficient, wherein the upper quartile, the quartile distance, the adjustment weight and the preset experience coefficient are positively correlated with the upper edge value. The formula model of the upper edge value may specifically be, for example:
wherein,representing the upper edge value +_>Representing the upper quartile,/->Represents the quarter bit distance, ">Representing adjustment weights, ++>Representing a preset empirical factor. />
Finally, the upper edge value is used as an abnormal concentration threshold value; and monitoring the real-time concentration data based on the abnormal concentration threshold, and performing early warning when the data value of the real-time concentration data is larger than the abnormal concentration threshold. It should be noted that the preset experience coefficient is 1.5 according to experience.
The embodiment also provides a combustible and toxic and harmful gas real-time monitoring system based on the Internet of things, which comprises a memory, a processor and a computer program, wherein the memory is used for storing the corresponding computer program, the processor is used for running the corresponding computer program, and the computer program can realize the steps of the combustible and toxic and harmful gas real-time monitoring method based on the Internet of things when running on the processor.
In summary, the method firstly acquires the real-time concentration data of the gas to be detected and the historical concentration data in a preset time period, and monitors the real-time concentration data by analyzing the historical concentration data; the data values of all sampling points in the historical concentration data can be initially classified according to the similarity conditions of the data values, so that at least two data categories are obtained; because the abnormal data are individual data and most of the data are normal data, the data which are distributed in a concentrated way are normal data to a great extent, so that the data intensity can be analyzed, and the final abnormal concentration threshold value can be determined; analyzing the distribution density condition of the data values in each data category, and acquiring a data set area; further, similarity and data dispersion analysis are performed on the data values in the data concentration area, so that the relative concentration of the data concentration area is obtained, the maximum data concentration area is further determined, the overall trend of the data is represented by the maximum data concentration area, and therefore, finally, the adjustment weight is obtained according to the distribution difference condition of the data values in the maximum data concentration area and the data values in the historical concentration data, the difference of the relative concentration of the maximum data concentration area and other data concentration areas, and the distribution condition of the data values in all data categories, and the abnormal concentration threshold is obtained by utilizing the adjustment weight and the distribution of the data values in the historical concentration data, so that the real-time concentration data is monitored. In summary, the invention analyzes the distribution aggregation condition of the historical concentration data to obtain the adjustment weight, so that the overall trend of the data can be more accurately mastered, the more accurate abnormal concentration threshold value is obtained to monitor the real-time data, and the accuracy and the reliability of the monitoring result are improved.
It should be noted that: the sequence of the 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 particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (6)
1. The method for monitoring combustible and toxic harmful gases in real time based on the Internet of things is characterized by comprising the following steps of:
acquiring real-time concentration data of gas to be detected and historical concentration data in a preset time period;
obtaining at least two data categories according to the similar conditions of all sampling point data values in the historical concentration data; taking any one data category as a data category to be measured, and acquiring a data concentration area according to the distribution density condition of data values in the data category to be measured;
acquiring a subsequence according to sequential continuous conditions of sampling points corresponding to data values in the data collection area; obtaining a similarity value according to the length difference and the similarity condition of adjacent subsequences; obtaining the data dispersion of the data concentration area according to the difference of the similarity values among the subsequences and the distribution condition of the subsequences on time sequence;
Obtaining the relative concentration of each data concentration area according to the data quantity and the data dispersion contained in the data concentration areas in all the data categories; screening the maximum data-intensive area from all the data-intensive areas according to the relative concentration degree; obtaining adjustment weights according to the distribution difference condition of the data values in the maximum data-intensive area and the data values in the historical concentration data, the difference of the relative densities of the maximum data-intensive area and other data-intensive areas and the distribution condition of the data values in all data categories;
acquiring an abnormal concentration threshold according to the distribution of the data values in the historical concentration data and the adjustment weight, and monitoring the real-time concentration data;
the data dispersion of the data set area is obtained according to the difference of similarity values among the subsequences and the distribution condition of the subsequences on time sequence, and a formula model of the data dispersion is as follows:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>Representing the category +.>Data dispersion of the region in the dataset, +.>Representing the category +.>Is +.>The time value of the last sample point of the sub-sequence is equal to +.>Difference in time values of first sample point of sub-sequence, +. >Representing the category +.>Is +.>The time value of the last sample point of the sub-sequence is equal to +.>Difference in time values of first sample point of sub-sequence, +.>Representing the category +.>Is +.>Subsequence and->Similarity between subsequences, < >>Representing the category +.>Is +.>Subsequence and->Similarity between subsequences, < >>Representing the category +.>The total number of subsequences in the region in the dataset;
the obtaining the relative concentration of each data concentration area according to the data quantity and the data dispersion included in the data concentration areas in all the data categories comprises the following steps:
taking the ratio of the data quantity contained in each data concentration area to the data quantity contained in all the data concentration areas as the data duty ratio of each data concentration area;
the value obtained by carrying out negative correlation mapping on the ratio of the data dispersion of each data set area to the data dispersion accumulation sum of all the data set areas is used as the data concentration of each data set area;
obtaining the relative concentration of each data concentration area according to the data occupation ratio and the data concentration of each data concentration area, wherein the data occupation ratio and the data concentration are positively correlated with the relative concentration;
The obtaining the adjustment weight according to the distribution difference condition of the data value in the maximum data-intensive area and the data value in the historical concentration data, the difference of the relative concentration of the maximum data-intensive area and other data-concentrated areas, and the distribution condition of the data value in all data categories comprises the following steps:
acquiring the lower quartile of all data values in the historical concentration data, and taking the difference value between the average value of all data values in the maximum data-intensive area and the lower quartile as a first important factor of the maximum data-intensive area;
taking the average value of the difference between the relative density of the maximum data-intensive area and the relative density of all other data-intensive areas as a second important factor of the maximum data-intensive area;
obtaining an adjustment weight according to the first important factor and the second important factor of the maximum data-intensive area;
calculating the average value of all the data values in each data category to obtain an average value characteristic value of each data category, and when the data in the maximum data-intensive area belongs to the data category corresponding to the maximum average value characteristic value, positively correlating the first important factor and the second important factor with the adjustment weight; when the data in the maximum data-intensive area does not belong to the data category corresponding to the maximum mean value characteristic value, the first important factor and the second important factor are in negative correlation with the adjustment weight;
The step of monitoring the real-time concentration data by obtaining an abnormal concentration threshold according to the distribution of the data values in the historical concentration data and the adjustment weight comprises the following steps:
acquiring an upper quartile of a data value in the historical concentration data, and acquiring a quartile range according to the upper quartile and the lower quartile corresponding to the historical concentration data;
acquiring an upper edge value in a box diagram of historical concentration data according to the upper quartile, the quartile range, the adjustment weight and the preset experience coefficient, wherein the upper quartile, the quartile range, the adjustment weight and the preset experience coefficient are positively correlated with the upper edge value; taking the upper edge value as the abnormal concentration threshold value;
and monitoring the real-time concentration data based on the abnormal concentration threshold, and performing early warning when the data value of the real-time concentration data is larger than the abnormal concentration threshold.
2. The method for monitoring combustible and toxic harmful gases in real time based on the internet of things according to claim 1, wherein the obtaining at least two data categories according to the similarity of all sampling point data values in the historical concentration data comprises:
Clustering the data values of all sampling points in the historical concentration data based on a K-means clustering algorithm to obtain at least two clusters, wherein the data in each cluster is a data category; wherein the distance measure at the time of clustering is the difference between the data values.
3. The method for monitoring combustible and toxic and harmful gases in real time based on the internet of things according to claim 1, wherein the acquiring the data concentration area according to the distribution density condition of the data values in the data category to be tested comprises:
and carrying out density analysis on the data in the data category to be detected based on a kernel density estimation algorithm, and acquiring a concentrated area of all the data in the data category to be detected as a data concentrated area.
4. The method for monitoring combustible and toxic and harmful gases in real time based on the internet of things according to claim 1, wherein the sub-sequence is obtained according to sequential continuous conditions of sampling points corresponding to data values in the data collection area, and comprises the following steps:
the sampling points of all the data values in the data collection area are sequentially ordered, and the data values of the sampling points with continuous moments are used as a subsequence; if the sampling point is not continuous with other sampling points, the data value of the sampling point is also used as a subsequence.
5. The method for monitoring combustible and toxic harmful gases in real time based on the internet of things according to claim 1, wherein the obtaining the similarity value according to the length difference and the similarity of the adjacent subsequences comprises:
obtaining the similarity value of the two adjacent subsequences according to the length difference and the DTW distance of the two adjacent subsequences; the length difference and the DTW distance are both inversely related to the similarity value.
6. The system for monitoring combustible and toxic harmful gases in real time based on the Internet of things comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, and is characterized in that the steps of the method according to any one of claims 1-5 are realized when the processor executes the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410021934.7A CN117517596B (en) | 2024-01-08 | 2024-01-08 | Method and system for monitoring combustible and toxic harmful gases in real time based on Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410021934.7A CN117517596B (en) | 2024-01-08 | 2024-01-08 | Method and system for monitoring combustible and toxic harmful gases in real time based on Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117517596A CN117517596A (en) | 2024-02-06 |
| CN117517596B true CN117517596B (en) | 2024-03-15 |
Family
ID=89742392
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410021934.7A Active CN117517596B (en) | 2024-01-08 | 2024-01-08 | Method and system for monitoring combustible and toxic harmful gases in real time based on Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117517596B (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004101778A1 (en) * | 2003-03-17 | 2004-11-25 | Affinium Pharmaceuticals, Inc. | Purified hydroxymethylglutaryl-coa reductases and structures thereof |
| WO2018122585A1 (en) * | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| CN114117852A (en) * | 2021-11-18 | 2022-03-01 | 华北电力大学 | Regional heat load rolling prediction method based on finite difference working domain division |
| CN114219157A (en) * | 2021-12-17 | 2022-03-22 | 西南石油大学 | Alkane gas infrared spectrum measurement method based on optimal decision and dynamic analysis |
| CN115034005A (en) * | 2022-05-30 | 2022-09-09 | 四川大学 | Model analysis visualization method for component residual service life prediction |
| WO2023001251A1 (en) * | 2021-07-22 | 2023-01-26 | 梅卡曼德（北京）机器人科技有限公司 | Dynamic picture-based 3d point cloud processing method and apparatus, device and medium |
| CN116701973A (en) * | 2023-08-09 | 2023-09-05 | 无锡市明通动力工业有限公司 | Motor abnormal vibration detection early warning method based on big data |
| CN116861354A (en) * | 2023-07-17 | 2023-10-10 | 国家电网有限公司华北分部 | Abnormality identification and mode distinguishing method and system for monitoring data of power transformer |
| CN116953488A (en) * | 2023-09-19 | 2023-10-27 | 深圳市东陆科技有限公司 | Monitoring method for integrated photoelectric chip |
| CN116975378A (en) * | 2023-09-21 | 2023-10-31 | 江苏云天精筑环境科技有限公司 | Equipment environment monitoring method and system based on big data |
| CN117171604A (en) * | 2023-11-03 | 2023-12-05 | 城资泰诺(山东)新材料科技有限公司 | Sensor-based insulation board production line abnormality monitoring system |
| CN117241306A (en) * | 2023-11-10 | 2023-12-15 | 深圳市银尔达电子有限公司 | Real-time monitoring method for abnormal flow data of 4G network |
| CN117349778A (en) * | 2023-12-04 | 2024-01-05 | 湖南蓝绿光电科技有限公司 | Online real-time monitoring system of consumer based on thing networking |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210073648A1 (en) * | 2019-09-10 | 2021-03-11 | AI Therapeutics, Inc. | Techniques for semi-supervised training and associated applications |
-
2024
- 2024-01-08 CN CN202410021934.7A patent/CN117517596B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004101778A1 (en) * | 2003-03-17 | 2004-11-25 | Affinium Pharmaceuticals, Inc. | Purified hydroxymethylglutaryl-coa reductases and structures thereof |
| WO2018122585A1 (en) * | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| WO2023001251A1 (en) * | 2021-07-22 | 2023-01-26 | 梅卡曼德（北京）机器人科技有限公司 | Dynamic picture-based 3d point cloud processing method and apparatus, device and medium |
| CN114117852A (en) * | 2021-11-18 | 2022-03-01 | 华北电力大学 | Regional heat load rolling prediction method based on finite difference working domain division |
| CN114219157A (en) * | 2021-12-17 | 2022-03-22 | 西南石油大学 | Alkane gas infrared spectrum measurement method based on optimal decision and dynamic analysis |
| CN115034005A (en) * | 2022-05-30 | 2022-09-09 | 四川大学 | Model analysis visualization method for component residual service life prediction |
| CN116861354A (en) * | 2023-07-17 | 2023-10-10 | 国家电网有限公司华北分部 | Abnormality identification and mode distinguishing method and system for monitoring data of power transformer |
| CN116701973A (en) * | 2023-08-09 | 2023-09-05 | 无锡市明通动力工业有限公司 | Motor abnormal vibration detection early warning method based on big data |
| CN116953488A (en) * | 2023-09-19 | 2023-10-27 | 深圳市东陆科技有限公司 | Monitoring method for integrated photoelectric chip |
| CN116975378A (en) * | 2023-09-21 | 2023-10-31 | 江苏云天精筑环境科技有限公司 | Equipment environment monitoring method and system based on big data |
| CN117171604A (en) * | 2023-11-03 | 2023-12-05 | 城资泰诺(山东)新材料科技有限公司 | Sensor-based insulation board production line abnormality monitoring system |
| CN117241306A (en) * | 2023-11-10 | 2023-12-15 | 深圳市银尔达电子有限公司 | Real-time monitoring method for abnormal flow data of 4G network |
| CN117349778A (en) * | 2023-12-04 | 2024-01-05 | 湖南蓝绿光电科技有限公司 | Online real-time monitoring system of consumer based on thing networking |
Non-Patent Citations (3)
| Title |
|---|
| 以VIIRS-DNB数据为基础的黄海渔船夜间灯光动态;陈冠宇;刘阳;田浩;叶振江;于海庆;李建超;田永军;;水产学报(第06期);全文 * |
| 基于优化的GA-BP网络算法的混合气体识别;杜鸿飞;谢光忠;;电子元件与材料(第12期);全文 * |
| 基于客户诉求的智能配电网建设需求研究;余学虎;石钟文;王慧;张兴松;甘莉;申健;;中国信息化;20200320(第03期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117517596A (en) | 2024-02-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101882925B1 (en) | Method for inspecting the quality of a solder joint | |
| CN111931868A (en) | Time series data abnormity detection method and device | |
| CN112284440B (en) | Sensor data deviation self-adaptive correction method | |
| US20080025591A1 (en) | Method and system for robust classification strategy for cancer detection from mass spectrometry data | |
| CN115982602B (en) | Photovoltaic transformer electrical fault detection method | |
| CN117196353B (en) | Environmental pollution assessment and monitoring method and system based on big data | |
| CN116559598B (en) | Smart distribution network fault positioning method and system | |
| CN116881745B (en) | Pressure transmitter abnormality monitoring method based on big data | |
| CN111338972A (en) | Machine learning-based software defect and complexity incidence relation analysis method | |
| US20220027230A1 (en) | Predicting Equipment Fail Mode from Process Trace | |
| CN117171604B (en) | Sensor-based insulation board production line abnormality monitoring system | |
| US20210374634A1 (en) | Work efficiency evaluation method, work efficiency evaluation apparatus, and program | |
| CN111832974A (en) | Vehicle fault early warning method and storage medium | |
| CN115409066A (en) | Method and device for detecting abnormality of time series data, and computer storage medium | |
| CN117556714B (en) | Preheating pipeline temperature data anomaly analysis method for aluminum metal smelting | |
| CN117517596B (en) | Method and system for monitoring combustible and toxic harmful gases in real time based on Internet of things | |
| CN112949735A (en) | Liquid hazardous chemical substance volatile concentration abnormity discovery method based on outlier data mining | |
| O'Reilly et al. | Revisiting the ROC curve for diagnostic applications with an unbalanced class distribution | |
| CN112199295A (en) | Deep neural network defect positioning method and system based on frequency spectrum | |
| CN116739619A (en) | Energy power carbon emission monitoring analysis modeling method and device | |
| CN111339986A (en) | Frequency law mining method and system for equipment based on time domain/frequency domain analysis | |
| CN114048808A (en) | Gross error identification and elimination method, system, server and computer readable storage medium for monitoring data | |
| CN107506824B (en) | Method and device for detecting bad observation data of power distribution network | |
| Singgih et al. | Comparison of M Estimation, S Estimation, with MM Estimation to Get the Best Estimation of Robust Regression in Criminal Cases in Indonesia: Perbandingan Estimasi M, Estimasi S, dengan Estimasi MM | |
| CN117893100B (en) | Construction method of quality evaluation data updating model based on convolutional neural network |
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 |