CN117972314A - Cloud platform monitoring method and system based on digital twinning - Google Patents
Cloud platform monitoring method and system based on digital twinning Download PDFInfo
- Publication number
- CN117972314A CN117972314A CN202410369936.5A CN202410369936A CN117972314A CN 117972314 A CN117972314 A CN 117972314A CN 202410369936 A CN202410369936 A CN 202410369936A CN 117972314 A CN117972314 A CN 117972314A
- Authority
- CN
- China
- Prior art keywords
- flow data
- data point
- points
- flow
- data
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000012544 monitoring process Methods 0.000 title claims abstract description 25
- 239000002245 particle Substances 0.000 claims abstract description 46
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 28
- 238000001914 filtration Methods 0.000 claims abstract description 19
- 238000012216 screening Methods 0.000 claims abstract description 9
- 230000008859 change Effects 0.000 claims description 23
- 230000008569 process Effects 0.000 claims description 23
- 230000006870 function Effects 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 2
- 230000009286 beneficial effect Effects 0.000 abstract 1
- 238000004519 manufacturing process Methods 0.000 description 7
- 239000012535 impurity Substances 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000001595 flow curve Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012800 visualization 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- 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
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to a cloud platform monitoring method and system based on digital twinning, wherein the method comprises the following steps: collecting flow data and pressure data of the water supply equipment, acquiring a plurality of extreme points of the local range of each flow data point, obtaining noise degree of each flow data point according to time difference and numerical value difference of every two adjacent extreme points, correcting the noise degree according to the pressure data, screening out target flow data points according to the corrected noise degree, initializing particles of the target flow data points, denoising the flow data through particle filtering, establishing a digital twin platform according to the denoised flow data to monitor the water supply equipment, improving the accuracy of particle filtering denoising, and being beneficial to establishing an accurate digital twin platform to accurately monitor the flow data of the water supply equipment.
Description
Technical Field
The present invention relates generally to the field of data processing technology. More particularly, the invention relates to a cloud platform monitoring method and system based on digital twinning.
Background
The digital twin system is used for copying the behavior of an engineering physical object simulation factory in a real environment in a digital mode and carrying out virtual simulation on the design, process and manufacture of a product or even the whole factory, so that the research and development and manufacture production efficiency of the product are improved, the possibility of error is prejudged in advance, and the purposes of saving the production cost and reducing the production loss are realized.
In the digital twin system, data collected by a sensor are required to be transmitted to a cloud platform in real time, so that the cloud platform can monitor a production line in real time, collected flow data are often required to be monitored in real time in the production test process of water supply equipment, so that the quality of the water supply equipment is ensured, but due to the fact that noise data exist in the collected flow data, the noise data in the collected flow data are removed by adopting a particle filtering algorithm in the prior art, and a digital twin model of the equipment is built by utilizing the denoised data.
Because the change of water quality in the operation process of the water supply equipment can influence the measurement result of the flow sensor to generate noise, especially when the water contains impurities or bubbles, fluctuation or instability of flow data can be caused, which means that the intensity of the noise generated in different time periods can be different, so that the condition of sampling deviation is easy to generate when the existing particle filtering algorithm adopts uniform particle quantity and random particles to filter and denoise, namely, the particle density of certain areas is lower and cannot better reflect real data distribution, or noise data points are directly used as particles to be initialized, the accuracy of establishing a digital twin model of the equipment is influenced, and errors exist in monitoring the flow data.
Disclosure of Invention
In order to solve one or more of the technical problems, the invention provides a cloud platform monitoring method and a system based on digital twinning, the method improves the accuracy of establishing a digital twinning model of equipment, and the system can accurately monitor flow data. The technical scheme is as follows: a cloud platform monitoring method based on digital twinning comprises the following steps:
collecting flow data and pressure data corresponding to the flow data when the water supply equipment operates; the flow data includes a plurality of flow data points, the pressure data includes a plurality of pressure data points;
presetting a local range of each flow data point, acquiring a plurality of extreme points of the local range, and obtaining the noise degree of each flow data point according to the time difference and the numerical difference of every two adjacent extreme points in the plurality of extreme points;
correcting the noise degree of each flow data point according to each pressure data point to obtain the corrected noise degree of each flow data point;
screening out target flow data points by utilizing the noise degrees corrected by all the flow data points, and initializing particles of the target flow data points;
Carrying out particle filtering denoising on the flow data by utilizing the target flow data point after particle initialization;
and establishing a digital twin platform according to the flow data after particle filtering denoising to monitor the water supply equipment.
Further, according to the time difference and the numerical difference of each two adjacent extremum points in the plurality of extremum points, obtaining the noise degree of each flow data point includes:
Dividing each two adjacent extreme points into a group to obtain a plurality of groups of extreme points in the local range;
Obtaining the data fluctuation degree and the data regularity in the local range of each flow data point according to the time difference and the numerical difference of two extreme points in each group of extreme points;
Based on the data fluctuation degree and regularity, calculating the noise degree of each flow data point, wherein the noise degree satisfies the following relation:
Where i is the sequence number of the traffic data point, For the noise level of the ith flow data point,/>For unifying quantized weights,/>As an exponential function based on a natural constant e,/>For the degree of data fluctuation in the local range of the ith flow data point,/>Is the data regularity in the local range of the ith traffic data point.
Further, the degree of fluctuation of the data in the local range of each flow data point includes:
Taking the reciprocal of the time difference between two extreme points in each group of extreme points as the frequency of each group of extreme points in the local range, and taking the average value of the frequencies of all groups of extreme points in the local range as the data change frequency in the local range;
Taking the numerical value difference of two extreme points in each group of extreme points as the amplitude difference of each group of extreme points in the local range, and taking the average value of the amplitude differences of all groups of extreme points in the local range as the data change amplitude in the local range;
the degree of data fluctuation in the local range is: and the sum of the data change frequency and the data change amplitude in the local range.
Further, the data regularity in the local range of each flow data point includes:
taking the reciprocal of the variance of the frequencies of all groups of extreme points in the local range as the consistency of the data frequency variation in the local range;
taking the reciprocal of the variance of the amplitude differences of all groups of extreme points in the local range as the consistency of the data amplitude variation in the local range;
The data regularity in the local range of each flow data point is: the consistency of the data frequency variation and the consistency of the data amplitude variation within the local range.
Further, correcting the noise degree of each flow data point according to each pressure data point to obtain the corrected noise degree of each flow data point, including:
First, a standard flow data point corresponding to each pressure data point is screened, including:
Obtaining a plurality of flow data points corresponding to each pressure data point, including: acquiring a plurality of flow data points corresponding to all the pressure data points with the same pressure data value as a plurality of flow data points corresponding to the pressure data point;
sequencing the collected pressure data points from small to large according to the values, and constructing a coordinate system by taking the values of the pressure data points as the abscissa and the values of a plurality of flow data points corresponding to each pressure data point as the ordinate;
Clustering a plurality of flow data points corresponding to each pressure data point according to the numerical value, and calculating a standard flow numerical value corresponding to each pressure data point according to a clustering result;
Then, fitting a functional relationship between the flow data point and the pressure data point, comprising:
Curve fitting is carried out on the standard flow values corresponding to all the pressure data points by using a least square method, so that the functional relation between the values of the pressure data points and the corresponding standard flow values is obtained Wherein the independent variable/>Is the value of the pressure data point,/>The standard flow value corresponding to the pressure data point;
finally, correcting the noise level of each flow data point based on the functional relation, including:
in the method, in the process of the invention, For the noise level of the ith flow data point,/>For the noise level corrected for the ith flow data point,/>Is the value of the ith flow data point,/>Is the value of the pressure data point corresponding to the ith flow data point,/>Meaning that the pressure data point corresponding to the ith flow data point is calculated according to the functional relation, and the standard flow value corresponding to the ith flow data point is calculated according to the functional relation,/>Is an exponential function based on a natural constant e.
Further, clustering the plurality of flow data points corresponding to each pressure data point according to the numerical value, and calculating the standard flow numerical value corresponding to each pressure data point according to the clustering result, wherein the method comprises the following steps:
Classifying a plurality of flow data points corresponding to each pressure data point in a coordinate system by using a DBSCAN clustering algorithm, setting the clustering radius to be 0.3, and obtaining a plurality of class clusters;
And taking one class cluster with the largest flow data points as a real class cluster, calculating the numerical average value of all the flow data points in the real class cluster, and taking the average value as a standard flow value corresponding to the pressure data.
Further, screening out the target flow data point by using the noise degree corrected by all the flow data points, including:
Firstly, sequencing all flow data points from small to large according to noise degree;
then, the data point of the flow of the first M% is taken as a flow data set, the iteration step length is 1% as the beginning of the iteration, and the ending condition of the iteration is that the data point of the flow of the first M% is taken % Flow data points, where M% and/>% Is the percentage of the flow data points, and is a preset value;
in each iteration process, calculating the preference of the flow data set participating in the iteration this time:
in the method, in the process of the invention, For iteration number,/>Preference of flow data set for mth participating iteration,/>For the number of flow data points contained in the flow data set of the mth participating iteration,/>For the sequence number of the traffic data point in the traffic data set,/>Get pass/>All integers within the range,/>For the/>, in the traffic datasetNoise level after correction of each flow data point,/>Data completeness of the flow data set for the mth participation iteration;
And finally, acquiring a flow data set with the maximum preference, and taking the flow data point contained in the flow data set as a target flow data point.
Further, the data completeness of the flow data set participating in iteration includes:
classifying the flow data points in the flow data sets participating in iteration according to the numerical value to obtain the flow data types contained in the flow data sets participating in iteration;
classifying all flow data points according to the numerical value to obtain the types of flow data contained in all flow data;
the data completeness of the flow data set participating in the iteration is: the ratio of the number of traffic data categories contained in the traffic data set participating in the iteration to the number of traffic data categories contained in all traffic data.
Further, the presetting the local range of each flow data point includes:
setting a certain flow data point as a center, selecting N flow data points on the left side and N flow data points on the right side of the flow data point, adding the flow data points, and taking the total 2N+1 flow data points as the local range of the flow data points.
The invention also provides a cloud platform monitoring system based on digital twinning, which comprises a memory and a processor, wherein the memory is stored with a computer program, and the processor executes the computer program to realize the steps of the monitoring method.
The invention has the following effects:
According to the invention, the noise degree of each flow data point is obtained through the data change characteristics of the local range of each flow data point in the operation process of the water supply equipment, the flow data point is screened according to the noise degree, a flow data set with small noise degree and good data completeness is screened out, the flow data point contained in the flow data set is used as initial particles for particle initialization, the self-adaptive particle initialization process in a particle filtering algorithm is realized, the characteristics of the collected flow data can be accurately reflected by the flow data point with small noise degree and good data completeness, the particle initialization is carried out on the flow data, the accuracy of particle filtering denoising of the flow data can be improved, and further the accurate monitoring of the flow data of the water supply equipment by a digital twin cloud platform is facilitated.
Drawings
The above, as well as additional purposes, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description when read in conjunction with the accompanying drawings. In the drawings, embodiments of the invention are illustrated by way of example and not by way of limitation, and like reference numerals refer to similar or corresponding parts and in which:
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.
Referring to fig. 1, a cloud platform monitoring method based on digital twinning includes steps S1-S5, specifically as follows:
S1: and collecting flow data and pressure data corresponding to the flow data when the water supply equipment operates.
The flow sensor and the pressure sensor are arranged on a pipeline of the water supply equipment to acquire data, the specific acquisition frequency is 10HZ, the acquisition time length is 5 minutes, the flow data and the pressure data in the operation process of the water supply equipment are obtained, the acquisition frequency and the acquisition time length can be adjusted by themselves and are not limited, the flow data comprise a plurality of flow data points, the pressure data comprise a plurality of pressure data points, no matter the pressure data points or the flow data points, each data point corresponds to a value, the value corresponding to each pressure data point is a pressure value, and the value corresponding to each flow data point is a flow value.
S2: presetting a local range of each flow data point, acquiring a plurality of extreme points of the local range, and obtaining the noise degree of each flow data point according to the time difference and the numerical difference of every two adjacent extreme points in the plurality of extreme points.
Because the traditional particle filtering algorithm adopts a fixed number of particles when initializing the particles, and the particles are selected to be initialized according to a random selection mode, noise points in flow data in the operation process of the water supply equipment can be contained in the selection process, so that the filtering result is influenced, therefore, the noise degree of each flow data point is obtained by analyzing the data change in the local range of the flow data point, the more the noise degree is higher, the more likely the flow data point is the noise data point, the collected flow data points are screened according to the noise degree of the flow data points, so that the reliability and the authenticity of the data obtained by initializing the particles in the particle filtering algorithm are higher, and the filtering denoising speed and the filtering accuracy of the algorithm can be improved.
Wherein setting the local range for each flow data point comprises: the local range of one flow data point is set to be the local range of the current flow data point, N flow data points on the left side and N flow data points on the right side of the current flow data point are selected, and the current flow data point is added with the current flow data point, and the total of 2N+1 flow data points are taken as the local range of the current flow data point, wherein the value of N is set to be 7 in the embodiment, namely 15 flow data points are selected in the local range of the current flow data point, namely the current flow data point, 7 flow data points on the left side of the current flow data point and 7 flow data points on the right side of the current flow data point are respectively selected. It should be noted that, in this embodiment, since the flow data is collected in a period of time, when the local range of each flow data point is divided, the division is performed based on the time sequence of the flow data points, and 7 flow data points are sequentially selected on the left side and the right side of the current flow data point according to the time sequence.
The obtaining the noise degree of each flow data point according to the time difference and the numerical difference of every two adjacent extremum points in the plurality of extremum points comprises the following steps:
s21: acquiring a plurality of extreme points in the local range of each flow data point, and grouping;
In the embodiment, a difference method is used to obtain a plurality of extreme points in a local range of each flow data point, and the obtaining method is not limited and can also be used for obtaining in other modes;
The method for grouping the extreme points comprises the following steps: every two adjacent extreme points are divided into a group;
In a specific example, for the ith flow data point in all flow data points, the extremum points in the local range are i1, i2, i3, i4, i5, i6, and total six extremum points, and the time corresponding to each extremum point is t1, t2, t3, t4, t5, t6, and five groups of extremum points can be obtained in total by taking every two adjacent extremum points as a group of extremum points, the first group is i1 and i2, the second group is i2 and i3, the third group is i3 and i4, the fourth group is i4 and i5, and the fifth group is i5 and i6.
S22: according to the time difference and the numerical difference of two extreme points in each group of extreme points, the data fluctuation degree and the data regularity in the local range of each flow data point are obtained, and the method is concretely as follows:
S221: acquiring the data fluctuation degree in the local range of each flow data point, wherein the data fluctuation degree comprises the following steps:
Firstly, acquiring data change frequency in the local range, namely taking the reciprocal of the time difference of two extreme points in each group of extreme points as the frequency of each group of extreme points in the local range, and taking the average value of the frequencies of all groups of extreme points in the local range as the data change frequency in the local range;
specifically, the frequency of the jth set of extreme points occurring in the local range of the ith flow data point is formulated as:
in the formula (i), Representing the frequency of occurrence of the jth set of extreme points in the local range of the ith flow data point,/>Time representing the first extreme point of the jth set of extreme points in the local range of the ith flow data point,/>Representing the time of the second extreme point in the j-th set of extreme points in the local range of the i-th flow data point, and representing the occurrence frequency of the extreme points by using the inverse of the time difference of the two extreme points;
in a specific example, taking a second set of extreme points within the local range of the ith flow data point as an example, the second set of extreme points occur at a frequency within the local range of: ；
and then, acquiring the data change amplitude in the local range, namely taking the numerical value difference of two extreme points in each group of extreme points as the amplitude difference of each group of extreme points in the local range, and taking the average value of the amplitude differences of all groups of extreme points in the local range as the data change amplitude in the local range.
Specifically, e represents the value of the extremum point, and the amplitude difference of the jth extremum point in the local range of the ith flow data point is expressed as:
in the formula (i), Representing the amplitude difference of the jth set of extreme points in the local range of the ith flow data point,/>Numerical value representing the first extreme point of the j-th set of extreme points in the local range of the i-th flow data point,/>The value representing the second extreme point of the j-th set of extreme points in the local range of the i-th flow data point represents the magnitude difference of the set of extreme points using the difference between the values of the two extreme points comprised by the set of extreme points.
In one specific example: taking the second set of extreme points as an example, the values of the two extreme points are e2 and e3, and the amplitude difference of the second set of extreme points is:；
Finally, the degree of data fluctuation in this local range is: the sum of the frequency of data change and the amplitude of data change in the local range is expressed as:
in the formula (i), For the degree of data fluctuation in the local range of the ith flow data point,/>Representing the number of extreme points in the local range,/>Representing that extreme points within the local range are totally divided into/>Group j represents the extreme point of the j group, and the value range of j is [/>The meaning of the method is that all groups of extreme points are traversed, the average frequency of all groups of extreme points in the local range represents the data change frequency (or fluctuation frequency) in the local range, the average amplitude difference of all groups of extreme points in the local range represents the data change amplitude (or fluctuation amplitude) in the local range, and the data fluctuation degree in the local range is obtained by combining the two.
It should be noted that, for each flow data point, the fluctuation degree of the flow data in the local range can be reflected by the change frequency and the change amplitude of the flow data point in the local range, that is, the magnitude of the up-down fluctuation of the value of the flow data point in the local range, and for one flow data point, if the value of the flow data point in the local range is faster and the magnitude of the up-down fluctuation is greater, the data in the local range of the flow data point has a high-frequency fluctuation characteristic, and the more likely the flow data point is a noise point, that is, the higher the noise degree of the flow data point is.
S222: acquiring data regularity in a local range of each flow data point, wherein the data regularity comprises;
taking the reciprocal of the variance of the frequencies of all groups of extreme points in the local range as the consistency of the data frequency variation in the local range;
taking the reciprocal of the variance of the amplitude differences of all groups of extreme points in the local range as the consistency of the data amplitude variation in the local range;
the data regularity in the local range of each flow data point is: the consistency of the data frequency variation and the sum of the consistency of the data amplitude variation in the local range;
specifically, the formula is as follows:
in the formula (i), Representing data regularity in a local range of an ith traffic data point,/>Representing the variance of the frequencies of all sets of extreme points within the local range of the ith flow data point, then/>Indicating the consistency of the data frequency variation in a local range, and the same applies/>Representing the variance of the amplitude differences of all sets of extreme points within the local range of the ith flow data pointThe consistency of the data amplitude variation in the local range is expressed, and the sum of the consistency and the data amplitude variation is the data regularity in the local range.
It should be noted that, for each flow data point, the regularity of the flow data in the local range may be reflected by the consistency of the frequency variation and the consistency of the amplitude variation of the flow data points in the local range, if the frequencies of each group of extreme points in the local range are more similar in the local range, the better the consistency of the frequency variation of the flow data points in the local range is, otherwise, the worse the consistency of the frequency variation is, specifically, the consistency of the frequency variation is represented by the variance of the frequencies of all groups of extreme points in the local range; if the amplitude differences of all the groups of extreme points are closer, the variation amplitude of the flow data points in the local range is always better, otherwise, the variation amplitude is worse, and specifically, the consistency of the amplitude variation is represented by the variance of the amplitude differences of all the groups of extreme points in the local range; further, if the consistency of the frequency variation and the consistency of the amplitude variation of the flow data points in the local range are better, the flow data in the local range are more regular, and conversely, the flow data in the local range are less regular.
S23: based on the data fluctuation degree and regularity, calculating the noise degree of each flow data point, wherein the noise degree satisfies the following relation:
in the formula (i), Representing the noise level of the ith flow data point,/>The weight representing the unification quantization prevents the data fluctuation degree or the data regularity in the local range of the ith flow data point from being excessively small, and the influence of the data fluctuation degree or the data regularity in the local range on the noise degree is difficult to be reflected because the value is excessively small, which is set in the embodimentThe value is an empirical value, and can be set by oneself, and the weight/>Is a number from 0 to 1,/>Meaning an exponential function based on a natural constant e.
It should be noted that, because the water quality change affects the measurement result of the flow sensor in the operation process of the water supply device, especially when the water contains impurities or bubbles, the flow data will be caused to fluctuate, and because the noise is the data fluctuation caused by the random factor, the flow data presents a certain irregular change, if the fluctuation degree of the flow data is larger and the random irregular feature is presented, the flow data is more likely to be noise data, therefore, the fluctuation degree of the flow data and the regularity of the flow data can be analyzed, and the noise degree of the flow data can be obtained.
It should be noted that, if the number of extreme points in the local range of a certain flow data point is less than or equal to 2, no obvious data fluctuation occurs in the local range of the flow data point, the flow data point is a normal flow data point, and the noise level is 0.
S3: and correcting the noise degree of each flow data point according to each pressure data point to obtain the corrected noise degree of each flow data point.
S31: screening the standard flow data point corresponding to each pressure data point, including:
All the collected pressure data points are ordered from small to large according to the numerical value, and then the pressure data points are taken as the abscissa, and a plurality of flow data corresponding to each pressure data point are taken as the ordinate, wherein the concrete method for acquiring a plurality of flow data corresponding to each pressure data point is as follows: acquiring a plurality of flow data points corresponding to all the pressure data points with the same value as the pressure data point as a plurality of flow data points corresponding to the pressure data point, so that one abscissa value corresponds to a plurality of ordinate values in a coordinate system, and the value of the same pressure data point corresponds to the value of a plurality of flow data points;
At each abscissa of the coordinate system, clustering a plurality of ordinate corresponding to each abscissa by using DBSCAN (distributed base station) clustering, namely classifying a plurality of flow data points corresponding to each pressure data point, setting a clustering radius to be 0.3, and obtaining a plurality of clusters, wherein the purpose of the clustering is to eliminate interference of noise data in the flow data, find a standard flow data value corresponding to each pressure data, and the clustering is performed based on the numerical values of the plurality of flow data points.
For the clustering result, the standard flow value corresponding to each pressure data is located in the cluster, the flow data points are more, and the flow data points are loitered around a certain fixed value, so that one cluster with the largest flow data point in the clustering result is taken as a real cluster, the average of the values of all the flow data points in the real cluster is calculated, and the average is taken as the standard flow value corresponding to the pressure data, and the standard flow value corresponding to each pressure data can be obtained according to the method.
S32: curve fitting is carried out on the standard flow values corresponding to all the pressure data points by utilizing a least square method to obtain a pressure-flow curve graph, and a functional relation between the values of the pressure data points and the corresponding standard flow values is obtainedWherein/>Representing the standard flow value corresponding to the pressure data point,/>A value representing a pressure data point;
S33: correcting the noise level of each flow data point based on the functional relationship, including:
in the method, in the process of the invention, For the noise level of the ith flow data point,/>For the noise level corrected for the ith flow data point,/>Is the value of the ith flow data point,/>Is the value of the pressure data point corresponding to the ith flow data point,/>Meaning that the pressure data point corresponding to the ith flow data point is calculated according to the functional relation, and the standard flow value corresponding to the pressure data point is calculated by using the methodAs an exponential function based on a natural constant e, then/>Representing the numerical deviation between the ith flow data point and the standard flow value,/>Is/>For any one flow data point, if the deviation between the flow data point and the standard flow value is larger, the flow data point is not in a proportional relation with the corresponding pressure data point, and the noise degree of the flow data point is higher if the deviation between the flow data point and the corresponding pressure data point is larger than the difference between the flow data point and the functional relation curve, and it is required to be noted that the pressure data point corresponding to each flow data point refers to the pressure data point corresponding to the moment when the flow data point is collected.
It should be noted that, because the flow data and the pressure data may show a certain proportional relationship or a functional relationship in the operation process of the water supply device, the noise degree of the flow data point can be corrected by using the function, so as to improve the accuracy of noise identification, and the noise degree after correction of all the flow data points can be obtained according to the method.
S4: and screening out target flow data points by utilizing the noise degrees corrected by all the flow data points, and initializing particles of the target flow data points.
S41: firstly, sorting all flow data points according to noise degree from small to large, then taking the flow data points of the first M% to obtain a flow data set, wherein the value of the preset M is 5 as the beginning of iteration, the iteration step length is 1% (namely, the flow data points of the first 6% of the next iteration are taken as the flow data set) and the ending condition of the iteration is that the flow data set is taken beforeAs the flow data set, preset/>, in this embodimentThe value of (2) is 30, and can be specifically set by self;
s42: then, in each iteration process, calculating the preference of the flow data set participating in the iteration at this time:
in the method, in the process of the invention, For iteration number,/>Preference of flow data set for mth participating iteration,/>For the number of flow data points contained in the flow data set of the mth participating iteration,/>For the sequence number of the traffic data point in the traffic data set,/>Get pass/>All integers within the range,/>For the/>, in the traffic datasetNoise level after correction of each flow data point,/>Data completeness for the flow data set for the mth participating iteration.
It should be noted that, for the corresponding flow data set in each iteration, the noise level may be represented by the average value of the noise levels of all flow data points in the flow data set, the greater the average value is, the greater the noise level of the flow data set is, the completeness of the data may be represented by the ratio of the number of types of flow data in the flow data set participating in the iteration to the number of types of flow data in all flow data, the greater the ratio is, the greater the variety of flow data contained in the flow data set in the iteration process is, the better the completeness is, the more the feature of the flow data can be reflected, the more suitable the particle initialization is performed, the lower the noise level of one flow data set is, the better the data completeness is, the higher the preference is, the more should be the initialized particle, and conversely, the lower the preference is the less can not be the initialized particle.
The method for acquiring the data completeness of the flow data set participating in iteration comprises the following steps:
Classifying the flow data points in the flow data sets participating in iteration according to the numerical value to obtain the flow data types contained in the flow data sets participating in iteration, classifying according to the numerical value, specifically, taking the flow data points with the same numerical value as a class, and taking the flow data points as a class if no other flow data points with the same numerical value exist in a certain flow data point;
classifying all flow data points according to the numerical value to obtain the types of flow data contained in all flow data;
the data completeness of the flow data set participating in the iteration is: the ratio of the number of traffic data categories contained in the traffic data set participating in the iteration to the number of traffic data categories contained in all traffic data.
Expressed by the formula:
in the formula (i), For the number of categories of flow data points contained in the flow data set participating in the mth iteration,/>The number of categories of the flow data points included in all the flow data points.
In one specific example: if the flow data set involved in the iteration contains 5 flow data points, the values of the 5 flow data points are in turn: 1,1,2,3,2, the first flow data point and the second flow data point are in a class, the third flow data point and the fifth flow data point are in a class, the fourth flow data point is in a class alone, the number of classes of the flow data points contained in the flow data set participating in iteration is 3, and the method for classifying all the flow data points according to the numerical values is the same as the method.
S43: and finally, taking the flow data point contained in the flow data set with the maximum preference as a target flow data point, and initializing the particles by using the target flow data point.
It should be noted that, in particle initialization (i.e., selecting some flow data points to perform particle initialization), not only accuracy of the flow data points but also completeness of the flow data are considered, i.e., the more and better the flow data can represent, i.e., the more and better the types of the flow data points are, so that in the screening process, the noise degree of the data points and the completeness of the data points can be considered simultaneously to calculate the screening preference, and a flow data set with small noise degree and good completeness of the data points is screened, and the flow data points included in the flow data set are used as initial particles to perform particle initialization.
S5: and carrying out particle filtering denoising on the flow data by utilizing the target flow data points after particle initialization, and establishing a digital twin platform according to the flow data after particle filtering denoising to monitor the water supply equipment.
According to the embodiment, particle filter algorithm is improved, particle initialization is particularly improved, initialized particles are screened, flow data points which can accurately reflect data characteristics are screened out and serve as particles to be initialized, filtering denoising is conducted on flow data in the operation process of water supply equipment based on the improved particle filter algorithm, accurate flow data are obtained after denoising, a digital twin platform is built according to the accurate flow data, the operation condition of the water supply equipment is monitored in the digital twin platform, the existing abnormality can be intuitively and accurately represented through a 3D visualization means, hidden danger can be conveniently detected by staff, and accurate and efficient monitoring of the water supply equipment is achieved through the digital twin platform.
The invention also provides a cloud platform monitoring system based on digital twinning, which comprises a memory and a processor, wherein the memory is stored with a computer program, the processor executes the computer program to realize any step of S1-S5, filtering and denoising of flow data in the running process of the water supply equipment are realized, accurate flow data are obtained after denoising, a digital twinning platform is constructed according to the accurate flow data, and the running condition of the water supply equipment is monitored in the digital twinning platform.
In the description of the present specification, the meaning of "a plurality", "a number" or "a plurality" is at least two, for example, two, three or more, etc., unless explicitly defined otherwise.
While various embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Many modifications, changes, and substitutions will now occur to those skilled in the art without departing from the spirit and scope of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention.
Claims (10)
1. The cloud platform monitoring method based on digital twinning is characterized by comprising the following steps of:
collecting flow data and pressure data corresponding to the flow data when the water supply equipment operates; the flow data includes a plurality of flow data points, the pressure data includes a plurality of pressure data points;
presetting a local range of each flow data point, acquiring a plurality of extreme points of the local range, and obtaining the noise degree of each flow data point according to the time difference and the numerical difference of every two adjacent extreme points in the plurality of extreme points;
correcting the noise degree of each flow data point according to each pressure data point to obtain the corrected noise degree of each flow data point;
screening out target flow data points by utilizing the noise degrees corrected by all the flow data points, and initializing particles of the target flow data points;
Carrying out particle filtering denoising on the flow data by utilizing the target flow data point after particle initialization;
and establishing a digital twin platform according to the flow data after particle filtering denoising to monitor the water supply equipment.
2. The method for monitoring a cloud platform based on digital twinning according to claim 1, wherein obtaining the noise level of each flow data point according to the time difference and the numerical difference of each two adjacent extreme points in the plurality of extreme points comprises:
Dividing each two adjacent extreme points into a group to obtain a plurality of groups of extreme points in the local range;
Obtaining the data fluctuation degree and the data regularity in the local range of each flow data point according to the time difference and the numerical difference of two extreme points in each group of extreme points;
Based on the data fluctuation degree and regularity, calculating the noise degree of each flow data point, wherein the noise degree satisfies the following relation:
；
Where i is the sequence number of the traffic data point, For the noise level of the ith flow data point,/>Is weight,/>As an exponential function based on a natural constant e,/>For the degree of data fluctuation in the local range of the ith flow data point,/>Is the data regularity in the local range of the ith traffic data point.
3. The cloud platform monitoring method based on digital twinning of claim 2, wherein the degree of data fluctuation in the local range of each flow data point comprises:
Taking the reciprocal of the time difference between two extreme points in each group of extreme points as the frequency of each group of extreme points in the local range, and taking the average value of the frequencies of all groups of extreme points in the local range as the data change frequency in the local range;
Taking the numerical value difference of two extreme points in each group of extreme points as the amplitude difference of each group of extreme points in the local range, and taking the average value of the amplitude differences of all groups of extreme points in the local range as the data change amplitude in the local range;
the degree of data fluctuation in the local range is: and the sum of the data change frequency and the data change amplitude in the local range.
4. The cloud platform monitoring method based on digital twinning of claim 2, wherein the data regularity in the local range of each traffic data point comprises:
taking the reciprocal of the variance of the frequencies of all groups of extreme points in the local range as the consistency of the data frequency variation in the local range;
taking the reciprocal of the variance of the amplitude differences of all groups of extreme points in the local range as the consistency of the data amplitude variation in the local range;
The data regularity in the local range of each flow data point is: the consistency of the data frequency variation and the consistency of the data amplitude variation within the local range.
5. The digital twinning-based cloud platform monitoring method of claim 2, wherein correcting the noise level of each flow data point according to each pressure data point to obtain the corrected noise level of each flow data point comprises:
First, a standard flow data point corresponding to each pressure data point is screened, including:
Obtaining a plurality of flow data points corresponding to each pressure data point, including: acquiring a plurality of flow data points corresponding to all the pressure data points with the same pressure data value as a plurality of flow data points corresponding to the pressure data point;
sequencing the collected pressure data points from small to large according to the values, and constructing a coordinate system by taking the values of the pressure data points as the abscissa and the values of a plurality of flow data points corresponding to each pressure data point as the ordinate;
Clustering a plurality of flow data points corresponding to each pressure data point according to the numerical value, and calculating a standard flow numerical value corresponding to each pressure data point according to a clustering result;
Then, fitting a functional relationship between the flow data point and the pressure data point, comprising:
Curve fitting is carried out on the standard flow values corresponding to all the pressure data points by using a least square method, so that the functional relation between the values of the pressure data points and the corresponding standard flow values is obtained Wherein the independent variable/>Is the value of the pressure data point,/>The standard flow value corresponding to the pressure data point;
finally, correcting the noise level of each flow data point based on the functional relation, including:
；
in the method, in the process of the invention, For the noise level of the ith flow data point,/>For the noise level corrected for the ith flow data point,/>Is the value of the ith flow data point,/>Is the value of the pressure data point corresponding to the ith flow data point,/>Meaning that the standard flow value of the pressure data point corresponding to the ith flow data point calculated according to the functional relation,/>Is an exponential function based on a natural constant e.
6. The cloud platform monitoring method based on digital twinning of claim 5, wherein clustering the plurality of flow data points corresponding to each pressure data point according to a numerical value, and calculating a standard flow value corresponding to each pressure data point according to a clustering result comprises:
Classifying a plurality of flow data points corresponding to each pressure data point in a coordinate system by using a DBSCAN clustering algorithm, setting the clustering radius to be 0.3, and obtaining a plurality of class clusters;
And taking one class cluster with the largest flow data points as a real class cluster, calculating the numerical average value of all the flow data points in the real class cluster, and taking the average value as a standard flow value corresponding to the pressure data.
7. The method for monitoring a cloud platform based on digital twinning according to claim 5, wherein the step of screening out the target flow data point by using the noise level corrected by all the flow data points comprises the steps of:
Firstly, sequencing all flow data points from small to large according to noise degree;
then, the data point of the flow of the first M% is taken as a flow data set, the iteration step length is 1% as the beginning of the iteration, and the ending condition of the iteration is that the data point of the flow of the first M% is taken Wherein M% and/>The percentages of the flow data points are all preset values;
in each iteration process, calculating the preference of the flow data set participating in the iteration this time:
；
in the method, in the process of the invention, For iteration number,/>Preference of flow data set for mth participating iteration,/>For the number of flow data points contained in the flow data set of the mth participating iteration,/>For the sequence number of the traffic data point in the traffic data set,Get pass/>All integers within the range,/>For the/>, in the traffic datasetNoise level after correction of each flow data point,/>Data completeness of the flow data set for the mth participation iteration;
And finally, acquiring a flow data set with the maximum preference, and taking the flow data point contained in the flow data set as a target flow data point.
8. The cloud platform monitoring method based on digital twinning of claim 7, wherein the data integrity of the iteration-involved flow data set comprises:
classifying the flow data points in the flow data sets participating in iteration according to the numerical value to obtain the flow data types contained in the flow data sets participating in iteration;
classifying all flow data points according to the numerical value to obtain the types of flow data contained in all flow data;
the data completeness of the flow data set participating in the iteration is: the ratio of the number of traffic data categories contained in the traffic data set participating in the iteration to the number of traffic data categories contained in all traffic data.
9. The cloud platform monitoring method based on digital twinning of claim 8, wherein the presetting of the local range of each flow data point includes:
setting a certain flow data point as a center, selecting N flow data points on the left side and N flow data points on the right side of the flow data point, adding the flow data points, and taking the total 2N+1 flow data points as the local range of the flow data points.
10. A digital twinning-based cloud platform monitoring system, characterized in that the monitoring system comprises a memory and a processor, the memory having stored thereon a computer program, the processor executing the computer program to carry out the steps of the monitoring method according to any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410369936.5A CN117972314B (en) | 2024-03-29 | 2024-03-29 | Cloud platform monitoring method and system based on digital twinning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410369936.5A CN117972314B (en) | 2024-03-29 | 2024-03-29 | Cloud platform monitoring method and system based on digital twinning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117972314A true CN117972314A (en) | 2024-05-03 |
| CN117972314B CN117972314B (en) | 2024-06-04 |
Family
ID=90855861
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410369936.5A Active CN117972314B (en) | 2024-03-29 | 2024-03-29 | Cloud platform monitoring method and system based on digital twinning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117972314B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118133053A (en) * | 2024-05-08 | 2024-06-04 | 山东瑞福锂业有限公司 | Industrial data collecting and processing system and method |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109989021A (en) * | 2017-12-30 | 2019-07-09 | 魏永强 | A kind of combination field and the compound vacuum coating method of liner bias conical pipe |
| CN109989016A (en) * | 2017-12-30 | 2019-07-09 | 魏永强 | A kind of combination field, compound tube and the compound vacuum coating method of perforated baffle |
| CN110161958A (en) * | 2019-06-06 | 2019-08-23 | 山东华立供水设备有限公司 | A kind of single frequency conversion automatic control system and its implementation of no-negative-pressure water supply equipment |
| CN113703410A (en) * | 2021-08-31 | 2021-11-26 | 华电国际电力股份有限公司深圳公司 | Wisdom operation platform of wisdom power plant |
| CN216108791U (en) * | 2021-08-25 | 2022-03-22 | 山东华立供水设备有限公司 | Integrated mute pump set |
| CN114925529A (en) * | 2022-05-23 | 2022-08-19 | 西安交通大学 | State correction method, system, equipment and medium for digital twin model of condenser |
| US20230123322A1 (en) * | 2021-04-16 | 2023-04-20 | Strong Force Vcn Portfolio 2019, Llc | Predictive Model Data Stream Prioritization |
| US20230173395A1 (en) * | 2021-07-14 | 2023-06-08 | Strong Force TX Portfolio 2018, LLC | Systems and methods with integrated gaming engines and smart contracts |
| CN116382088A (en) * | 2023-04-20 | 2023-07-04 | 中国海洋大学 | Autonomous control method and system for underwater throttle valve based on digital twin |
| CN116819382A (en) * | 2023-06-25 | 2023-09-29 | 哈尔滨工业大学 | DC-DC converter state monitoring method based on digital twin |
| CN117074064A (en) * | 2023-08-18 | 2023-11-17 | 浙江迪弗莱包装科技股份有限公司 | Big data-based operation monitoring method for air column scalding machine |
| CN117235447A (en) * | 2023-11-10 | 2023-12-15 | 天津六百光年智能科技有限公司 | Building energy data management method and system |
-
2024
- 2024-03-29 CN CN202410369936.5A patent/CN117972314B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109989021A (en) * | 2017-12-30 | 2019-07-09 | 魏永强 | A kind of combination field and the compound vacuum coating method of liner bias conical pipe |
| CN109989016A (en) * | 2017-12-30 | 2019-07-09 | 魏永强 | A kind of combination field, compound tube and the compound vacuum coating method of perforated baffle |
| CN110161958A (en) * | 2019-06-06 | 2019-08-23 | 山东华立供水设备有限公司 | A kind of single frequency conversion automatic control system and its implementation of no-negative-pressure water supply equipment |
| US20230123322A1 (en) * | 2021-04-16 | 2023-04-20 | Strong Force Vcn Portfolio 2019, Llc | Predictive Model Data Stream Prioritization |
| US20230173395A1 (en) * | 2021-07-14 | 2023-06-08 | Strong Force TX Portfolio 2018, LLC | Systems and methods with integrated gaming engines and smart contracts |
| CN216108791U (en) * | 2021-08-25 | 2022-03-22 | 山东华立供水设备有限公司 | Integrated mute pump set |
| CN113703410A (en) * | 2021-08-31 | 2021-11-26 | 华电国际电力股份有限公司深圳公司 | Wisdom operation platform of wisdom power plant |
| CN114925529A (en) * | 2022-05-23 | 2022-08-19 | 西安交通大学 | State correction method, system, equipment and medium for digital twin model of condenser |
| CN116382088A (en) * | 2023-04-20 | 2023-07-04 | 中国海洋大学 | Autonomous control method and system for underwater throttle valve based on digital twin |
| CN116819382A (en) * | 2023-06-25 | 2023-09-29 | 哈尔滨工业大学 | DC-DC converter state monitoring method based on digital twin |
| CN117074064A (en) * | 2023-08-18 | 2023-11-17 | 浙江迪弗莱包装科技股份有限公司 | Big data-based operation monitoring method for air column scalding machine |
| CN117235447A (en) * | 2023-11-10 | 2023-12-15 | 天津六百光年智能科技有限公司 | Building energy data management method and system |
Non-Patent Citations (2)
| Title |
|---|
| 彭新俊;王翼飞;: "快速原空间孪生支持向量回归算法", 模式识别与人工智能, no. 01, 15 February 2011 (2011-02-15) * |
| 王志有;李欢;刘自增;吴加敏;施祖贤;: "基于深度学习算法的卫星影像变化监测", 计算机系统应用, no. 01, 15 January 2020 (2020-01-15) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118133053A (en) * | 2024-05-08 | 2024-06-04 | 山东瑞福锂业有限公司 | Industrial data collecting and processing system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117972314B (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116186634B (en) | Intelligent management system for construction data of building engineering | |
| CN117972314B (en) | Cloud platform monitoring method and system based on digital twinning | |
| CN105868164B (en) | A kind of soft-measuring modeling method based on the linear dynamic system model for having supervision | |
| CN115081795B (en) | Enterprise energy consumption anomaly cause analysis method and system in multidimensional scene | |
| CN113298162A (en) | Bridge health monitoring method and system based on K-means algorithm | |
| CN115438726A (en) | Device life and fault type prediction method and system based on digital twin technology | |
| CN117556366B (en) | Data abnormality detection system and method based on data screening | |
| CN116738353B (en) | Pharmaceutical workshop air filter element performance detection method based on data analysis | |
| CN109840671A (en) | Operational development effect calculates equipment, operational development effect calculation method and recording medium | |
| CN114266289A (en) | Complex equipment health state assessment method | |
| CN113139880A (en) | Wind turbine generator actual power curve fitting method, device, equipment and storage medium | |
| CN117992894B (en) | Agricultural greenhouse environment abnormal data monitoring method based on Internet of things | |
| CN104376078A (en) | Abnormal data detection method based on knowledge entropy | |
| CN112202630A (en) | Network quality abnormity detection method and device based on unsupervised model | |
| CN112348055A (en) | Clustering evaluation measurement method, system, device and storage medium | |
| CN116341911A (en) | Alternating-current interference corrosion risk evaluation method and system based on FAHP-SVM | |
| CN114755010A (en) | Rotary machine vibration fault diagnosis method and system | |
| CN110533109A (en) | A kind of storage spraying production monitoring data and characteristic analysis method and its device | |
| CN117272055B (en) | Electric energy meter abnormality detection method and device based on filtering enhancement self-encoder | |
| WO2022134579A1 (en) | Method and apparatus for detecting abnormalities of service transaction data, and computer device | |
| CN112241512A (en) | Method and system for improving audit precision | |
| CN106295712A (en) | A kind of fault detection method and system | |
| CN116957361A (en) | Ship task system health state detection method based on virtual-real combination | |
| CN112949816A (en) | ANN-based method and system for identifying abnormal data of hydrostatic level | |
| Barlowe et al. | Multivariate visual explanation for high dimensional datasets |
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 |