CN117034043A - Intelligent building comprehensive energy consumption monitoring method and system based on multi-energy Internet of things - Google Patents
Intelligent building comprehensive energy consumption monitoring method and system based on multi-energy Internet of things Download PDFInfo
- Publication number
- CN117034043A CN117034043A CN202311294922.3A CN202311294922A CN117034043A CN 117034043 A CN117034043 A CN 117034043A CN 202311294922 A CN202311294922 A CN 202311294922A CN 117034043 A CN117034043 A CN 117034043A
- Authority
- CN
- China
- Prior art keywords
- center point
- clustering center
- moment
- data
- time axis
- 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
- 238000005265 energy consumption Methods 0.000 title claims abstract description 91
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000012544 monitoring process Methods 0.000 title claims abstract description 29
- 238000012216 screening Methods 0.000 claims abstract description 97
- 230000002159 abnormal effect Effects 0.000 claims abstract description 37
- 238000012937 correction Methods 0.000 claims abstract description 8
- 238000000513 principal component analysis Methods 0.000 claims abstract description 6
- 238000004364 calculation method Methods 0.000 claims description 21
- 230000001186 cumulative effect Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 239000002131 composite material Substances 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 abstract description 7
- 125000004122 cyclic group Chemical group 0.000 abstract description 3
- 238000004422 calculation algorithm Methods 0.000 description 11
- 230000008859 change Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000011218 segmentation Effects 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 5
- 230000035772 mutation Effects 0.000 description 3
- 238000012847 principal component analysis method Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 239000002075 main ingredient Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/08—Construction
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Business, Economics & Management (AREA)
- Pure & Applied Mathematics (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method and a system for monitoring comprehensive energy consumption of an intelligent building based on a multi-energy Internet of things, which belong to the technical field of monitoring data processing, and comprise the following steps: acquiring multidimensional energy consumption data of a building at a plurality of moments; performing principal component analysis on the acquired multidimensional energy consumption data to obtain screening principal components; selecting a plurality of clustering center points when the screening main components are subjected to clustering circulation; calculating a comprehensive expansion coefficient of the moment corresponding to each clustering center point extending along any direction of a time axis, determining a search range after correction of each clustering center point by using the comprehensive expansion coefficient, updating the moment corresponding to the clustering center point, and intercepting abnormal data according to the relevance of the data in a final clustering result after cyclic clustering; the method and the device are used for solving the problem that in the prior art, error scratch is easy to occur when abnormal energy consumption data are monitored, so that the time period in which the abnormal data are located cannot be accurately intercepted.
Description
Technical Field
The invention belongs to the technical field of energy consumption data processing, and particularly relates to a comprehensive energy consumption monitoring method and system for an intelligent building based on a multi-energy Internet of things.
Background
The intelligent building is a great product integrating modern science and technology, mainly comprises modern building technology, modern computer technology and modern communication technology, and optimally combines the structure, system, service and management of the building according to the requirements of users, so that a humanized building environment with high efficiency, comfort and convenience is provided for the users; the energy consumption monitoring system is used for collecting and monitoring the use data of various energy sources in the intelligent building, such as the use data of electric power consumption, gas consumption, energy load and the like, through the internet of things technology.
At present, a central processing unit is generally utilized to analyze and process data obtained by various energy sensors, and the central processing unit is utilized to identify a time period corresponding to an energy waste and low-efficiency area, so that a user can be helped to know the use condition of energy, and can be helped to identify the problems of energy consumption peak time period and low equipment efficiency, and further, a building manager can take targeted measures to optimize energy distribution and reduce energy consumption, thereby improving energy utilization efficiency.
However, in the prior art, when analyzing and processing data obtained by various energy sensors by using a central processing unit, the time periods with different energy consumption are generally divided by using the thought of SLIC super-pixel division so as to find a time period with abnormal energy consumption, but due to the complexity of multi-dimensional data, the situation that abnormal data and normal data are intersected with each other can occur, and for general SLIC super-pixel division, the situation of wrong division can occur in the dividing process, so that the time period in which the abnormal data are located cannot be accurately identified, that is, the abnormal energy consumption situation cannot be accurately monitored.
Disclosure of Invention
The invention provides a comprehensive energy consumption monitoring method and system for an intelligent building based on a multi-energy Internet of things, which are used for solving the problems that in the prior art, when monitoring energy consumption data, the time period in which abnormal data are located cannot be accurately identified, and the abnormal condition of the energy consumption cannot be accurately monitored.
The intelligent building comprehensive energy consumption monitoring method based on the multi-energy Internet of things adopts the following technical scheme:
s1, acquiring multidimensional energy consumption data of a building at a plurality of moments;
s2, performing principal component analysis on the obtained multidimensional energy consumption data at a plurality of moments to obtain a plurality of principal components, extracting variance contribution rate of each principal component, and selecting a screening principal component according to the variance contribution rate of each principal component;
s3, selecting a plurality of clustering center points when the screening main components are subjected to clustering circulation, dividing a time axis into a plurality of sections by utilizing corresponding moments of each clustering center point, and extending and selecting two adjacent first sections and second sections along any direction of the time axis from each clustering center point;
s4, acquiring data of each moment on each screening main component, and calculating expansion coefficients of the moment corresponding to each clustering center point extending in any direction along a time axis in each screening main component according to the similarity of the data in a first interval and a second interval in the same screening main component;
S5, calculating the comprehensive expansion coefficient of each clustering center point corresponding moment extending along any direction of the time axis according to the correlation of the data in the first interval in any two main screening components, the correlation of the data in the second interval in any two main screening components and the expansion coefficient of each clustering center point corresponding moment extending along any direction of the time axis in each main screening component;
s6, correcting the preset initial search range of each clustering center point according to the comprehensive expansion coefficient of the corresponding moment of each clustering center point along any direction of the time axis, and obtaining a corrected search range of the corresponding moment of each clustering center point;
s7, correcting the corresponding moment of each clustering center point to obtain a search range, and clustering the multidimensional energy consumption data to obtain an initial clustering result;
s8, updating the corresponding moment of each clustering center point according to the obtained initial clustering result, and repeatedly executing the steps from S3 to S7 to start the next cycle when the sum of the absolute value of the moment difference value corresponding to each clustering center point of the current cycle and the moment difference value corresponding to the next cycle clustering center point is larger than or equal to a preset first threshold value; terminating the circulation when the sum of the absolute value of the moment difference value corresponding to each clustering center point of the circulation and the moment corresponding to the next circulating clustering center point is smaller than a preset first threshold value, and outputting a final clustering result;
S9, intercepting abnormal data according to the correlation of the data in the final clustering result.
Further, the step of calculating the expansion coefficient of each cluster center point corresponding time extending in any direction along the time axis in each screening principal component according to the similarity of the data in the first section and the second section in the same screening principal component comprises:
performing linear fitting on the data in the first interval to obtain a first straight line;
performing linear fitting on the data in the second interval to obtain a second straight line;
and taking the similarity of the slopes of the first straight line and the second straight line as the expansion coefficient of each clustering center point extending along any direction of the time axis in each screening principal component.
Further, the calculation formula of the expansion coefficient extending along any direction of the time axis in each screening principal component at the moment corresponding to each cluster center point is as follows:
wherein,indicate->The corresponding moment of the clustering center points>In->The individual screening principle components are located along the time axis +.>Expansion coefficient extending in the direction; />Indicate->Screening main components; />Indicate->The clustering center points correspond to moments; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis; Indicate->The screening principle is at the->Performing linear fitting on the data in each interval to obtain the slope of the straight line;indicate->The screening principle is at the->Performing linear fitting on the data in each interval to obtain the slope of the straight line; />An exponential function based on a natural constant is represented.
Further, the calculation formula of the comprehensive expansion coefficient extending along any direction of the time axis at the moment corresponding to each cluster center point is as follows:
wherein,indicate->The corresponding moment of the clustering center points>Along the time axis->A comprehensive expansion coefficient extending in the direction; />Representing the total number of the screening main components; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis; />Expression sequence->And sequence->Pearson correlation coefficient therebetween; />Indicate->The screening principle is at the->A data sequence within a respective interval; />Indicate->The screening principle is at the->A data sequence within a respective interval;expression sequence->And->Pearson correlation coefficient therebetween; />Indicate->The main component is at->A data sequence within a respective interval; />Indicate->The main component is at the firstA data sequence within a respective interval; />Representing an existing normalization function; / >Indicate->The corresponding moment of the clustering center points>In->The individual screening principle components are located along the time axis +.>Expansion coefficient extending in the direction.
Further, the step of correcting the preset initial search range of each cluster center point according to the comprehensive expansion coefficient extending along any direction of the time axis at the time corresponding to each cluster center point to obtain the corrected search range of the time corresponding to each cluster center point comprises the following steps:
arbitrarily selecting two adjacent clustering center points to be marked as a first clustering center point and a second clustering center point;
acquiring a comprehensive expansion coefficient extending along the positive direction of the time axis at the moment corresponding to the first clustering center point, and simultaneously acquiring a comprehensive expansion coefficient extending along the negative direction of the time axis at the moment corresponding to the second clustering center point;
calculating the sum of the comprehensive expansion coefficients extending along the positive direction of the time axis at the moment corresponding to the first clustering center point and the comprehensive expansion coefficients extending along the negative direction of the time axis at the moment corresponding to the second clustering center point, and recording the sum as the sum of the comprehensive expansion coefficients;
calculating the ratio of the comprehensive expansion coefficient of the first cluster center point extending along the positive direction of the time axis to the sum of the comprehensive expansion coefficient and the value, and taking the product of the ratio and the preset initial search range as the search range after correction at the corresponding moment of the first cluster center point;
And calculating the corrected search range of each clustering center point corresponding to the moment according to the calculation method of the corrected search range of the first clustering center point.
Further, the calculation formula of the search range after the correction of the corresponding moment of each cluster center point is as follows:
wherein,indicate->The corresponding moment of the clustering center points>Correcting the search range; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis;indicate->The corresponding moment of the clustering center points>A comprehensive expansion coefficient extending in the positive direction of the time axis; />Indicate->The corresponding moment of the clustering center points>A composite expansion coefficient extending in a negative direction of the time axis; />Representing a preset initial search range->Indicate->The length of the interval, i.e. edge + ->The lengths of the adjacent sections in the direction.
Further, the step of calculating the sum of the absolute value of the time difference value corresponding to the clustering center point of the next cycle and the time corresponding to each clustering center point of the current cycle includes:
selecting any one of the initial clustering results as a target category, and taking a clustering center point in the target category as a target clustering center point;
acquiring a first moment corresponding to a target clustering center point in the current cycle, and simultaneously acquiring a second moment corresponding to a target clustering center point in the next cycle;
Calculating the absolute value of the time difference between the first moment and the second moment, and calculating the absolute value of the time difference corresponding to each clustering center point according to a calculation method of the absolute value of the time difference corresponding to the target clustering center point;
and adding the absolute values of the time differences corresponding to all the clustering center points to obtain the sum of the absolute values of the time differences corresponding to the moment of each clustering center point in the current cycle and the moment of the next cycle clustering center point.
Further, the step of selecting the screening principal component according to the variance contribution rate of each principal component includes:
and calculating the cumulative variance contribution rate according to the variance contribution rate of each principal component, and selecting principal components with the cumulative variance contribution rate larger than a preset second threshold as screening principal components.
Further, the step of intercepting the abnormal data according to the correlation of the data in the final clustering result comprises the following steps:
selecting any one of the final clustering results as a target category, and marking the categories adjacent to the target category on a time axis as target adjacent categories;
calculating the correlation of the data in the target category in any two dimensions, and simultaneously calculating the correlation of the data in the adjacent category of the target in any two dimensions;
And calculating a correlation difference value of the correlation of the data in the target category in any two dimensions and the correlation of the target adjacent category in the corresponding two dimensions, and intercepting multi-dimensional energy consumption data in the target category and the target neighborhood category as abnormal data when the absolute value of the correlation difference value is larger than a preset third threshold value.
The intelligent building comprehensive energy consumption monitoring system based on the multi-energy Internet of things comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the processor realizes the steps of the method when executing the computer program.
The beneficial effects of the invention are as follows:
the method is used for solving the problem that in the prior art, when monitoring the energy consumption data, the time period in which the abnormal data are located cannot be accurately identified; when the clustering circulation is carried out, firstly, dividing a time axis into a plurality of intervals according to a selected clustering center point; for a certain screening principal component, if the similarity between corresponding data sequences in two intervals extending from a clustering center point to a positive direction of a time axis is high, the energy consumption data represented by the screening principal component has small change in the time period, and a search range of the clustering center point should be given a higher expansion coefficient, so that more similar data corresponding times are divided into the same intervals, and therefore, according to the similarity of the data of the same screening principal component in adjacent intervals, the expansion coefficient of each clustering center point corresponding time extending in any direction along the time axis in each screening principal component is obtained.
Because the expansion coefficient of each clustering center point extending along any direction of a time axis in each screening main component at the corresponding moment represents the difference of data in adjacent intervals in a single screening main component, and each screening main component represents energy consumption data with different dimensions, the calculation result of the expansion coefficient is different possibly due to hysteresis and other factors among the data, so that the expansion coefficient of each clustering center point is adjusted according to the correlation of the multidimensional energy consumption data of different main components in the same interval to obtain a comprehensive expansion coefficient; according to the invention, the expansion coefficient of each clustering center point is adjusted, so that the probability of occurrence of abnormal mutation energy consumption data in the same category is reduced, the accuracy of segmentation of the time period where the abnormal data are located is improved, interception of the abnormal time period in the monitoring data is facilitated, and further, the energy waste is reduced by adjusting the energy distribution.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a general flow chart of steps of a method for monitoring comprehensive energy consumption of a smart building based on multi-energy Internet of things;
FIG. 2 is a diagram showing the calculation of the cumulative variance contribution rate according to the variance contribution rate of each principal component in the present invention;
FIG. 3 is a schematic diagram of dividing a time axis into a plurality of intervals by utilizing corresponding moments of each cluster center point in the invention;
FIG. 4 is a schematic diagram of a target class and a target neighbor class according to the present invention;
FIG. 5 is a schematic diagram of a polyline obtained by fitting data of a target class and a target adjacent class in any two dimensions in the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. 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.
Example 1:
the method for monitoring comprehensive energy consumption of intelligent building based on multi-energy Internet of things provided by the embodiment is as shown in fig. 1, and comprises the following steps:
S1, acquiring multidimensional energy consumption data of a building at a plurality of moments.
The application scenario for this embodiment is that a plurality of sensors on a smart building are utilized to collect multidimensional energy consumption data, the multidimensional energy consumption data in one day is collected, the data sampling interval is 20s, each time the multidimensional energy consumption data at one moment is obtained through sampling, the multidimensional energy consumption data obtained in this embodiment comprises electric power energy consumption data A1, gas energy consumption data A2, air conditioner energy consumption data A3, water energy consumption data A4, equipment energy consumption data A5, temperature energy consumption data A6 and solar energy consumption data A7.
The method comprises the steps that electric power energy consumption data A1, gas energy consumption data A2, air conditioner energy consumption data A3, water energy consumption data A4, equipment energy consumption data A5, temperature energy consumption data A6 and solar energy consumption data A7 collected at each moment form a seven-dimensional vector, the seven-dimensional vector is recorded as multi-dimensional energy consumption data at each moment, and multi-dimensional energy consumption data at all moments in a day are obtained up to this point;
s2, performing principal component analysis on the obtained multidimensional energy consumption data at a plurality of moments to obtain a plurality of principal components, extracting variance contribution rate of each principal component, and selecting a screening principal component according to the variance contribution rate of each principal component.
The step of selecting the screening principal components according to the variance contribution rate of each principal component comprises the following steps: and calculating the cumulative variance contribution rate according to the variance contribution rate of each principal component, and selecting principal components with the cumulative variance contribution rate larger than a preset second threshold as screening principal components.
In the multidimensional energy consumption data, redundant dimensions or variables exist, and the variables possibly have no obvious distinguishing capability in analysis, so that redundant information of the multidimensional energy consumption data can be identified and removed by using a PCA principal component analysis method, and the definition of the data is improved. And carrying out principal component analysis on the obtained multidimensional energy consumption data at all moments by using a PCA principal component analysis method to obtain a plurality of principal components.
It should be noted that the principal component analysis method of PCA is an existing data analysis method, and is commonly used for dimension reduction of high-dimensional data, and is used for extracting principal feature components of the data. After each principal component is obtained, the variance contribution rate of each principal component is extracted as a known technique. The variance contribution rate refers to the contribution degree of each principal component to the variance of the original data, the larger the general variance contribution rate is, the more important the principal component is to explain the original data, the variance contribution rate is calculated by dividing the variance of each principal component by the total variance of the original data to obtain a percentage, and in practical application, the principal component can be obtained by using PCA analysis software and the variance contribution rate can be extracted.
The method for calculating the accumulated variance contribution rate according to the variance contribution rate of each principal component comprises the following steps: sorting the obtained variance contribution rates of all the main components from large to small, calculating the accumulated variance contribution rate after obtaining the variance contribution rate of each main component, presetting a second threshold value as 85% in the embodiment, and selecting the front of the accumulated variance contribution rate being more than 85%The main component is selected as the main component of the screening, and the main component is marked as +.>,/>,…,/>The following analysis was performed for a plurality of screening principal components.
As shown in fig. 2, if a graph is obtained in which the cumulative variance contribution rate is calculated from the variance contribution rate of each principal componentMain ingredient of will->The variance contribution rate of the main components is sequenced from big to small and accumulated to obtain the accumulated variance contribution rate, if accumulated to the +.>The cumulative variance contribution rate after each principal component is 86%, then the former +.>The individual principal components are used as screening principal components.
S3, selecting a plurality of clustering center points when the screening main components are subjected to clustering circulation, dividing a time axis into a plurality of sections by utilizing corresponding moments of each clustering center point, and extending and selecting two adjacent first sections and second sections along any direction of the time axis from each clustering center point.
As shown in fig. 3, a schematic diagram of dividing a time axis into a plurality of intervals by using corresponding moments of each cluster center point is shown; in fig. 3, the position of the 0 point is the origin of the time axis, and the extension along the left side of the time axis is the extension along the negative direction of the time axis, and the extension along the right side of the time axis is the square along the time axisAnd extending, each screening principal component corresponds to one dimension data, so that each clustering center point corresponds to a plurality of principal component data at the same time. The clustering center points are required to be selected equidistantly during the clustering cycle, and the time interval between two adjacent clustering center points selected during the primary cycle is recorded as,/>The recommended value of (2) is +.>If the starting time is 0, interval +.>The cluster center points after the clustering are marked as 1 moment, and the interval after 1 moment is +.>The cluster center points after the clustering are marked as 2 moments, a plurality of cluster center points are sequentially selected, one interval is formed between the 0 moment and the 1 moment, the other interval is formed between the 1 moment and the 2 moment, and the time axis is divided into a plurality of intervals by utilizing the corresponding moment of each cluster center point in the same way; meanwhile, taking the 5 th moment as an example, two adjacent first sections and second sections are selected from the 5 th moment along the positive direction of the time axis, namely, the section between the 5 th moment and the 6 th moment is the first section, and the section between the 6 th moment and the 7 th moment is the second section.
It should be noted that, the clustering algorithm used in this embodiment is a SLIC super-pixel segmentation algorithm, which is a clustering algorithm, and the method for approximately implementing the algorithm is as follows: firstly randomly selecting a plurality of clustering center points distributed at equal intervals, updating the clustering center points according to the distance between data points in the searching range of the clustering center points, and finally continuously updating the positions of the clustering center points in a plurality of times of circulation to obtain a final clustering result.
It should be further noted that, the SLIC super-pixel segmentation algorithm is used for image data, and the multi-dimensional energy consumption data in the embodiment is time sequence data, so that the embodiment inputs the image with the time sequence multi-dimensional energy consumption data as the behavior one into the SLIC super-pixel segmentation algorithm, and the specific SLIC super-pixel segmentation algorithm is a known technology, and the specific implementation method thereof is not repeated in the embodiment.
S4, acquiring data of each moment on each screening principal component, and calculating expansion coefficients of the moment corresponding to each clustering center point extending in any direction along a time axis in each screening principal component according to similarity of data in a first interval and a second interval in the same screening principal component.
According to the similarity of the data in the first interval and the second interval in the same screening principal component, the step of calculating the expansion coefficient of each clustering center point corresponding moment extending along any direction of the time axis in each screening principal component comprises the following steps: performing linear fitting on the data in the first interval to obtain a first straight line; performing linear fitting on the data in the second interval to obtain a second straight line; and taking the similarity of the slopes of the first straight line and the second straight line as the expansion coefficient of each clustering center point extending along any direction of the time axis in each screening principal component.
After the multidimensional energy consumption data of each moment is known, the data of each screening principal component at each moment is obtained, which belongs to the prior known technology, and a specific acquisition method is to calculate the projection value of the multidimensional energy consumption data of each moment on each screening principal component and record the projection value as the data of each moment on each screening principal component, wherein each data corresponds to one moment, so that all the data on each screening principal component form a data sequence of each screening principal component according to the sequence of the moments. It should be noted that, according to the principal component analysis algorithm, each screening principal component is a unit vector, and the projection value of all the multidimensional energy consumption data on each screening principal component is equal to the inner product of the multidimensional energy consumption data and the screening principal component.
For a certain screening principal component, if the similarity between corresponding data sequences is higher in two intervals extending and selected from the clustering center point in any direction of the time axis, the energy consumption data represented by the principal component has smaller change in the time period, and a search range of the clustering center point should be given a higher expansion coefficient, so that more similar data are correspondingly divided into the same intervals at the same time.
The calculation formula of the expansion coefficient extending along any direction of the time axis in each screening principal component at the moment corresponding to each clustering center point is as follows:
wherein,indicate->The corresponding moment of the clustering center points>In->The individual screening principle components are located along the time axis +.>Expansion coefficient extending in the direction; />Indicate->Screening main components; />Indicate->The clustering center points correspond to moments; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis;indicate->The screening principle is at the->Performing linear fitting on the data in each interval to obtain the slope of the straight line;indicate->The screening principle is at the->Performing linear fitting on the data in each interval to obtain the slope of the straight line; / >An exponential function based on a natural constant is represented.
In the calculation formula of the expansion coefficient of each cluster center point in any extending direction along the time axis in each screening principal component, for the firstThe search range of each cluster center point should be different in the positive and negative directions along the time axis, so that the change of the data in the positive and negative directions along the time axis needs to be considered, and the search range is analyzed by the analysis>When the value of (2) is 1,Indicate->The first interval of the cluster center points extending in the positive direction of the time axis, i.e. the first and the +.>The interval adjacent to the cluster center is also called the first interval,>indicate->The second interval, the second and the first, of the cluster center points extending in the positive direction of the time axis>The interval adjacent to the clustering center points is also called a second interval; when->When the value of (2) is-1, ">Indicate->A first interval in which the cluster center points extend in the negative direction of the time axis,/>Indicate->And a second interval in which the clustering center points extend along the positive direction of the time axis is used for judging whether the adjacent area needs to be expanded or not by measuring the change condition of the data points in two adjacent intervals.
For example: if in the present embodimentThen->When the value of (2) is 1, the first interval is the interval between the corresponding moment of the 5 th clustering center point and the corresponding moment of the 6 th clustering center point, the second interval is the interval between the corresponding moment of the 6 th clustering center point and the corresponding moment of the 7 th clustering center point, and meanwhile, the data acquired in the step are the data of the same screening main component.
S5, according to the correlation of the data in the first interval in any two screening main components, the correlation of the data in the second interval in any two screening main components and the expansion coefficient of each clustering center point corresponding moment extending along any direction of a time axis in each screening main component, calculating the comprehensive expansion coefficient of each clustering center point corresponding moment extending along any direction of the time axis.
In step S4, the expansion coefficient of each cluster center point in each screening principal component along any extending direction of the time axis is calculated, which is characterized in that in a single screening principal component, the difference of the data in adjacent intervals may cause the calculation result of the expansion coefficient to be different due to the hysteresis between the data and other factors. For some mutation data in the screening principal components, the mutation data may be caused by normal energy consumption change, but the expansion coefficient of the corresponding center point is smaller, so that the obtained searching range is poor in error only according to the data change of a single principal component; consider that when energy consumption data is abnormal, the correlation coefficient between different principal components of the corresponding interval will be suddenly changed, and otherwise will be kept at a relatively stable level. For example, in one day of data, the correlation coefficient between the temperature data and the air conditioner energy consumption data should be stable, when the correlation coefficient has a large difference, the abnormal probability of the corresponding time period is high, and the corresponding expansion coefficient should be adjusted to be small, so as to prevent the partial abnormal data and the normal data from being divided into the same category, and therefore, the expansion coefficient of each cluster center point is adjusted according to the correlation of the data of different screening main components in the same interval to obtain the comprehensive expansion coefficient.
The calculation formula of the comprehensive expansion coefficient extending along any direction of the time axis at the moment corresponding to each clustering center point is as follows:
wherein,indicate->The corresponding moment of the clustering center points>Along the time axis->A comprehensive expansion coefficient extending in the direction; />Representing the total number of the screening main components; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis; />Expression sequence->And sequence->Pearson correlation coefficient therebetween; />Indicate->The screening principle is at the->A data sequence within a respective interval; />Indicate->The screening principle is at the->A data sequence within a respective interval;expression sequence->And->Pearson correlation coefficient therebetween; />Indicate->The main component is at->A data sequence within a respective interval; />Indicate->The main component is at the firstA data sequence within a respective interval; />Representing an existing normalization function; />Indicate->The corresponding moment of the clustering center points>In->The individual screening principle components are located along the time axis +.>Expansion coefficient extending in the direction.
In the calculation formula of the comprehensive expansion coefficient extending along any direction of the time axis at the moment corresponding to each clustering center point, the first is utilizedThe corresponding moment of the clustering center points >In the individual screening principle components +.>The expansion coefficients extending in the direction are weighted and averaged, and the weight coefficients are obtained by utilizing the change of the correlation coefficients of the main components in the corresponding intervals; if%>Major component and->The main components have smaller correlation coefficient difference between two adjacent regions, namely corresponding to two regionsIn the middle, the influence between the two main components is changed synchronously, corresponding to +.>The principal components are at the cluster center>Time lower edge->The confidence of the expansion coefficient in the direction should be higher, whereas the expansion coefficient thereof should be compressed. Meanwhile, it should be noted that +.>When using the data sequences of different screening principal components in the same interval, calculating +.>Data sequences with different screening principal components within the same interval are also utilized.
And S6, correcting the preset initial search range of each clustering center point according to the comprehensive expansion coefficient of the moment corresponding to each clustering center point, which extends along any direction of the time axis, so as to obtain a corrected search range of the moment corresponding to each clustering center point.
Correcting the initial search range of each clustering center point according to the comprehensive expansion coefficient extending along any direction of the time axis at the moment corresponding to each clustering center point, and obtaining the corrected search range of each clustering center point comprises the following steps: arbitrarily selecting two adjacent clustering center points to be marked as a first clustering center point and a second clustering center point; acquiring a comprehensive expansion coefficient extending along the positive direction of the time axis at the moment corresponding to the first clustering center point, and simultaneously acquiring a comprehensive expansion coefficient extending along the negative direction of the time axis at the moment corresponding to the second clustering center point; calculating the sum of the comprehensive expansion coefficients extending along the positive direction of the time axis at the moment corresponding to the first clustering center point and the comprehensive expansion coefficients extending along the negative direction of the time axis at the moment corresponding to the second clustering center point, and recording the sum as the sum of the comprehensive expansion coefficients; calculating the ratio of the comprehensive expansion coefficient of the first cluster center point extending along the positive direction of the time axis to the sum of the comprehensive expansion coefficient and the value, and taking the product of the ratio and the preset initial search range as the search range after correction at the corresponding moment of the first cluster center point; and calculating the corrected search range of each clustering center point corresponding to the moment according to the calculation method of the corrected search range of the first clustering center point.
The calculation formula of the search range after the correction of the corresponding moment of each cluster center point is as follows:
wherein,indicate->The corresponding moment of the clustering center points>Correcting the search range; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis;indicate->The corresponding moment of the clustering center points>A comprehensive expansion coefficient extending in the positive direction of the time axis; />Indicate->The corresponding moment of the clustering center points>A composite expansion coefficient extending in a negative direction of the time axis; />Representing a preset initial search range->Indicate->The length of the interval, i.e. edge + ->The lengths of the adjacent sections in the direction.
In a calculation formula of a search range after the correction of the corresponding moment of each cluster center point, on the basis of a preset initial search range, determining a search interval of each cluster center point by utilizing the proportion between expansion coefficients between adjacent moments;the method aims to ensure that the search ranges of two adjacent moments along the intersecting direction ensure certain overlapping, and prevent the occurrence of excessive points from occurring in the search range of a single clustering center point.
It should be noted that in the SLIC superpixel segmentation algorithm, the distance between two cluster center points, i.e., the length of two intervals, is twice the preset initial search range, and the positions of the cluster center points are changed during the clustering cycle, i.e., the distance between two cluster center points is changed, so that the initial search is re-preset once according to the cluster center point position during each cycle The cable range, i.e. the preset initial search range isWhen->The preset initial search range may also change when the change occurs.
And S7, correcting the corresponding moment of each clustering center point, searching the range, and clustering the multidimensional energy consumption data to obtain an initial clustering result.
The method comprises the steps of obtaining a search range after the corresponding moment of each clustering center point is corrected, and clustering the multidimensional energy consumption data to obtain an initial clustering result, wherein the steps comprise: calculating the Euclidean distance between each clustering center point and each multidimensional energy consumption data in the corrected search range; and clustering according to the Euclidean distance between the multidimensional energy consumption data and each clustering center point to obtain an initial clustering result. It should be noted that in the existing superpixel segmentation algorithm, after the search range of the cluster center point is obtained, the step of clustering the data to obtain a clustering result is to calculate the euclidean distance between each data point and the cluster center point in each search range of the cluster center point, and the search range between a plurality of cluster center points has an overlapping portion, so that one data point needs to calculate the distance between the data point and a plurality of cluster center points, select the cluster center point with the minimum euclidean distance, divide the data point and the cluster center point into clusters, and finally output the clustering result, and the specific division mode of the data point and the cluster center point is also the existing known technology.
And (2) after the search range is obtained after the corresponding moment of each clustering center point is corrected, clustering the multidimensional energy consumption data of each moment obtained in the step (S1), wherein the result obtained by the first cyclic clustering is an initial clustering result.
S8, updating the corresponding moment of each clustering center point according to the obtained initial clustering result, and repeatedly executing the steps from S3 to S7 to start the next cycle when the sum of the absolute value of the moment difference value corresponding to each clustering center point of the current cycle and the moment difference value corresponding to the next cycle clustering center point is larger than or equal to a preset first threshold value; and stopping the circulation when the sum of the absolute value of the moment difference value corresponding to each clustering center point of the circulation and the moment corresponding to the next circulating clustering center point is smaller than a preset first threshold value, and outputting a final clustering result.
And updating the corresponding moment of each clustering center point according to the obtained initial clustering result, and obtaining the next cyclic clustering center point by using the existing SLIC algorithm to update the clustering center point position according to the target clustering center point.
The calculating step of the sum of the absolute value of the time difference value corresponding to the time corresponding to each clustering center point of the current cycle and the next cycle clustering center point comprises the following steps:
Selecting any one of the initial clustering results as a target category, and taking a clustering center point in the target category as a target clustering center point; acquiring a first moment corresponding to a target clustering center point in the current cycle, and simultaneously acquiring a second moment corresponding to a target clustering center point in the next cycle; calculating the absolute value of the time difference between the first moment and the second moment, and calculating the absolute value of the time difference corresponding to each clustering center point according to a calculation method of the absolute value of the time difference corresponding to the target clustering center point; and adding the absolute values of the time differences corresponding to all the clustering center points to obtain the sum of the absolute values of the time differences corresponding to the moment of each clustering center point in the current cycle and the moment of the next cycle clustering center point.
It should be noted that, when the time corresponding to each cluster center point hardly changes, that is, the position of the cluster center point hardly changes, the cycle is terminated, and the condition of the cycle is as follows:
wherein,indicating the%>The moments corresponding to the clustering center points; />Indicating the next cycle +.>The moments corresponding to the clustering center points; />Representing the total number of clustering center points; />Representing a preset first threshold value +.in this embodiment >Is 10.
S9, intercepting abnormal data according to the correlation of the data in the final clustering result.
The step of detecting the abnormal data according to the correlation of the data in the final clustering result comprises the following steps: selecting any one of the final clustering results as a target category, and marking the categories adjacent to the target category on a time axis as target adjacent categories; calculating the correlation of the data in the target category in any two dimensions, and simultaneously calculating the correlation of the data in the adjacent category of the target in any two dimensions; and calculating a correlation difference value of the correlation of the data in the target category in any two dimensions and the correlation of the target adjacent category in the corresponding two dimensions, and intercepting multi-dimensional energy consumption data in the target category and the target neighborhood category as abnormal data when the absolute value of the correlation difference value is larger than a preset third threshold value.
As shown in fig. 4, a schematic diagram of a target class and a target adjacent class is shown, it should be noted that in this embodiment, the target class corresponds to seven-dimensional data, the target adjacent class corresponds to seven-dimensional data, and the correlation between any two dimensional data is calculated when intercepting abnormal data. As shown in fig. 5, in order to fit the data of the target class and the target adjacent class in any two dimensions to obtain a broken line schematic diagram, for example, if there is a sudden change in the correlation between A3 and A6, it is considered that the data in the target class and the data in the target adjacent class are abnormal, in this embodiment, the abnormal data is intercepted to find out the position of the abnormal data, and then the subsequent analysis and processing are performed, so that the problem of low equipment efficiency is found out, so that only the section of the abnormal data needs to be found out, the position of the abnormal data is quickly located, and when the problem is analyzed and processed in the subsequent analysis, the data before and after the section of the abnormal data is analyzed, so that the problem of low energy consumption efficiency is accurately found out.
The suggested value of the third threshold in this embodiment isWhen the correlation difference between the correlation of the data in the target category in any two dimensions and the correlation of the target adjacent category in the corresponding two dimensions is larger than a preset third threshold value +.>When the correlation between the target class and the target neighborhood class is abnormal, the data in the target class and the target neighborhood class are cut out as abnormal data as long as the correlation between any two dimensions is abnormal.
Example 2:
the intelligent building comprehensive energy consumption monitoring system based on the multi-energy Internet of things provided by the embodiment comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the processor realizes the steps of an intelligent building comprehensive energy consumption monitoring method based on the multi-energy Internet of things when executing the computer program.
The invention provides a comprehensive energy consumption monitoring method for an intelligent building based on a multi-energy Internet of things, which is used for solving the problems that in the prior art, when monitoring energy consumption data, the time period in which abnormal data are located cannot be accurately identified, and the abnormal condition of the energy consumption cannot be accurately monitored.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy Internet of things is characterized by comprising the following steps of:
s1, acquiring multidimensional energy consumption data of a building at a plurality of moments;
s2, performing principal component analysis on the obtained multidimensional energy consumption data at a plurality of moments to obtain a plurality of principal components, extracting variance contribution rate of each principal component, and selecting a screening principal component according to the variance contribution rate of each principal component;
s3, selecting a plurality of clustering center points when the screening main components are subjected to clustering circulation, dividing a time axis into a plurality of sections by utilizing corresponding moments of each clustering center point, and extending and selecting two adjacent first sections and second sections along any direction of the time axis from each clustering center point;
s4, acquiring data of each moment on each screening main component, and calculating expansion coefficients of the moment corresponding to each clustering center point extending in any direction along a time axis in each screening main component according to the similarity of the data in a first interval and a second interval in the same screening main component;
S5, calculating the comprehensive expansion coefficient of each clustering center point corresponding moment extending along any direction of the time axis according to the correlation of the data in the first interval in any two main screening components, the correlation of the data in the second interval in any two main screening components and the expansion coefficient of each clustering center point corresponding moment extending along any direction of the time axis in each main screening component;
s6, correcting the preset initial search range of each clustering center point according to the comprehensive expansion coefficient of the corresponding moment of each clustering center point along any direction of the time axis, and obtaining a corrected search range of the corresponding moment of each clustering center point;
s7, correcting the corresponding moment of each clustering center point to obtain a search range, and clustering the multidimensional energy consumption data to obtain an initial clustering result;
s8, updating the corresponding moment of each clustering center point according to the obtained initial clustering result, and repeatedly executing the steps from S3 to S7 to start the next cycle when the sum of the absolute value of the moment difference value corresponding to each clustering center point of the current cycle and the moment difference value corresponding to the next cycle clustering center point is larger than or equal to a preset first threshold value; terminating the circulation when the sum of the absolute value of the moment difference value corresponding to each clustering center point of the circulation and the moment corresponding to the next circulating clustering center point is smaller than a preset first threshold value, and outputting a final clustering result;
S9, intercepting abnormal data according to the correlation of the data in the final clustering result.
2. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the step of calculating the expansion coefficient of each clustering center point corresponding moment extending in any direction along the time axis in each screening principal component according to the similarity of data in the first interval and the second interval in the same screening principal component comprises:
performing linear fitting on the data in the first interval to obtain a first straight line;
performing linear fitting on the data in the second interval to obtain a second straight line;
and taking the similarity of the slopes of the first straight line and the second straight line as the expansion coefficient of each clustering center point extending along any direction of the time axis in each screening principal component.
3. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 2, wherein the calculation formula of the expansion coefficient extending along any direction of a time axis in each screening principal component at the moment corresponding to each clustering center point is as follows:
;
wherein,indicate->The corresponding moment of the clustering center points>In->The individual screening principle components are located along the time axis +. >Expansion coefficient extending in the direction; />Indicate->Screening main components; />Indicate->The clustering center points correspond to moments; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis;indicate->The screening principle is at the->Performing linear fitting on the data in each interval to obtain the slope of the straight line; />Indicate->The screening principle is at the->Performing linear fitting on the data in each interval to obtain the slope of the straight line; />An exponential function based on a natural constant is represented.
4. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the calculation formula of the comprehensive expansion coefficient extending along any direction of a time axis at the moment corresponding to each clustering center point is as follows:
;
wherein,indicate->The corresponding moment of the clustering center points>Along the time axis/>A comprehensive expansion coefficient extending in the direction; />Representing the total number of the screening main components; />Representing a direction representative value, ++>When->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis; />Expression sequence->And sequencePearson correlation coefficient therebetween; / >Indicate->The screening principle is at the->A data sequence within a respective interval; />Indicate->The screening principle is at the->A data sequence within a respective interval;expression sequence->And->Pearson correlation coefficient therebetween; />Indicate->The main component is at->A data sequence within a respective interval; />Indicate->The main component is at the firstA data sequence within a respective interval; />Representing an existing normalization function; />Indicate->The corresponding moment of the clustering center points>In->The individual screening principle components are located along the time axis +.>Expansion coefficient extending in the direction.
5. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the step of correcting the preset initial search range of each cluster center point according to the comprehensive expansion coefficient of each cluster center point corresponding to the moment extending along any direction of the time axis to obtain the corrected search range of each cluster center point corresponding to the moment comprises the following steps:
arbitrarily selecting two adjacent clustering center points to be marked as a first clustering center point and a second clustering center point;
acquiring a comprehensive expansion coefficient extending along the positive direction of the time axis at the moment corresponding to the first clustering center point, and simultaneously acquiring a comprehensive expansion coefficient extending along the negative direction of the time axis at the moment corresponding to the second clustering center point;
Calculating the sum of the comprehensive expansion coefficients extending along the positive direction of the time axis at the moment corresponding to the first clustering center point and the comprehensive expansion coefficients extending along the negative direction of the time axis at the moment corresponding to the second clustering center point, and recording the sum as the sum of the comprehensive expansion coefficients;
calculating the ratio of the comprehensive expansion coefficient of the first cluster center point extending along the positive direction of the time axis to the sum of the comprehensive expansion coefficient and the value, and taking the product of the ratio and the preset initial search range as the search range after correction at the corresponding moment of the first cluster center point;
and calculating the corrected search range of each clustering center point corresponding to the moment according to the calculation method of the corrected search range of the first clustering center point.
6. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the calculation formula of the search range after the correction of the corresponding moment of each clustering center point is as follows:
;
wherein,indicate->The corresponding moment of the clustering center points>Correcting the search range; />The representative value of the direction is indicated,when->When 1 is taken, it means extending in the positive direction of the time axis when +.>Taking-1 to indicate extending in the negative direction of the time axis; />Indicate->The corresponding moment of the clustering center points >A comprehensive expansion coefficient extending in the positive direction of the time axis; />Represent the firstThe corresponding moment of the clustering center points>A composite expansion coefficient extending in a negative direction of the time axis; />Representing a preset initial search range->Indicate->The length of the interval, i.e. edge + ->The lengths of the adjacent sections in the direction.
7. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the calculating step of the sum of the absolute value of the time difference value corresponding to each clustering center point of the current cycle and the next cycle clustering center point comprises the following steps:
selecting any one of the initial clustering results as a target category, and taking a clustering center point in the target category as a target clustering center point;
acquiring a first moment corresponding to a target clustering center point in the current cycle, and simultaneously acquiring a second moment corresponding to a target clustering center point in the next cycle;
calculating the absolute value of the time difference between the first moment and the second moment, and calculating the absolute value of the time difference corresponding to each clustering center point according to a calculation method of the absolute value of the time difference corresponding to the target clustering center point;
and adding the absolute values of the time differences corresponding to all the clustering center points to obtain the sum of the absolute values of the time differences corresponding to the moment of each clustering center point in the current cycle and the moment of the next cycle clustering center point.
8. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the step of selecting the screening principal component according to the variance contribution rate of each principal component comprises the steps of:
and calculating the cumulative variance contribution rate according to the variance contribution rate of each principal component, and selecting principal components with the cumulative variance contribution rate larger than a preset second threshold as screening principal components.
9. The intelligent building comprehensive energy consumption monitoring method based on the multi-energy internet of things according to claim 1, wherein the step of intercepting abnormal data according to the correlation of the data in the final clustering result comprises the following steps:
selecting any one of the final clustering results as a target category, and marking the categories adjacent to the target category on a time axis as target adjacent categories;
calculating the correlation of the data in the target category in any two dimensions, and simultaneously calculating the correlation of the data in the adjacent category of the target in any two dimensions;
and calculating a correlation difference value of the correlation of the data in the target category in any two dimensions and the correlation of the target adjacent category in the corresponding two dimensions, and intercepting multi-dimensional energy consumption data in the target category and the target neighborhood category as abnormal data when the absolute value of the correlation difference value is larger than a preset third threshold value.
10. The intelligent building comprehensive energy consumption monitoring system based on the multi-energy Internet of things comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and is characterized in that the steps of the method according to any one of claims 1-9 are realized when the processor executes the computer program.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311294922.3A CN117034043B (en) | 2023-10-09 | 2023-10-09 | Intelligent building comprehensive energy consumption monitoring method and system based on multi-energy Internet of things |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311294922.3A CN117034043B (en) | 2023-10-09 | 2023-10-09 | Intelligent building comprehensive energy consumption monitoring method and system based on multi-energy Internet of things |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117034043A true CN117034043A (en) | 2023-11-10 |
CN117034043B CN117034043B (en) | 2024-01-16 |
Family
ID=88639844
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311294922.3A Active CN117034043B (en) | 2023-10-09 | 2023-10-09 | Intelligent building comprehensive energy consumption monitoring method and system based on multi-energy Internet of things |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117034043B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117421618A (en) * | 2023-11-24 | 2024-01-19 | 上海东方低碳科技产业股份有限公司 | Building energy consumption monitoring method and system |
CN117436006A (en) * | 2023-12-22 | 2024-01-23 | 圣道天德电气(山东)有限公司 | Intelligent ring main unit fault real-time monitoring method and system |
CN117455725A (en) * | 2023-12-22 | 2024-01-26 | 广惠建设工程集团有限公司 | Building energy consumption management method and system based on BIM (building information modeling) building |
CN117609813A (en) * | 2024-01-23 | 2024-02-27 | 山东第一医科大学附属省立医院(山东省立医院) | Intelligent management method for intensive patient monitoring data |
CN117668684A (en) * | 2024-01-31 | 2024-03-08 | 新风光电子科技股份有限公司 | Power grid electric energy data anomaly detection method based on big data analysis |
CN117952569A (en) * | 2024-03-27 | 2024-04-30 | 山东省科学院能源研究所 | Public building collaborative energy supply management system based on multisource renewable energy sources |
CN117977717A (en) * | 2024-04-01 | 2024-05-03 | 国网黑龙江省电力有限公司佳木斯供电公司 | Cold region wind-solar-thermal energy storage comprehensive energy collaborative management method and system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2016206784A (en) * | 2015-04-17 | 2016-12-08 | 株式会社Ihi | Data analysis device and data analysis method |
EP3839680A1 (en) * | 2019-12-20 | 2021-06-23 | Robert Bosch GmbH | Method and device for controlling a machine using principal component analysis |
CN115329446A (en) * | 2022-10-13 | 2022-11-11 | 江苏航运职业技术学院 | Digital twinning modeling method for intelligent hoisting process of prefabricated parts of fabricated building |
CN116109039A (en) * | 2023-02-22 | 2023-05-12 | 扬州市职业大学(扬州开放大学) | Data-driven anomaly detection and early warning system |
CN116308456A (en) * | 2023-02-23 | 2023-06-23 | 国网新疆电力有限公司信息通信公司 | User electricity consumption behavior analysis method, system, device and medium in big data environment |
CN116431975A (en) * | 2023-06-12 | 2023-07-14 | 陕西巨人商务信息咨询有限公司 | Environment monitoring method and system for data center |
CN116579506A (en) * | 2023-07-13 | 2023-08-11 | 陕西通信规划设计研究院有限公司 | Building energy consumption data intelligent management method and system based on big data |
-
2023
- 2023-10-09 CN CN202311294922.3A patent/CN117034043B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2016206784A (en) * | 2015-04-17 | 2016-12-08 | 株式会社Ihi | Data analysis device and data analysis method |
EP3839680A1 (en) * | 2019-12-20 | 2021-06-23 | Robert Bosch GmbH | Method and device for controlling a machine using principal component analysis |
CN115329446A (en) * | 2022-10-13 | 2022-11-11 | 江苏航运职业技术学院 | Digital twinning modeling method for intelligent hoisting process of prefabricated parts of fabricated building |
CN116109039A (en) * | 2023-02-22 | 2023-05-12 | 扬州市职业大学(扬州开放大学) | Data-driven anomaly detection and early warning system |
CN116308456A (en) * | 2023-02-23 | 2023-06-23 | 国网新疆电力有限公司信息通信公司 | User electricity consumption behavior analysis method, system, device and medium in big data environment |
CN116431975A (en) * | 2023-06-12 | 2023-07-14 | 陕西巨人商务信息咨询有限公司 | Environment monitoring method and system for data center |
CN116579506A (en) * | 2023-07-13 | 2023-08-11 | 陕西通信规划设计研究院有限公司 | Building energy consumption data intelligent management method and system based on big data |
Non-Patent Citations (5)
Title |
---|
HANG SU ET.AL.: ""Nonlocal feature learning based on a variational graph auto-encoder network for small area change detection using SAR imagery"", 《ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING》, vol. 193, pages 137 - 149, XP087198486, DOI: 10.1016/j.isprsjprs.2022.09.006 * |
张雪茹;尹志强;姚亦锋;胡美娟;洪永胜;: "安徽省城市建设用地变化及驱动力分析", 长江流域资源与环境, vol. 25, no. 04, pages 544 - 551 * |
李琳: ""建筑能效分析技术及能耗预测模型研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, vol. 2020, no. 12, pages 038 - 46 * |
阮孟丽等: ""基于SLIC-DPC算法的车辆检测研究"", 《齐齐哈尔大学学报》, vol. 37, no. 02, pages 41 - 45 * |
陈伟: ""群体智能算法及其在基因表达数据聚类中的应用"", 《中国博士学位论文全文数据库信息科技辑》, vol. 2011, no. 12, pages 140 - 32 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117421618A (en) * | 2023-11-24 | 2024-01-19 | 上海东方低碳科技产业股份有限公司 | Building energy consumption monitoring method and system |
CN117421618B (en) * | 2023-11-24 | 2024-04-05 | 上海东方低碳科技产业股份有限公司 | Building energy consumption monitoring method and system |
CN117436006B (en) * | 2023-12-22 | 2024-03-15 | 圣道天德电气(山东)有限公司 | Intelligent ring main unit fault real-time monitoring method and system |
CN117436006A (en) * | 2023-12-22 | 2024-01-23 | 圣道天德电气(山东)有限公司 | Intelligent ring main unit fault real-time monitoring method and system |
CN117455725A (en) * | 2023-12-22 | 2024-01-26 | 广惠建设工程集团有限公司 | Building energy consumption management method and system based on BIM (building information modeling) building |
CN117455725B (en) * | 2023-12-22 | 2024-03-12 | 广惠建设工程集团有限公司 | Building energy consumption management method and system based on BIM (building information modeling) building |
CN117609813A (en) * | 2024-01-23 | 2024-02-27 | 山东第一医科大学附属省立医院(山东省立医院) | Intelligent management method for intensive patient monitoring data |
CN117609813B (en) * | 2024-01-23 | 2024-04-23 | 山东第一医科大学附属省立医院(山东省立医院) | Intelligent management method for intensive patient monitoring data |
CN117668684A (en) * | 2024-01-31 | 2024-03-08 | 新风光电子科技股份有限公司 | Power grid electric energy data anomaly detection method based on big data analysis |
CN117668684B (en) * | 2024-01-31 | 2024-04-16 | 新风光电子科技股份有限公司 | Power grid electric energy data anomaly detection method based on big data analysis |
CN117952569A (en) * | 2024-03-27 | 2024-04-30 | 山东省科学院能源研究所 | Public building collaborative energy supply management system based on multisource renewable energy sources |
CN117977717A (en) * | 2024-04-01 | 2024-05-03 | 国网黑龙江省电力有限公司佳木斯供电公司 | Cold region wind-solar-thermal energy storage comprehensive energy collaborative management method and system |
CN117977717B (en) * | 2024-04-01 | 2024-06-11 | 国网黑龙江省电力有限公司佳木斯供电公司 | Cold region wind-solar-thermal energy storage comprehensive energy collaborative management method and system |
Also Published As
Publication number | Publication date |
---|---|
CN117034043B (en) | 2024-01-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN117034043B (en) | Intelligent building comprehensive energy consumption monitoring method and system based on multi-energy Internet of things | |
CN117093879B (en) | Intelligent operation management method and system for data center | |
CN112699913B (en) | Method and device for diagnosing abnormal relationship of household transformer in transformer area | |
CN117411189B (en) | Monitoring data enhancement method of micro-grid coordination controller | |
CN112800231B (en) | Power data verification method and device, computer equipment and storage medium | |
CN110503153B (en) | Photovoltaic system fault diagnosis method based on differential evolution algorithm and support vector machine | |
CN116955895B (en) | Transformer operation abnormity monitoring method and system | |
CN111738348B (en) | Power data anomaly detection method and device | |
CN117312997B (en) | Intelligent diagnosis method and system for power management system | |
CN116992322B (en) | Smart city data center management system | |
CN115601197B (en) | Abnormal state detection method and device for photovoltaic power station | |
CN111179108A (en) | Method and device for predicting power consumption | |
CN115664038A (en) | Intelligent power distribution operation and maintenance monitoring system for electrical safety management | |
CN117436006B (en) | Intelligent ring main unit fault real-time monitoring method and system | |
CN113127464B (en) | Agricultural big data environment feature processing method and device and electronic equipment | |
CN113094448B (en) | Analysis method and analysis device for residence empty state and electronic equipment | |
CN117237678B (en) | Method, device, equipment and storage medium for detecting abnormal electricity utilization behavior | |
CN117150244B (en) | Intelligent power distribution cabinet state monitoring method and system based on electrical parameter analysis | |
CN111506636A (en) | System and method for analyzing residential electricity consumption behavior based on autoregressive and neighbor algorithm | |
CN117236022A (en) | Training method and application method of residual life prediction model of transformer and electronic equipment | |
CN116561692A (en) | Dynamic update real-time measurement data detection method | |
CN117575176B (en) | Processing method and system for abnormal value in power data | |
CN110942196B (en) | Predicted irradiation correction method and device | |
CN117791626B (en) | Intelligent comprehensive power box power supply optimization method | |
CN118052540A (en) | Intelligent management method and system for equipment maintenance scheme |
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 |