CN111695792B - Subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering - Google Patents
Subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering Download PDFInfo
- Publication number
- CN111695792B CN111695792B CN202010473632.5A CN202010473632A CN111695792B CN 111695792 B CN111695792 B CN 111695792B CN 202010473632 A CN202010473632 A CN 202010473632A CN 111695792 B CN111695792 B CN 111695792B
- Authority
- CN
- China
- Prior art keywords
- attribute
- data
- cluster
- energy consumption
- clustering
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000005265 energy consumption Methods 0.000 title claims abstract description 104
- 238000005286 illumination Methods 0.000 title claims abstract description 58
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 57
- 238000004458 analytical method Methods 0.000 title claims abstract description 28
- 239000011159 matrix material Substances 0.000 claims abstract description 29
- 239000013598 vector Substances 0.000 claims abstract description 24
- 230000009467 reduction Effects 0.000 claims abstract description 11
- 230000008859 change Effects 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 39
- 238000012545 processing Methods 0.000 claims description 10
- 238000012544 monitoring process Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 description 7
- 230000005856 abnormality Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004134 energy conservation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] 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
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/80—Management or planning
- Y02P90/82—Energy audits or management systems therefor
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Educational Administration (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Probability & Statistics with Applications (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Circuit Arrangement For Electric Light Sources In General (AREA)
Abstract
The invention relates to a subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering, which comprises the following steps: the energy consumption data of the lighting system are corresponding to the respective equipment attributes to form an energy consumption analysis multi-attribute value system; converting each non-numerical illumination system attribute into a numerical value, normalizing each attribute, converting the attribute into an attribute value, and establishing a vector of each attribute value to all data; reducing the dimension of a multi-attribute value system of the data, and carrying out Kmeans clustering on the data matrix after the dimension reduction; sequentially comparing the distances from each main component to each cluster center in the rest non-clustering operation data, adding the distances from the main attribute of each lighting device to each cluster, and distributing the data to the cluster closest to the cluster center to obtain k class clusters; for each cluster, calculating the mean value of each principal component of all points in the cluster, and taking the mean value of the principal components as a mass center until the change of the center of the cluster tends to be stable, so as to form k clusters; based on the k clusters formed, a result analysis is performed.
Description
Technical Field
The invention relates to the field of rail transit, in particular to a subway station lighting system abnormal energy consumption analysis method based on multi-attribute clustering.
Background
With the rapid development of the urban process in China, urban rail transit construction is rapidly developed. The subway generated by the era is gradually an important transportation means for citizens to travel, and huge energy consumption is brought. The subway mainly uses electricity as energy, but the power consumption of the subway is extremely high, such as Beijing, shanghai and Shenzhen, the power consumption of the subway accounts for more than 30% of the operation cost, the annual power consumption is about 1 hundred million KWh, and the electric charge cost is huge. Independent lighting systems are arranged at subway stations and intervals to provide convenient and comfortable light sources for passengers and staff. According to a series of related policy regulations such as GB50034-2004 building lighting design standard, GB50016-2006 building design fireproof standard, GB/T162752008 urban rail transit lighting and the like, the subway lighting system is definitely specified to account for 14.2-16.1% of the average equipment load of a subway station, and in actual cases, the subway lighting system accounts for 20-30% of the average equipment load of the subway station, and the first-line urban ratio is more obvious. Although subway systems have been developed to an intelligent stage, the problem of excessively large energy consumption ratio of the lighting system is still not well solved.
The load level, the function, the power supply and distribution mode, the power consumption, the power supply time and the like of the lighting system equipment are basic attributes influencing the energy consumption of the lighting system, and are basic elements providing comfortable light sources for passengers and staff, ensuring the normal illumination of subways and timely processing the occurrence of emergency conditions. Therefore, the abnormal energy consumption of the subway illumination system, especially the abnormality of the section work illumination, the safety illumination and the emergency illumination, is found in time, and is an important task for guaranteeing the normal operation of the subway.
However, from the view point of the current equipment and equipment technical conditions of subway stations, in the operation and maintenance of a lighting system, most of the equipment and equipment technical conditions are based on periodic manual detection and random manual monitoring, so that the energy consumption abnormality judgment has a certain limitation, on one hand, the timely discovery of equipment energy consumption abnormality cannot be guaranteed, on the other hand, specific equipment with abnormal energy consumption is difficult to accurately locate, and the time is long, so that the equipment has no actual energy saving effect, a specific effective algorithm is used for analyzing historical lighting energy consumption from the background, abnormal energy consumption is found, and the equipment is controlled in a targeted manner, so that the energy saving effect is obvious, and the operation and maintenance cost can be reduced.
Abnormal energy consumption refers to data that is completely independent in all energy consumption data sets, and these energy consumption data are not randomly biased but are generated by completely different mechanisms. The energy consumption abnormality of the subway illumination system has the characteristics of randomness, complexity and periodicity. The abnormal energy consumption of the subway illumination system is highly uncertain, and the problem that the energy consumption abnormal condition of the illumination system is difficult to accurately detect and the accurate abnormal energy consumption detection mathematical model is difficult to establish in the subway operation process exists.
Anomaly detection is widely used in a variety of application fields, including disease monitoring, equipment status monitoring, building energy consumption, network intrusion, and the like. Monitoring methods for energy consumption anomalies include regression, k-nearest neighbor (KNN), entropy, and clustering. Clustering is the most commonly used method for anomaly monitoring, and is very sensitive to outliers in the data. The class of the clustering algorithm may be defined in a classification based on the devices of the lighting system, but not with accuracy; the clustering center is randomly selected, so that the time complexity of abnormal energy consumption judgment is increased, and the situation is more serious under the condition of large data volume. In addition, for the subway illumination system, due to different attributes among devices, the abnormal state of the subway illumination system cannot be simply detected from the energy consumption value, and abnormal energy consumption analysis is performed by integrating multiple attributes.
Disclosure of Invention
The invention aims to provide a subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering, which is used for solving the problems in the prior art.
The invention discloses a subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering, which comprises the following steps: energy consumption data of lighting systemCorresponding to the respective equipment attributes, an energy consumption analysis multi-attribute value system A= { a is formed 1 ，a 2 ，...，a i Wherein i represents the number of lighting system attributes, a i Representing an attribute name; converting each non-numerical illumination system attribute into a numerical value, normalizing each attribute, converting the attribute into an attribute value, and establishing a vector of each attribute value to all data; reducing the dimension of a multi-attribute value system of data, constructing a principal component parameter system which comprises a plurality of principal component data, determining the number k of clusters according to an inflection point method, selecting k pieces of equipment data as cluster centers of Kmeans initial clusters, and performing Kmeans clustering on a data matrix after the dimension reduction; sequentially comparing the distances from each main component to each cluster center in the rest non-clustering operation data, adding the distances from the main attribute of each lighting device to each cluster, and distributing the data to the cluster closest to the cluster center to obtain k class clusters; for each cluster, calculating the mean value of each principal component of all points in the cluster, and taking the mean value of the principal components as a mass center until the change of the center of the cluster tends to be stable, so as to form k clusters; based on the k clusters formed, a result analysis is performed.
According to an embodiment of the subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering, the illumination system attributes comprise: device model Type, device Power P, load Level, power supply and distribution mode Power, energy consumption EC, time Time and service life Age.
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, the analysis according to the result comprises the following steps: the data which is not distributed in any cluster is abnormal data, and the energy consumption of the corresponding equipment is abnormal; and the abnormal equipment is brought into the daily important monitoring range.
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, each non-numerical illumination system attribute is converted into a numerical type, wherein setting a certain numerical value to represent each type of the non-numerical illumination system attribute, and correspondingly converting each type into a numerical value.
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering calculates average values of principal components of all points in a cluster and takes the average values of the principal components as mass centers, and comprises the following steps: after the primary division, a plurality of pieces of data belong to the same cluster, each cluster is taken as one piece of data, each attribute column is averaged to convert the plurality of pieces of data into one piece of data, new data is formed until the change of the cluster center tends to be stable, and k clusters are formed.
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, the converting of the non-numerical attribute into the numerical attribute comprises: the first level, the second level and the third level are respectively corresponding to 1,2 and 3, and the centralized power supply, the distributed power supply and the mixed power supply of the power supply and distribution mode are respectively corresponding to 0,1 and 2.
According to one embodiment of the subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering, normalization processing is carried out on each attribute data, vectors of each attribute on all data are established, and an illumination system equipment multi-attribute matrix M is established ni The normalization processing formula includes:
wherein Value represents a certain attribute Value corresponding to each device of the energy consumption data, value represents a Value corresponding to the device under the attribute, and all devices have a plurality of values Value under the attribute min The minimum Value, of all values in the representation max Represents the maximum of all values in the attribute column and n represents the lighting system device number.
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, dimension reduction is performed on a multi-attribute value system of data, and a principal component matrix is constructed, wherein the principal component matrix comprises the following components:
1) Assuming an illumination system having n devices, each device having i attributes, constructing an n x i original moment based on the obtained vectors for all data for each attributeArray M n×i Each row represents a lighting device, each column represents an attribute, and the average vector of each attribute of the original matrix is calculated
2) Solving an original matrix M n×i Covariance matrix S of (2) n×n ＝Cov(X)；
3) Solving eigenvalues and eigenvectors of the covariance matrix;
4) Calculating the contribution rate and the accumulated contribution rate of principal components, taking j principal components with the accumulated contribution rate of more than 90% to form new attribute data, and forming a principal component parameter system MC= { MC 1 ，mc 2 ，...，mc j }。
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, kmeans clustering is performed on a data matrix after dimension reduction, and the method comprises the following steps: determining the number k of clusters according to an inflection point method; each piece of data of all the devices is regarded as a node, one piece of data is randomly selected as a first cluster center point, the data which is least similar to the piece of data is selected as a second cluster center point, a plurality of the points which are least similar to the first two pieces of data are selected as a third cluster center point, and the like until k cluster center points are selected; dividing and extracting principal component data in principal component columns to form a principal component column vector space, clustering the principal component column vector space, and gathering similar principal components into the same class; sequentially comparing distances from each principal component in the residual data to each cluster center, adding the distances from the principal attribute of each lighting device to each cluster, distributing the data to the cluster of the cluster center closest to the cluster center to obtain k class clusters, and for each cluster, calculating the average value of each principal component of all points in the cluster and taking the average value of the principal components as a centroid; until the change of cluster center tends to be stable, forming final k clusters and isolated points.
According to an embodiment of the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, kmeans clustering is performed on a data matrix after dimension reduction, and the method comprises the following steps: and in the principal component parameter system, the data principal component data in each attribute column is divided into principal component columns, extracted to form a principal component column vector space, and the principal component column vector space is clustered to gather similar principal components in the same class.
According to the method for analyzing the abnormal energy consumption of the subway illumination system based on the multi-attribute clustering, the energy consumption data of the illumination system in the subway operation process and the energy consumption attributes of all illumination devices can be acquired in real time through the energy management system of a rail transit company, and the abnormal energy consumption behavior in the operation process can be detected through the main attribute clustering analysis of the illumination system, so that the accuracy and timeliness of detecting the abnormal energy consumption of the subway illumination system in the prior art are improved, energy conservation and emission reduction are realized for the subway operation company, and benefits are increased.
Drawings
Fig. 1 is a flowchart of an embodiment of a method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering.
Detailed Description
For the purposes of clarity, content, and advantages of the present invention, a detailed description of the embodiments of the present invention will be described in detail below with reference to the drawings and examples.
The invention discloses a subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering, which comprises the following steps: the energy consumption data of the lighting system are corresponding to the respective equipment attributes to form an energy consumption analysis multi-attribute value system A= { a 1 ，a 2 ，...，a i -i represents the number of attributes of the lighting system, -a represents that the lighting system energy consumption comes from a plurality of devices, each device being affected by aspects, the aspects constituting an attribute set of energy consumption data, -a i The attribute names of the device are represented, such as a device model, a device Power and the like, and specifically comprise a device model Type, a device Power P, a load Level, a Power supply and distribution mode Power, energy consumption EC, time Time and service life Age;
converting non-numeric attributes of each piece of data (such as { equipment model, equipment power, load grade, …, energy consumption value }) into numeric attributes, wherein equipment in the lighting system has a plurality of attributes, such as the equipment power is numeric, and if a power supply and distribution mode (centralized power supply, distributed power supply and mixed power supply) is the non-numeric attribute, converting the non-numeric attributes into numeric data, respectively corresponding to (0, 1 and 2), normalizing each attribute data, converting the non-numeric data into an attribute value convenient to process, converting the non-numeric data into numeric data, and normalizing the data; meaning the attribute value. For example, before numerical conversion, {1100w, centralized power, …,50}, after conversion, {1100w,0, …,50}, a vector of each attribute to all data is established;
the dimension of a multi-attribute value system of the data subjected to the normalization processing of the equipment energy consumption multi-attribute data in the lighting system is reduced, namely each piece of energy consumption data consists of a plurality of attribute values, and normalization processing is carried out on each attribute column, so that the condition that the excessive data and the excessive data are small is avoided, the attribute dimension is reduced, and a main component parameter system MC= { MC is constructed 1 ，mc 2 ，...，mc j The energy consumption of the lighting system is influenced by various attributes, but the related attributes are excessive, and the time complexity is increased, so that the attribute with less influence on the energy consumption is deleted in the dimension reduction processing, and the rest attributes have more influence on the energy consumption, so that the main attribute, namely the main component, is formed;
determining the number k of clusters according to an inflection point method;
selecting k energy consumption data of the illumination system after normalization processing and dimension reduction processing as cluster centers of kmeans initial clusters; dividing and extracting data of main components in each attribute column by using the main component column, wherein each attribute column is the main component column, forming a main component column vector space to cluster the main component column vector space, and gathering similar main components into the same class;
the method specifically comprises the following steps: firstly, k pieces of energy consumption data are selected as cluster centers, the data of a main component column are used as a data set, the data set is clustered, and the data set and the data of the cluster centers are classified, so that the data are also in the cluster. For example, the data set is { a, b, c, d, e, f }, where { a, b, c } is the initial cluster center, and the clustering is performed on { a, d, e }, { b, f }, { c }, and d, e and a are in the same cluster.
And comparing distances from each principal component in the residual data (only one attribute column is subjected to clustering operation each time, and the residual data is the attribute columns which are not subjected to clustering) to each clustering center in sequence, wherein one piece of energy consumption data is from a plurality of attributes, and the attribute columns are divided, so that one piece of data is possibly divided into different clusters according to different attribute divisions, and therefore judgment is carried out based on the sum of the distances from all the attributes to each cluster. This calculation from the usual distance formula for a cluster, such as Euclidean distance, the smaller the distance, the greater the likelihood that the data is attributed to this cluster. Adding the distances from the main attribute of each lighting device to each cluster, and distributing data to the clusters closest to the cluster center to obtain k class clusters; through clustering, the energy consumption data of all devices are divided into different classes, and the data in the same class form a set, namely a cluster. Clusters belong to a definition inherent to the clustering algorithm.
For each cluster, calculating the mean value of each principal component of all points in the cluster and taking the mean value of the principal components as a centroid, wherein the method specifically comprises the following steps: and when the clustering is not finished, the clustering is performed for a plurality of times until the partitioning result is stable. After the primary division, a plurality of data belong to the same cluster, at this time, each cluster is taken as one data, and a plurality of clusters are formed, so that new energy consumption data is formed, and aiming at each cluster, the plurality of data in each cluster are converted into one data, namely, each attribute column of the data is averaged, so that new data is formed. Until the change of the cluster center tends to be stable, forming final k clusters;
and finally, analyzing, wherein the data which are not distributed in any cluster are abnormal data, namely, the energy consumption of the corresponding equipment is abnormal.
The equipment is brought into the daily key monitoring range, and is purposefully adjusted and overhauled, so that the equipment can be used normally.
Fig. 1 is a flowchart of an embodiment of a method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering, and as shown in fig. 1, the method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering of the present invention includes:
s101, selecting a subway station operation station from a household energy consumption analysis module of an energy management system of the company, selecting month data, extracting lighting system data from a database, and displaying the lighting system data on a household energy consumption analysis interface; selecting trend chart display on a household energy consumption analysis interface, and intuitively acquiring energy consumption EC of each lighting device in the site on each day md (m represents a certain lighting device, d represents the day of the current month);
s102, acquiring specific attributes of each lighting device of the lighting system, such as power, load level, power supply and distribution modes and the like;
s103, each energy consumption data EC of the lighting system md Corresponding to respective equipment attributes, an energy consumption analysis multi-attribute value system A= { EC, type, P, level, power, time, age … }, wherein the equipment model Type, the equipment Power P, the load Level, the Power supply and distribution mode Power, the Time Time, the service life Age and the like are formed;
s104, analyzing and abstracting multiple attributes of each piece of data, converting non-numerical attributes into numerical attributes, respectively corresponding to (1, 2 and 3) according to levels (primary, secondary and tertiary), respectively corresponding to (0, 1 and 2) in power supply and distribution modes (centralized power supply, distributed power supply and mixed power supply), normalizing the data of each attribute, establishing vectors of each attribute on all the data, and establishing a multi-attribute matrix M of lighting system equipment ni (n represents a lighting system device);
wherein, the normalization processing formula is as follows:
wherein Value represents a certain attribute Value corresponding to each device of the energy consumption data, value represents a Value corresponding to the device under the attribute, and all devices have a plurality of values Value under the attribute min What is shown in the representationWith minimum Value of values max Representing the maximum of all values in the attribute column
S105: PCA dimension reduction is carried out on the multi-attribute value system, and the original multi-attribute value system A= { a is adopted 1 ，a 2 ，...，a i The (i represents the number of attributes of the lighting system) is updated to a principal component parameter system MC= { MC 1 ，mc 2 ，...，mc j And j represents the number of principal components of the lighting system (j < i), and establishes a principal component matrix MC of the lighting system device nj (n represents a lighting system device). The specific process is as follows:
1) Assuming that the illumination system has n devices, each device having i attributes, constructing an n x i original matrix M based on the vectors of all data for each attribute obtained in step 4 n×i Each row represents a lighting device, each column represents an attribute, and the average vector of each attribute of the original matrix is calculated
2) Solving an original matrix M n×i Covariance matrix S of (2) n×n ＝Cov(X)；
3) Solving eigenvalues and eigenvectors of the covariance matrix;
4) Calculating the contribution rate and the accumulated contribution rate of principal components, taking j principal components with the accumulated contribution rate of more than 90% to form new attribute data, and forming a principal component parameter system MC= { MC 1 ，mc 2 ，...，mc j }；
S106: kmeans clustering is carried out on the data matrix after dimension reduction, and a clustering result C= { C is obtained 1 ，C 2 ，...，C k }∪{C e }, wherein C 1 ，C 2 ，...，C k Representing the clustering result, C e Is a set of nodes that do not belong to each cluster, i.e., a set of outliers. The specific flow is as follows:
5) Determining the number k of clusters according to an inflection point method (elbow rule);
6) Selection of initial center device. Each piece of data of all the devices is regarded as a node, one piece of data is randomly selected as a first cluster center point, then the data which is least similar to the piece of data is selected as a second cluster center point, then the most dissimilar number of the data which is the first two pieces of data is selected as a third cluster center point, and so on until k cluster center points are selected;
7) Dividing and extracting principal component data in principal component columns to form a principal component column vector space, clustering the principal component column vector space, and gathering similar principal components into the same class;
8) Sequentially comparing the distances from each main component in the residual data to each cluster center, adding the distances from the main attribute of each lighting device to each cluster, and distributing the data to the cluster of the cluster center closest to the main attribute to obtain k class clusters;
9) And (4) for each cluster, calculating the average value of each principal component of all points in the cluster, taking the average value of the principal components as the mass center, and repeatedly executing the step (4) until the change of the center of the cluster tends to be stable, and forming final k clusters and isolated points, namely the data of abnormal energy consumption of the lighting system, so that the purpose of finding abnormal energy consumption is achieved.
The energy management system acquires energy consumption data in the subway operation process in real time, and establishes a multi-attribute value system of the energy consumption of the lighting system according to the acquired energy consumption data of the lighting system and the characteristics of each lighting device. And (3) reducing the dimension of the multi-attribute value system, establishing a main component parameter system, classifying the main component parameter system by using a clustering algorithm to obtain a final energy consumption analysis result, and finding out abnormal energy consumption. The invention is not only suitable for subway illumination systems, but also suitable for air conditioner energy consumption analysis, water pump energy consumption analysis and the like, and has strong accuracy, applicability and expansibility.
The subway illumination system abnormal energy consumption analysis based on the multi-attribute clustering can accurately find abnormal energy consumption events in real time, determine the occurrence time and specific equipment of the abnormal energy consumption, facilitate maintenance personnel to overhaul in time, and provide reliable guarantees for safe operation of subways, safe traveling of passengers and the like. The invention has strong practicability in the aspect of energy conservation of the rail transit subway. The invention can effectively realize the abnormal energy consumption detection function of the lighting system.
The foregoing is merely a preferred embodiment of the present invention, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present invention, and such modifications and variations should also be regarded as being within the scope of the invention.
Claims (9)
1. A subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering comprises the following steps:
the energy consumption data of the lighting system are corresponding to the respective equipment attributes to form an energy consumption analysis multi-attribute value system A= { a 1 ，a 2 ，...，a i Wherein i represents the number of lighting system attributes, a i Representing an attribute name;
converting each non-numerical illumination system attribute into a numerical value, normalizing each attribute, converting the attribute into an attribute value, and establishing a vector of each attribute value to all data;
reducing the dimension of a multi-attribute value system of data, constructing a principal component parameter system which comprises a plurality of principal component data, determining the number k of clusters according to an inflection point method, selecting k pieces of equipment data as cluster centers of Kmeans initial clusters, and performing Kmeans clustering on a data matrix after the dimension reduction;
sequentially comparing the distances from each main component to each cluster center in the rest non-clustering operation data, adding the distances from the main attribute of each lighting device to each cluster, and distributing the data to the cluster closest to the cluster center to obtain k class clusters; for each cluster, calculating the mean value of each principal component of all points in the cluster, and taking the mean value of the principal components as a mass center until the change of the center of the cluster tends to be stable, so as to form k clusters;
according to the k clusters formed, carrying out result analysis;
wherein the lighting system attributes include: device model Type, device Power P, load Level, power supply and distribution mode Power, energy consumption EC, time Time and service life Age.
2. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering according to claim 1, wherein the analyzing according to the result comprises: the data which is not distributed in any cluster is abnormal data, and the energy consumption of the corresponding equipment is abnormal; and the abnormal equipment is brought into the daily important monitoring range.
3. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering according to claim 1, wherein converting each non-numerical illumination system attribute into a numerical type includes setting a certain numerical value to represent each type of the non-numerical illumination system attribute, and converting each type into a numerical value correspondingly.
4. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering according to claim 1, wherein calculating the mean value of each principal component of all points in the cluster and taking the mean value of the principal components as the centroid comprises: after the primary division, a plurality of pieces of data belong to the same cluster, each cluster is taken as one piece of data, each attribute column is averaged to convert the plurality of pieces of data into one piece of data, new data is formed until the change of the cluster center tends to be stable, and k clusters are formed.
5. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering according to claim 1, wherein converting the non-numerical attribute into a numerical attribute comprises: the first level, the second level and the third level are respectively corresponding to 1,2 and 3, and the centralized power supply, the distributed power supply and the mixed power supply of the power supply and distribution mode are respectively corresponding to 0,1 and 2.
6. The method for analyzing abnormal energy consumption of subway illumination system based on multi-attribute clustering as set forth in claim 1, wherein each attribute is normalized to all data, and each attribute is establishedIs used for establishing a multi-attribute matrix M of the lighting system equipment ni The normalization processing formula includes:
wherein Value represents a certain attribute Value corresponding to each device of the energy consumption data, value represents a Value corresponding to the device under the attribute, and all devices have a plurality of values Value under the attribute min The minimum Value, of all values in the representation max Represents the maximum of all values in the attribute column and n represents the lighting system device number.
7. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering according to claim 1, wherein the step of reducing the dimension of the multi-attribute value system of the data and constructing the principal component matrix comprises the steps of:
1) Assuming an illumination system having n devices, each device having i attributes, constructing an n x i original matrix M based on the obtained vectors of each attribute for all data n×i Each row represents a lighting device, each column represents an attribute, and the average vector of each attribute of the original matrix is calculated
2) Solving an original matrix M n×i Covariance matrix S of (2) n×n ＝Cov(X)；
3) Solving eigenvalues and eigenvectors of the covariance matrix;
4) Calculating the contribution rate and the accumulated contribution rate of principal components, taking j principal components with the accumulated contribution rate of more than 90% to form new attribute data, and forming a principal component parameter system MC= { MC 1 ，mc 2 ，...，mc j }。
8. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering as set forth in claim 1, wherein performing Kmeans clustering on the reduced-dimension data matrix includes:
determining the number k of clusters according to an inflection point method;
each piece of data of all the devices is regarded as a node, one piece of data is randomly selected as a first cluster center point, the data which is least similar to the piece of data is selected as a second cluster center point, a plurality of the points which are least similar to the first two pieces of data are selected as a third cluster center point, and the like until k cluster center points are selected;
dividing and extracting principal component data in principal component columns to form a principal component column vector space, clustering the principal component column vector space, and gathering similar principal components into the same class;
sequentially comparing distances from each principal component in the residual data to each cluster center, adding the distances from the principal attribute of each lighting device to each cluster, distributing the data to the cluster of the cluster center closest to the cluster center to obtain k class clusters, and for each cluster, calculating the average value of each principal component of all points in the cluster and taking the average value of the principal components as a centroid; until the change of cluster center tends to be stable, forming final k clusters and isolated points.
9. The method for analyzing abnormal energy consumption of a subway illumination system based on multi-attribute clustering as set forth in claim 1, wherein performing Kmeans clustering on the reduced-dimension data matrix includes: and in the principal component parameter system, the data principal component data in each attribute column is divided into principal component columns, extracted to form a principal component column vector space, and the principal component column vector space is clustered to gather similar principal components in the same class.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010473632.5A CN111695792B (en) | 2020-05-29 | 2020-05-29 | Subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010473632.5A CN111695792B (en) | 2020-05-29 | 2020-05-29 | Subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111695792A CN111695792A (en) | 2020-09-22 |
| CN111695792B true CN111695792B (en) | 2023-08-01 |
Family
ID=72478763
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010473632.5A Active CN111695792B (en) | 2020-05-29 | 2020-05-29 | Subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111695792B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112131388B (en) * | 2020-09-28 | 2024-02-06 | 范馨月 | Abnormal data detection method containing text data types |
| CN116893297B (en) * | 2023-09-11 | 2024-01-12 | 常州旭泰克系统科技有限公司 | Method and system for monitoring energy consumption of rotating equipment |
| CN116991219B (en) * | 2023-09-26 | 2024-01-12 | 北京国电通网络技术有限公司 | Abnormal production state monitoring method, device, electronic equipment and medium |
| CN118096120B (en) * | 2024-03-04 | 2024-09-13 | 江苏丰登光电科技有限公司 | Data real-time monitoring method and system based on multiple data collection |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102789599A (en) * | 2012-07-06 | 2012-11-21 | 西北工业大学 | Operation shop bottleneck recognition method based on cluster analysis and multiple attribute decision making |
| CN103913630A (en) * | 2014-04-18 | 2014-07-09 | 上海电科智能系统股份有限公司 | Centralized electric energy measuring and energy consumption analyzing system for rail transit |
| CN104570083A (en) * | 2013-10-29 | 2015-04-29 | 中国石油天然气集团公司 | Multi-dimensional seismic attribute-based automatic geologic body identification method |
| CN105446139A (en) * | 2015-12-18 | 2016-03-30 | 华南理工大学 | Construction energy consumption analysis method and construction energy consumption analysis system based on BP neural network |
| CN105764162A (en) * | 2016-05-10 | 2016-07-13 | 江苏大学 | Wireless sensor network abnormal event detecting method based on multi-attribute correlation |
| CN105825323A (en) * | 2016-03-10 | 2016-08-03 | 山东建筑大学 | Building energy consumption main influence factor analysis method based on big data |
| CN106570581A (en) * | 2016-10-26 | 2017-04-19 | 东北电力大学 | Attribute association based load prediction system and method in energy Internet environment |
| CN108304978A (en) * | 2018-05-08 | 2018-07-20 | 国网江西省电力有限公司经济技术研究院 | A kind of mid-term Electric Power Load Forecast method based on data clusters theory |
| CN108549754A (en) * | 2018-03-29 | 2018-09-18 | 中交隧道局电气化工程有限公司 | Tramcar system energy consumption analysis method based on control figure |
| CN109214522A (en) * | 2018-07-17 | 2019-01-15 | 西安交通大学 | One kind being based on multiattribute equipment performance degradation assessment method |
| CN109858832A (en) * | 2019-02-25 | 2019-06-07 | 合肥工业大学 | A kind of more attribute green Index grading Coordination Evaluation methods |
| CN110727257A (en) * | 2019-08-27 | 2020-01-24 | 华润置地控股有限公司 | Equipment operation diagnosis method and device based on K-means clustering algorithm |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060259360A1 (en) * | 2005-05-16 | 2006-11-16 | Manyworlds, Inc. | Multiple Attribute and Behavior-based Advertising Process |
-
2020
- 2020-05-29 CN CN202010473632.5A patent/CN111695792B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102789599A (en) * | 2012-07-06 | 2012-11-21 | 西北工业大学 | Operation shop bottleneck recognition method based on cluster analysis and multiple attribute decision making |
| CN104570083A (en) * | 2013-10-29 | 2015-04-29 | 中国石油天然气集团公司 | Multi-dimensional seismic attribute-based automatic geologic body identification method |
| CN103913630A (en) * | 2014-04-18 | 2014-07-09 | 上海电科智能系统股份有限公司 | Centralized electric energy measuring and energy consumption analyzing system for rail transit |
| CN105446139A (en) * | 2015-12-18 | 2016-03-30 | 华南理工大学 | Construction energy consumption analysis method and construction energy consumption analysis system based on BP neural network |
| CN105825323A (en) * | 2016-03-10 | 2016-08-03 | 山东建筑大学 | Building energy consumption main influence factor analysis method based on big data |
| CN105764162A (en) * | 2016-05-10 | 2016-07-13 | 江苏大学 | Wireless sensor network abnormal event detecting method based on multi-attribute correlation |
| CN106570581A (en) * | 2016-10-26 | 2017-04-19 | 东北电力大学 | Attribute association based load prediction system and method in energy Internet environment |
| CN108549754A (en) * | 2018-03-29 | 2018-09-18 | 中交隧道局电气化工程有限公司 | Tramcar system energy consumption analysis method based on control figure |
| CN108304978A (en) * | 2018-05-08 | 2018-07-20 | 国网江西省电力有限公司经济技术研究院 | A kind of mid-term Electric Power Load Forecast method based on data clusters theory |
| CN109214522A (en) * | 2018-07-17 | 2019-01-15 | 西安交通大学 | One kind being based on multiattribute equipment performance degradation assessment method |
| CN109858832A (en) * | 2019-02-25 | 2019-06-07 | 合肥工业大学 | A kind of more attribute green Index grading Coordination Evaluation methods |
| CN110727257A (en) * | 2019-08-27 | 2020-01-24 | 华润置地控股有限公司 | Equipment operation diagnosis method and device based on K-means clustering algorithm |
Non-Patent Citations (2)
| Title |
|---|
| 基于多属性决策TOPSIS的簇头选择;王林;《计算机工程》;第45卷(第01期);第45-49、54页 * |
| 应用GIS的城市照明规划系统;边宇;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》(第04期);第C038-51页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111695792A (en) | 2020-09-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111695792B (en) | Subway illumination system abnormal energy consumption analysis method based on multi-attribute clustering | |
| CN110110870B (en) | Intelligent equipment fault monitoring method based on event map technology | |
| CN111429027B (en) | Regional power transmission network operation multidimensional analysis method based on big data | |
| CN103337043B (en) | The method for early warning of electric power communication device running status and system | |
| CN107886235A (en) | A kind of Fire risk assessment method for coupling certainty and uncertainty analysis | |
| CN106022592A (en) | Power consumption behavior anomaly detection and public security risk early warning method and device | |
| CN112488371B (en) | Pipeline intelligent early warning method and system based on big data | |
| CN104038375A (en) | Alarm processing analysis system and method of broadcasting and TV access network | |
| CN116090819A (en) | Power distribution network risk situation prediction method based on association rule | |
| CN115796390A (en) | Data-driven building energy consumption abnormity diagnosis method and system | |
| CN103914482B (en) | Centralized Monitoring event influence property based on CMDB determines method | |
| CN108921438B (en) | Power distribution network regulation and control management weak link identification method based on cascade weight | |
| CN116365519B (en) | Power load prediction method, system, storage medium and equipment | |
| CN108446202A (en) | A kind of judgment method of the safe condition of calculator room equipment | |
| CN112783939A (en) | Low-voltage distribution network running state evaluation method based on data mining | |
| CN104865959B (en) | A kind of fault self-diagnosis method of fire-fighting power supply control system | |
| CN109193624B (en) | Load classification method based on improved K-means clustering | |
| CN113052250A (en) | Decision support method, system, device and medium based on meteorological disaster | |
| CN112991093A (en) | Electricity stealing detection method and system based on edge calculation | |
| Chumnanvanichkul et al. | Three-level Classification of Air Conditioning Energy Consumption for Building Energy Management System Using Data Mining Techniques | |
| CN111651493A (en) | Full life cycle forward guidance and feedback optimization management method of passenger lifting equipment | |
| Zhang et al. | [Retracted] Substation Operation Information Maintenance Based on Intelligent Data Mining | |
| CN114399177B (en) | Scheduling treatment rule mining and generating method and system based on Apriori | |
| CN115659832B (en) | Enterprise operation analysis and early warning monitoring method and system based on big data analysis | |
| Zhao et al. | Research on machine learning-based correlation analysis method for power equipment alarms |
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 |