CN106570729A - Air conditioner reliability influence factor-based regional clustering method - Google Patents
Air conditioner reliability influence factor-based regional clustering method Download PDFInfo
- Publication number
- CN106570729A CN106570729A CN201610996880.1A CN201610996880A CN106570729A CN 106570729 A CN106570729 A CN 106570729A CN 201610996880 A CN201610996880 A CN 201610996880A CN 106570729 A CN106570729 A CN 106570729A
- Authority
- CN
- China
- Prior art keywords
- air
- conditioning
- region
- provinces
- 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.)
- Pending
Links
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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses an air conditioner reliability influence factor-based regional clustering method. The method includes the following steps that: a system analyzes the regional differences of working environment factors and user use habit factors that influence the reliability of air conditioners and extracts working environment reliability key influence factors and user use habit reliability key influence factors; an air conditioner reliability influence factor-based regional clustering analysis comprehensive evaluation model is constructed; judgment criterion of air conditioner startup refrigeration and heating are formulated, the average consumption tendency indexes of the air conditioners are accurately quantified; and a weighted Ward clustering algorithm is adopted to carry out clustering analysis in the aspects of the working environment influence factors and the user use habit factor influence factors, so that an working environment influence factor clustering analysis result and a user use habit factor influence factor clustering analysis result are obtained, a secondary clustering method is adopted to integrate the working environment influence factor clustering analysis result and the user use habit factor influence factor clustering analysis result, so that final regional distribution can be obtained. With the air conditioner reliability influence factor-based regional clustering method of the invention adopted, the use reliability difference of air conditioners distributed in different areas can be minimum, and more scientific and more precise regional classification results can be obtained.
Description
Technical field
The present invention relates to a kind of region clustering method, concretely relate to a kind of based on air-conditioning reliability effect factor
Region clustering method.
Background technology
In increasingly competitive air-conditioning market, reliability becomes the core index and enterprise for weighing air-conditioning qualitative attribute
Obtain the key factor of product competition advantage.In recent years, the research of air-conditioning reliability has been increasingly subject to theoretical circles and industrial quarters
Pay attention to, these research work are concentrated mainly on failure analysis, Reliability evaluation and the reliability of air-conditioning parts and system
The aspects such as property influence factor.Due to the difference of the factors such as the natural environmental condition of different geographic regions, economic situation and living habit
It is different so that working condition residing for air-conditioning and use time are inconsistent, there is larger difference using reliability so as to cause which.Cause
This, from affect air-conditioning reliability factor angle, science sophisticated category is carried out to geographic area, contribute to more precisely
Hold and be distributed in the air-conditioning of different geographic regions and use reliability level, so as to make more efficiently product design, manufacture, sale
With the decision-making of after-sales service strategy.
It is with regard to air conditioning system reliability effect factorial analysiss, main at present to affect empty by the means researchs such as experiment and checking
Adjust the correlative factor or condition using reliability, for example, Yau etc.[1]From refrigeration and heating load, power consumption and outdoor design
Three aspect analysis impacts of the climate change to the Performance And Reliability of air conditioning system of condition；Li and Yau etc.[2][3]Respectively to the summer
The hot solar airconditioning of winter cryogenic region, the SPLIT AIR-CONDITIONING SYSTEM of torrid areas are tested, and analysis climate change is to air-conditioning system
The impact of cold, heating performance.From the point of view of the research of existing air conditioning system reliability effect factor, reliability effect factor is analyzed
It is extremely limited with the relatedness of area differentiation and the research for being applied to relevant range cluster analyses, only document【Liu Weidong, Song Hao
Wei, Zhao Zhiwei, Li Jie. the air-conditioning reliability assessment [J] based on cluster analyses and time concentration. Industrial Engineering and Management, 2013,
18(4)：156-160.】In analyze impact of the humiture weather conditions with regional differentiation to air-conditioning reliability, and should
For the cluster of the provincial capital of Subtropical China monsoon climatic region.
The classification of geographic area generally adopts clustering method, mainly according to specific cluster analyses angle survey region
Difference, according to the difference for choosing basic data model form, is divided into the cross-section data based on sample and the panel number based on sample
According to two big class of clustering method.Traditional cluster analyses object is usually the Different Individual cross-section data of single fixed time period,
For example, the basic model based on sample cross-section data carries out region division, Yu etc.[4]Using fuzzy clustering method to Chinese 30
Provinces and cities' CO2 discharge characteristicss are sorted out；Saracli[5]Several clustering methods and distance metric mode of analysis contrast hierarchical clustering
Various combination result, and Clustering Effect optimal Ward methods are applied to into osmanli region division.Said method mainly according to
According to two-dimensional section data, the loss of data message is easily caused, dynamic time sequence feature is also easily ignored, it is difficult to mining data behind
Profound information, therefore tend not to meet the needs of people's problem analysis.In recent years, occur in that according to a long term
Panel data carries out cluster analyses.Panel data is with can accommodate multi objective, to consider index active development feature etc. excellent
Feature, for example, the basic model based on panel data carries out territorial classification, Iyigun etc.[6]Based on temperature, precipitation and wet
Turkey is divided into 14 different climatic provinces using Ward methods by degree time sequential value；Document【Li Yinguo, Dai Yi, He Xiao
Group. the panel data clustering method [J] based on adaptive weighting. the system engineering theory and practice, 2013,33 (2)：388-
395.】The distance function and Ward functions of reconstruction panel data, using adaptive weighted clustering method to selected areas of China
Innovation ability carry out region division.From the point of view of current present Research, with regard to air-conditioning reliability effect factor regional differentiation
Research is still blank, also lacks the region clustering based on reliability effect factor regional differentiation and studies.
The content of the invention
The technical problem to be solved is to overcome the deficiencies in the prior art, there is provided a kind of to be based on air-conditioning reliability
The region clustering method of influence factor, is obtained in that more scientific, fine geographic area classification, so as to more precisely hold point
Cloth uses reliability level in the air-conditioning of different geographic regions.
The present invention employs the following technical solutions solution above-mentioned technical problem.A kind of area based on air-conditioning reliability effect factor
Domain clustering method, its step are as follows：
Step A, systematic analysiss affect the area difference of the Working Environments and user's use habit factor of air-conditioning reliability
The opposite sex, and two class reliability key influence factor of its working environment and user's use habit is extracted respectively；
Step B, using the corresponding region clustering analysis indexes obtained by step A, build based on air-conditioning reliability shadow
The comprehensive evaluation model of the region clustering analysis of the factor of sound；
Step C, the decision criteria for making air-conditioning start cooling and warming, accurately quantify air-conditioning average propensity to consume and refer to
Mark；
Step D, using the weight of each distance of sum of deviation square method objective computation, then with weighting Ward methods respectively from work
Making environment, user's use habit this two class influence factor carries out cluster analyses, obtains two kinds of classification region clustering results, then adopts
Synthesis is carried out to two kinds of classification region clustering results with secondary clustering procedure, so as to obtain final area distribution.
Preferably, the key influence factor of the working environment, refers to that selection includes temperature, humidity, sunshine and precipitation
Factor of natural environment；And using clustering target and measure be defined, wherein：
(1) temperature (DEG C)：With each province (city) all of region administrative unit mean temperature metric calculation monthly：
In formula：TMiptFor the p provinces t mean temperatures of i-th month, NpRepresent that p provinces include region administrative unit
Number, i represent month, the t expressions of years, xjiptFor j-th of the p provinces region administrative unit t mean temperatures of i-th month；
(2) relative humidity (%)：With each province (city) all of region administrative unit medial humidity metric calculation monthly：
In formula：HMiptFor the p provinces t medial humidities of i-th month, yjiptFor j-th of p provinces region administrative unit
The t medial humidities of i-th month；
(3) sunshine (h)：With each province (city) all of region administrative unit average sunshine time metric calculation monthly：
In formula：SMiptFor the p provinces t average sunshine times of i-th month, ZjiptIt is administrative single for j-th of p provinces region
The position t average sunshine times of i-th month；
(4) precipitation (mm)：Using each province (city) all of region administrative unit monthly (evening 8：00- evenings next day 8：00) when
Average precipitation metric calculation：
In formula：RMiptFor i-th month (evening 8 of p provinces t：00- evenings next day 8：00) average precipitation, UjiptSave for p
J-th of part (evenings 8 of region administrative unit t i-th month：00- evenings next day 8：00) average precipitation.
Preferably, the key influence factor of user's use habit, be converted into analysis natural environment, economic condition and
Impact of the user psychology adaptability to user's use habit.
Preferably, the natural environment includes temperature, humidity, sunshine and wind speed environments factor；And the clustering target for using
And measure is defined, wherein：
Temperature, humidity and sunshine are affected on human thermal comfort using season meansigma methodss quantitative measurement, and computing formula and (1)-
(3) formula is consistent, and now the unit of i is season, season meansigma methodss can be calculated by correspondence month value statistical average；
Mean wind speed metric calculation of the wind speed (m/s) using each province (city) all of region administrative unit quarterly：
In formula：VJkptFor the mean wind speed in p provinces t k seasons, WjiptFor j-th of p provinces region administrative unit
The mean wind speed in t k seasons.
Preferably, the economic condition includes two economic indicators of the total retail sales of consumer goods and average propensity to consume,
Influence degree of the level of economic development to user's use habit is weighed by above-mentioned economic indicator；And the clustering target that uses and degree
Amount method is defined, wherein：
Average propensity to consume tolerance of the average propensity to consume (%) using average each household quarterly, i.e., average each household is quarterly
The overall consumption expenditure of air-conditioning accounts for the ratio of average each household quarterly disposable income；Average propensity to consume is defined as：
In formula：APCkptFor air-conditioning average propensity to consume of the p provinces t annuals per user's kth season；ACkptFor p provinces
The disposable income in average per-user family t kth seasons；GCkptFor sky of the p provinces t annuals per user's kth season
Fixed consumption expenditure is adjusted, the price of one-time payment is converted in each season plus disposable mounting cost when being bought by user
In；FCkptAir-conditioning for p provinces average per-user t kth seasons uses the consumption expenditure, as producer typically has 6 years or so
Free repair guarantee, air-conditioning kth season cost of use can be simplified to kth season using produce the electricity charge, i.e.,
FCkpt=Mpt×Tkpt×Pk×Fpt (7)
In formula：MptIt is the every subscriber household air-conditioning owning amount of p provinces t annuals；PkFor the refrigeration or system in air-conditioning kth season
Heat consumption electrical power；FptFor the electricity price of p provinces t；TkptFor the available machine time of p provinces t kth season air-conditionings：
Tkpt=dkpt×tp (8)
In formula：dkptFor the start natural law in p provinces t kth seasons；tpAveragely heat daily or freeze for p provinces air-conditioning
Start duration.
Preferably, impact of the user psychology adaptability to user's use habit, that is, refer to NORTH CHINA area winter
Cold, carries out centralized and unified heating mode mostly, does not partly have central heating but the subscriber household in cold district tends to
Replace warming by air conditioner using other heating systems such as solar heating, burning coal heatings；SOUTHERN CHINA, middle part air conditioner user it is then general
All over tending to using warming by air conditioner mode.
Preferably, the decision criteria of air-conditioning start cooling and warming is made described in step C, in referring to judge 1 year
On the cooling and warming date of needs, each cooling and warming natural law is calculated in season, and determines that the air conditioner user of different regions is averagely every
Its start refrigeration, the duration for heating, so as to the air-conditioning available machine time of different regions various quarters in accurate calculating formula (8), the amount of obtaining
The air-conditioning average propensity to consume of change.
Preferably, in step D with weighting Ward clustering algorithms respectively from working environment, user's use habit this
When two class influence factors carry out cluster analyses, the similarity measure function of multi objective panel data covers absolute magnitude apart from dij
(AQED), speedup is apart from dij(ISED) and fluctuation apart from dij(VCED), represented between individuality using the weighted array of three kinds of distances
Comprehensive distance, i.e. Dij(CED)=α*·dij(AQED)+β*·dij(ISED)+γ*·dij(VCED), wherein α *, β * and γ *
Three kinds are represented respectively apart from corresponding weight coefficient, and meet α *+β *+γ *=1.
Preferably, described three kinds apart from corresponding weight coefficient, calculates according to the sum of deviation square of three kinds of distances, specifically such as
Under：
In formula：R is the root-mean-square standard deviation of the l time and class, and ν is the individual dimension of observation, and p is class LlIn observation
Individual number, L (AQED), L (ISED), L (VCED) are represented respectively and individually sample to be analyzed are clustered with three kinds of distances
Sum of deviation square in the whole class of analysis.
Preferably, synthesis is carried out to two kinds of region clustering result using secondary clustering procedure described in step D,
Concrete grammar is：First by natural environment influence factor region clustering result, user's use habit influence factor's region clustering result
Effectively combine, be under the jurisdiction of of a sort multiple geographic areas and be considered initial classes, the geographic area of belonging kinds cannot be determined
It is considered not sort out, independent groups of geographic area is considered separate class；After category label is finished, to comprising two and with Shangdi
The classification in reason region solves the barycenter of cluster centre, i.e. classification, all categories is gathered again with the sum of deviation square method of weighting
Class, both between class distances obtain the general area distribution situation after secondary cluster with balance for analysis.
Compared to existing technology, the invention has the advantages that：
1st, to the classification of geographic area residing for air-conditioning more science, fine；
2nd, using the quantitative model of scientific and reasonable clustering target, shadow of each factor to air-conditioning reliability has been measured exactly
The degree of sound.The index of cluster analyses is carried out to Provincial administrative unit division from belonging to Provincial administrative unit division
Region administrative unit statistical data is averagely obtained, and generally represents each province with main cities or provincial capital compared in existing research
Part, zone sample is not covered by the processing method in area in all parts of the country, more accuracy；The air-conditioning start of formulation freezes, heats judgement
Criterion, truly reflecting user's various quarters freezes, heats natural law, being capable of precise quantification air-conditioning average propensity to consume.
3rd, the complex nature of the problem is reduced using the secondary Clustering Comprehensive model in region, improve cluster efficiency.Using decomposition
The region clustering comprehensive evaluation model built with comprehensive thinking, considers working environment residing for air-conditioning and air conditioner user is used
It is accustomed to the impact of this two classes factor, is prevented effectively from the one-sidedness of single index；In the Ward clustering algorithms of panel data three kinds away from
From different weight coefficients be by real data reflect quantity of information objective computation be given, it is to avoid the subjectivity of weight coefficient assignment
Property, while so that cluster analyses have effectiveness and motility concurrently.Region clustering analysis is the result of overall merit, right so as to reach
Geographic area residing for air-conditioning carries out the purpose of more science, sophisticated category.
Description of the drawings
Fig. 1 is the comprehensive evaluation model schematic diagram of region clustering analysis；
Fig. 2 is the flow chart of region clustering analysis；
Fig. 3 (a) is the region division schematic diagram of the CONTINENTAL AREA OF CHINA cluster result based on factor of natural environment；
Fig. 3 (b) is the region division schematic diagram of the CONTINENTAL AREA OF CHINA cluster result based on user's use habit factor；
Fig. 3 (c) is the region division schematic diagram of the CONTINENTAL AREA OF CHINA based on second zone cluster result；
Fig. 4 (a) is the China's Mainland geography division figure based on Climatic ecology zoning；
Fig. 4 (b) is the China's Mainland geography division figure based on economic level subregion；
Fig. 5 (a) is the variation tendency schematic diagram of the coefficient of variation in group based on factor of natural environment Clustering；
Fig. 5 (b) is the variation tendency schematic diagram of the coefficient of variation in the group being grouped based on user's use habit cluster analysis；
Fig. 6 (a) is the variation tendency schematic diagram of the coefficient of variation in the group being grouped based on economic level；
Fig. 6 (b) is the variation tendency schematic diagram of the coefficient of variation in the group being grouped based on climate type；
Fig. 6 (c) is the variation tendency schematic diagram of the coefficient of variation in the group being grouped based on second zone cluster result.
Specific embodiment
With reference to the accompanying drawings and examples technical scheme is described in detail, referring to Fig. 1 to Fig. 6 (c).
A kind of region clustering method based on air-conditioning reliability effect factor, comprises the following steps：
Step A, systematic analysiss affect the area difference of the Working Environments and user's use habit factor of air-conditioning reliability
The opposite sex, and two class reliability key influence factor of its working environment and user's use habit is extracted respectively, it is specific as follows：
Step A1, the analysis of air-conditioning work environmental impact factor regional differentiation is carried out, and extract the key of its working environment
Influence factor；
On the one hand, air-conditioning and its parts or components and parts are by factor of natural environments such as temperature, humidity, sunshine, precipitation
Occur tired, aging or consume etc. under Circulation, and then have influence on the use reliability of air-conditioning；On the other hand, air-conditioning is
No start operation is which uses or the working time is also closely related with the factor of natural environment such as temperature, humidity, sunshine, precipitation.
Present invention preferably uses including temperature, humidity, sunshine, precipitation factor of natural environment as working environment crucial effect
Factor；And using clustering target and measure be defined, wherein：
(1) temperature (DEG C)：With each province (city) all of region administrative unit mean temperature metric calculation monthly：
In formula：TMiptFor the p provinces t mean temperatures of i-th month, NpRepresent that p provinces include region administrative unit
Number, i represent month, the t expressions of years, xjiptFor j-th of the p provinces region administrative unit t mean temperatures of i-th month；
(2) relative humidity (%)：With each province (city) all of region administrative unit medial humidity metric calculation monthly：
In formula：HMiptFor the p provinces t medial humidities of i-th month, yjiptFor j-th of p provinces region administrative unit
The t medial humidities of i-th month；
(3) sunshine (h)：With each province (city) all of region administrative unit average sunshine time metric calculation monthly：
In formula：SMiptFor the p provinces t average sunshine times of i-th month, ZjiptIt is administrative single for j-th of p provinces region
The position t average sunshine times of i-th month；
(4) precipitation (mm)：Using each province (city) all of region administrative unit monthly (evening 8：00- evenings next day 8：00) when
Average precipitation metric calculation：
In formula：RMiptFor i-th month (evening 8 of p provinces t：00- evenings next day 8：00) average precipitation, UjiptSave for p
J-th of part (evenings 8 of region administrative unit t i-th month：00- evenings next day 8：00) average precipitation.
Step A2, the analysis of user use habit influence factor regional differentiation is carried out, and by the key of user's use habit
Influence factor is converted into the analysis impact of natural environment, economic condition and user psychology adaptability to user's use habit；
(1) natural environment：The different natural environment of weather conditions necessarily causes to show to each department air conditioner user use habit
Writing affects, preferably impact of the environmental factorss such as analysis temperature, humidity, sunshine, wind speed to human thermal comfort herein；And use
Clustering target and measure are defined, wherein：
Temperature, humidity and sunshine are affected on human thermal comfort using season meansigma methodss quantitative measurement, and computing formula and (1)-
(3) formula is consistent, and now the unit of i is season, season meansigma methodss can be calculated by correspondence month value statistical average；
Mean wind speed metric calculation of the wind speed (m/s) using each province (city) all of region administrative unit quarterly：
In formula：VJkptFor the mean wind speed in p provinces t k seasons, WjiptFor j-th of p provinces region administrative unit
The mean wind speed in t k seasons.
(2) economic condition：Economic condition affects the use time of air-conditioning to a certain extent, disappears preferably by society herein
Take two economic indicators of product total volume of retail sales rate of increase and average propensity to consume the level of economic development is weighed to user's use habit
Influence degree；And using clustering target and measure be defined, wherein：
Total retail sales of consumer goods season of the total retail sales of consumer goods rate of increase (%) using each province (city) kth season
Degree rate of increase (%) metric calculation.
Average propensity to consume tolerance of the average propensity to consume (%) using average each household quarterly, i.e., average each household is quarterly
The overall consumption expenditure of air-conditioning accounts for the ratio of average each household quarterly disposable income；Average propensity to consume is defined as：
In formula：APCkptFor air-conditioning average propensity to consume of the p provinces t annuals per user's kth season；ACkptFor p provinces
The disposable income in average per-user family t kth seasons；GCkptFor sky of the p provinces t annuals per user's kth season
Fixed consumption expenditure is adjusted, the price of one-time payment is converted in each season plus disposable mounting cost when being bought by user
In；FCkptAir-conditioning for p provinces average per-user t kth seasons uses the consumption expenditure, as producer typically has 6 years or so
Free repair guarantee, air-conditioning kth season cost of use can be simplified to kth season using produce the electricity charge, i.e.,
FCkpt=Mpt×Tkpt×Pk×Fpt (7)
In formula：MptIt is the every subscriber household air-conditioning owning amount of p provinces t annuals；PkFor the refrigeration or system in air-conditioning kth season
Heat consumption electrical power；FptFor the electricity price of p provinces t；TkptFor the available machine time of p provinces t kth season air-conditionings：
Tkpt=dkpt×tp (8)
In formula：dkptFor the start natural law in p provinces t kth seasons；tpAveragely heat daily or freeze for p provinces air-conditioning
Start duration.
(3) user psychology adaptability：Impact of the user psychology adaptability to user's use habit, that is, refer to NORTH CHINA ground
Area's winter is cold, carries out mostly centralized and unified heating mode, does not partly have central heating but the subscriber household in cold district
Tend to replace warming by air conditioner using other heating systems such as solar heating, burning coal heatings；SOUTHERN CHINA, the idle call at middle part
Then generally tend to using warming by air conditioner mode at family.
Step B, using the corresponding region clustering analysis indexes obtained by step A, build based on air-conditioning reliability shadow
The comprehensive evaluation model of the region clustering analysis of the factor of sound；
The temperature of extraction air-conditioning reliability effect factor, humidity, sunshine, rainfall, wind in this step specific embodiment
Speed, total retail sales of consumer goods rate of increase, average propensity to consume totally 7 indexs, according to first decomposing, thinking comprehensive again, structure
Build the comprehensive evaluation model that the region clustering based on air-conditioning reliability effect factor shown in Fig. 1 is analyzed.The model will be based on can
It is region clustering based on working environment influence factor and is used based on user by the region clustering PROBLEM DECOMPOSITION of property influence factor
Two subproblems of region clustering of custom influence factor, are that secondary cluster is required finally by the solution of comprehensive two subproblems
Cluster result.The complex nature of the problem can be reduced using the model, cluster efficiency is improved.
Step C, the decision criteria for making air-conditioning start cooling and warming, accurately quantify air-conditioning average propensity to consume and refer to
Mark；
The inventive method is scientifically and accurately to quantify influence degree of user's use habit factor to air-conditioning reliability, enters one
The decision criteria that step is made rational air-conditioning start refrigeration, heated, to the cooling and warming date for judging to need in 1 year, counts
Calculate each cooling and warming natural law in season, and determine the air conditioner user of different regions averagely start refrigeration, the duration for heating daily,
So as to accurately calculate the air-conditioning available machine time of different regions various quarters, the air-conditioning average propensity to consume for quantifying is obtained.
From energy-conservation principle in this specific embodiment, under winter room internal satisfaction comfort conditions, colder environment is selected,
Winter mainly by temperature cause it is uncomfortable, humidity affect it is less；Air-conditioning start behavior is typically only occurred in before door and window closes
Put, wind speed affects little to winter indoor environment.Summer selects partial heat environment in the case where comfort conditions are met, and works as indoor temperature and humidity
Beyond comfort standard when, air conditioner user may strengthen indoor air flows, therefore wind speed using the regulative mode such as gravity-flow ventilation
Summer Indoor environment is had a certain impact.Take and be respectively the 5-10 months, each 6 months of the 11-4 months 1 refrigeration, heat the time period.Root
According to ASHRAE55-2010 standards【ASHRAE.2010.ANSI/ASHRAE Standard 55-2010, Thermal
Environmental Conditions for Human Occupancy.Atlanta：ASHRAE.】, GB50736-2012 rule
Model【GB50736-2012. civil buildings heating ventilator and In Air Conditioning Design specification [S]. Beijing：Chinese architecture scientific research
Institute, 2012.】With the pertinent literature of human thermal comfort【Tian Yuanyuan, Xu Weiquan. the experiment of human thermal response under thermal and humidity environment is ground
Study carefully [J]. HVAC, 2003,33：27-30.】, the characteristics of in combination with China's air conditioner user hot adaptability, be given shown in table 1
The start decision criteria that air conditioner user is heated, freezed, wherein average every air-conditioning starts shooting duration daily according to document【Fructus Pruni salicinae love,
Shi Wenxing, Wang Baolong, etc. China's residential air conditioner time of origin investigation [C]. the 12nd national electric refrigerator (cabinet), air-conditioning
Device and compressor academic exchange conference collection of thesis, 2014,34-39.】The residential air conditioner operation of the China's Mainland different regions for being given
Time calculates.
1 air conditioner refrigerating of table heats start decision criteria
Note：- the sub- factor is represented without impact, --- represent the heating of this area's heating installation
Step D, using the weight of each distance of sum of deviation square method objective computation, then with the Ward clustering algorithms point of weighting
Cluster analyses are not carried out from working environment, user's use habit this two class influence factor, obtain two kinds of classification region clustering results,
Then synthesis is carried out to two kinds of classification region clustering results using secondary clustering method, so as to obtain final area distribution.
Idiographic flow is as follows：
(1) the similarity measure index of construction multi objective panel data, and determine the weight coefficient of three kinds of distances, so as to
To the Ward clustering algorithms of weighting；
Panel data has space sample, the data structure of three dimensional informations of time serieses and index, is widely used in
The territorial classification research of natural environment, social life and economic dispatch field.Present invention preferably uses multi objective panel data
Based on model carry out cluster analyses.
A simple class formation is produced from a complex set of multi objective panel data, inevitable requirement carries out similarity measurements
Amount.The similarity measure distance function and Ward clustering algorithms of panel data is prior art, and detailed content can refer to document【Lee
Cause and effect, Dai Yi, He Xiaoqun. the panel data clustering method [J] based on adaptive weighting. the system engineering theory and practice, 2013,
33(2)：388-395.】, specific configuration method is as follows：
If it is X={ x that analysis object to be clustered is all1, x2..., xi..., xn(1≤i≤n), each sample xiWith m
Index, time span are T, xijtRepresent numerical value of j-th index of individuality i in t.Similarity measure distance function is covered absolutely
To span from dij(AQED), speedup is apart from dij(ISED) and fluctuation apart from dij(VCED), using the weighted array table of three kinds of distances
Show the comprehensive distance between individuality,
Dij(CED)=α*·dij(AQED)+β*·dij(ISED)+γ*·dij(VCED) (9)
Wherein α*, β*And γ*Three kinds are represented respectively apart from corresponding weight coefficient, and meet α*+β*+γ*=1.
Ward hierarchical clustering methods highlight the homogeneity inside class area, are suitable for the collection clustering with region as sample
Class, subpartition decision-making.Each sample in set is first constituted a class by itself by the method, when categories combination is carried out, deviation is put down
Two minimum classes of side and increasing degree merge first, then merge all categories step by step successively；Its specific algorithm is as follows：
By nIt is individualSample is divided into k classes：G1, G2... Gk, xilFor class GlIn i-thIt is individualIndividuality, NlFor class GlIn individual amount,
According to Ward method classificating thoughts, it is the impact that can consider three kinds of distances, it is to avoid Ward statistics deteriorate to traditional Ward systems
Metering, panel data classification GlMiddle Nl It is individualThe sum of deviation square function of sample is
In formula：xiltRepresent GlThe absolute magnitude of middle t periods individuality i, yiltRepresent GlThe incremental velocity of middle t periods individuality i, Zilt
Represent GlThe degree of variation of middle individual i.
α is given preferably by the sum of deviation square objective computation of three kinds of distances herein*, β*andγ*, it is to avoid weight system
The subjectivity of number assignment, concrete grammar are as follows：
In formula：R is the root-mean-square standard deviation of the l time and class, and ν is the individual dimension of observation, and p is class LlIn observation
Individual number, L (AQED), L (ISED), L (VCED) are represented respectively and individually sample to be analyzed are clustered with three kinds of distances
Sum of deviation square in the whole class of analysis.
It is assumed that being divided into k classes, then total sum of deviation square function is
The amplitude that sum of deviation square increases after certain two class merges in Ward methods, that is, be counted as between class and class square away from
From if by certain two class GAAnd GBMerge into new class GC, then class GAAnd GBThe distance between be
Other classes GKWith new class GCRange formula be
In formula：NA、NB、NCAnd NKIt is class G respectivelyA、GB、GCAnd GKCorresponding number of individuals.By formula (15) obtain Ward length of normals from
Matrix, minimum corresponding classification G of screening distanceiAnd GjIt is polymerized.
(2) Panel Data Model A is built using the key influence factor factor of natural environment, and by the index in model A
Quantization, monthly data normalization, are carried out based on natural environment according to the sorting procedure of the weighting Ward methods of multi objective panel data
The region clustering analysis of factor, obtains the region clustering result based on factor of natural environment.Then using user's use habit
Key influence factor builds Panel Data Model B, by the quantification of targets in Model B, season data normalization, according to weighting Ward
The sorting procedure of method carries out the region clustering based on user's use habit factor and analyzes, and obtains based on user's use habit factor
Region clustering result.
(3) synthesis is carried out to the other region clustering result of two species using secondary clustering procedure, specifically in accordance with the following methods：First
Natural environment influence factor region clustering result, user's use habit influence factor's region clustering result are effectively combined, is subordinate to
Belong to of a sort multiple geographic areas be considered initial classes, cannot determine the geographic area of belonging kinds be considered not sort out,
Independent groups of geographic area is considered separate class；After category label is finished, to the classification comprising two and above geographic area
The barycenter of cluster centre, i.e. classification is solved, all categories is clustered again with the sum of deviation square method of weighting, is analyzed and weigh
Both between class distances, finally give the general area distribution results after secondary cluster.
The region clustering flow chart schematic diagram adopted in this specific embodiment is as shown in Figure 2.
Embodiment：
In order to verify The effect of invention, by the aforementioned region clustering comprehensive evaluation model based on air-conditioning reliability effect factor
The region clustering analysis of the 31 provincial administrative unit divisions in China's Mainland is applied to the idiographic flow of region clustering analysis, is such as schemed
1～2.
According to the method for regional economic statistics yearbook zoning, with the provincial administrative unit division in 31, China's Mainland
For sample to be analyzed.Due to State Statistics Bureau issue with regard to cities and towns and rural resident's family it is average per one hundred houses air-conditioning owning amount with
And the statistical data of urban residents' disposable income per capita and rural per-capita income was by the end of 2012, China Meteorological science
The meteorological data that data sharing service net is issued was by the end of 2013, therefore the span of sample interval selects 2005-2012, extracted
Factor of natural environment and user using air-conditioning custom factor build respectively Panel Data Model A (31 × 8 × 48), B (31 × 8 ×
24).Wherein the Data Source of temperature, relative humidity, sunshine, precipitation and wind speed is in China Meteorological Sharing Services for Scientific Data net
Chinese terrestrial climate data moon Value Data collection, Chinese terrestrial climate data earning in a day data set；The total retail sales of consumer goods increase
Long rate, average each household air-conditioning owning amount, average each household disposable income are then bases《Regional economic statistics yearbook》It is related
Data are computed drawing.
(1) region clustering based on factor of natural environment is analyzed
The weight of absolute magnitude distance, speedup distance and fluctuation apart from three is calculated according to formula (11), (12), respectively
α*=0.6219, β*=0.2238, γ*=0.1544
Region clustering analysis is carried out according to sorting procedure, the provincial region dividing unit in 31, China's Mainland is based on factor of natural environment
Panel data be divided into 10 classes, be shown in Table 2.
Region clustering analysis result of the table 2 based on factor of natural environment
| Classification | Area | Classification | Area |
| Group I | Beijing, Tianjin, Hebei, Shanxi | Group VI | Chongqing, Sichuan, Guizhou |
| Group II | The Inner Mongol, Gansu, Ningxia, Xinjiang | Group VII | Shandong, Henan, Shaanxi |
| Group III | Liaoning, Jilin, Heilungkiang | Group VIII | Qinghai, Tibet |
| Group IV | Shanghai, Zhejiang, Jiangsu, Anhui, Hubei | Group IX | Yunnan |
| Group V | Fujian, Guangdong, Guangxi, Jiangxi, Hunan | Group X | Hainan |
(2) region clustering based on user's use habit factor is analyzed
The cooling and warming be given using table 1 is started shooting criterion, and the climatic data daily to each department 2005-2012 is one by one time
Go through screening, the annual various quarters start refrigeration of the provincial administrative unit division of itemized record 31, heat in be belonging respectively to day off and
Workaday natural law, table 3 give the data of wherein 8 provincial administrative unit divisions, wherein Beijing, Liaoning, Henan and Xinjiang
Deng northern territory due to the mental aptitude of consideration resident, winter is often tended to replace air-conditioning heating using heating, therefore
Air-conditioning start natural law is 0.
3 some areas various quarters of table air conditioner refrigerating/heat natural law
According to each department refrigeration, start natural law is heated, quarterly average consumption is inclined air-conditioning to be calculated by formula (6)-(8)
To.Absolute magnitude distance, speedup distance and fluctuation are calculated according to formula (11), (12) to be respectively apart from the weight of three
α*=0.6707, β*=0.2536, γ*=0.0757
Region clustering is carried out according to sorting procedure, the provincial region dividing unit in 31, China's Mainland is based on air conditioner user use habit
The panel data of factor is divided into 10 classes, is shown in Table 4.
Region clustering analysis result of the table 4 based on user's use habit factor
| Classification | Area | Classification | Area |
| Group I | Beijing, Tianjin, Hebei, Shanxi, Shandong | Group VI | Zhejiang, Guangdong, Fujian |
| Group II | The Inner Mongol, Gansu, Ningxia, Xinjiang, Qinghai | Group VII | Henan, Shaanxi, Yunnan |
| Group III | Liaoning, Jilin, Heilungkiang | Group VIII | Shanghai |
| Group IV | Jiangsu, Anhui, Hubei | Group IX | Hainan |
| Group V | Guangxi, Jiangxi, Hunan, Chongqing, Sichuan, Guizhou | Group X | Tibet |
(3) second zone cluster analyses
Tied according to the region clustering result based on factor of natural environment and the region clustering based on user's use habit factor
Really, there are 8 classes in the big class city that can be under the jurisdiction of same class city, is considered as initial classes；Remaining Shanghai, Zhejiang, Shandong,
They in natural environment and use habit, in different classifications, therefore are respectively seen as 6 and are not returned by Yunnan, Tibet, Qinghai
Class；Hainan is considered as separate class.This 15 classifications are carried out into second cluster analysis in the lump.
By initial classes, do not sort out with separate class again respectively build Panel Data Model C (15 × 8 × 48), D (15 × 8 ×
24), according to step cluster analyses again, the second zone cluster analysis result of the 31 provincial region dividing units in China's Mainland is obtained,
As shown in table 5 and Fig. 3 (a), (b), (c).By Fig. 3 (a), (b), (c) as can be seen that second zone cluster result is compared to certainly
So environment, the geographic area of the independent cluster analyses of two class factor of user's use habit divide, internuncial regional context is met
Meanwhile, territorial classification is more fine；Cluster can preferably reflect natural environment or user's use habit type factor for the first time
In macroscopical difference, second cluster is then ideal in terms of microcosmic, constitutes complete area division scheme.
5 second zone cluster analysis result of table
| Classification | Area | Classification | Area |
| Group I | Beijing, Tianjin, Hebei, Shanxi, Shandong | Group VII | Guangdong, Fujian |
| Group II | The Inner Mongol, Gansu, Ningxia, Xinjiang | Group VIII | Henan, Shaanxi |
| Group III | Liaoning, Jilin, Heilungkiang | Group IX | Qinghai, Tibet |
| Group IV | Jiangsu, Zhejiang, Anhui, Hubei | Group X | Shanghai |
| Group V | Guangxi, Hunan, Jiangxi | XI groups | Hainan |
| Group VI | Chongqing, Sichuan, Guizhou | XII groups | Yunnan |
Select herein by the region clustering result based on factor of natural environment with according to climate in china type general zoning,
Region clustering result based on user's use habit factor and the zoning by China's economic level and second zone cluster
Analysis result and document【Liu Weidong, Song Haowei, Zhao Zhiwei, Li Jie. commented based on the air-conditioning reliability of cluster analyses and time concentration
Estimate [J]. Industrial Engineering and Management, 2013,18 (4)：156-160.】The cluster result of middle subtropical zone monsoon climatic region carries out three groups
Relatively, shown in wherein Climatic ecology zoning, economic regionalism such as Fig. 4 (a), (b).From Fig. 3 (a), (b), (c) and Fig. 4 (a), (b) can
To find out, the geographic area carried out from weather or economic angle divides all excessively rough, not fully suitable for air-conditioning reliability
The research of assessment.Above-mentioned document, is incited somebody to action with subtropical monsoon climate district provincial capital as province representative, humiture as clustering target
Subtropical monsoon climate district is divided into four groups, and the second zone cluster result of the inventive method is being classified compared with subtropical monsoon climate district
It is even better in precision.
The convergent journey of each department reliability effect factor in the group that different geographic regions are divided is assessed using VC Method
Degree.In view of the regional negligible amounts after the 7th group, study and only first group～the 7th group are verified.Intuitively to show
All data are standardized by 2005-2012 group differences level and situation of change, as a result such as Fig. 5 (a), (b)
It is shown with Fig. 6 (a), (b), (c).Can be seen by Fig. 5 (a), (b), based on the Clustering that user's use habit factor is carried out
Group in the coefficient of variation change over slowly, maintain essentially in a radix level, and be based on factor of natural environment cluster point
In the group of group, the coefficient of variation is then relatively large with time fluctuating margin, is air-conditioning reliability convergent degree in each department in regulation group
Principal element.The coefficient of variation in the group being grouped based on this paper second zones cluster result be can be seen that from Fig. 6 (a), (b), (c)
Significantly lower than the regional classification based on climate type and economic level, the homoplasy after new Clustering in group is significantly improved,
Its similarity is shown more.
Comprehensive Fig. 5 (a), (b) and Fig. 6 (a), (b), the relative analyses of (c) understand that the region clustering result of context of methods is divided
In same group of group, each department reliability effect factor shows preferable homoplasy in recent years, the air-conditioning of different regions in group
It is minimum using reliability difference.
The quantitative model science of the clustering target that the present invention is adopted has measured influence degree of each factor to air-conditioning reliability；
The air-conditioning start made freezes, heats decision criteria, can precisely quantify air-conditioning average propensity to consume；And the region for building
Clustering Comprehensive model has considered the impact of working environment residing for air-conditioning and air conditioner user use habit this two classes factor；
The quantity of information objective computation that the different weight coefficients of three kinds of distances are reflected by real data in the Ward clustering algorithms of panel data
Be given, it is to avoid the subjectivity of weight coefficient assignment.Therefore the region clustering method of the present invention is to geographic area residing for air-conditioning
Classification more science, fine.And the inventive method on affected by factor of natural environment refrigerator, electric fan, water heater, the sun
Energy streetlight and affected the similar research of the products such as television set, washing machine that also there is good reference valency by user's use habit factor
Value.
List of references
[1] Yau, Y.H., and H.L.Pean.The climate change impact on air conditioner
system and reliability in Malaysia—A review[J].Renewable and Sustainable
Energy Reviews, 2011,15 (9)：4939-4949.
[2] Li, Y., G.Zhang, G.Z.Lv, A.N.Zhang and R.Z.Wang.Performance study of a
solar photovoltaic air conditioner in the hot summer and cold winter zone[J]
.Solar Energy, 2015,117：167-179.
[3] Yau, Y.H.and H.L.Pean.The performance study of a split type air
Conditioning system in the tropics, as affected by weather [J] .Energy and
Buildings, 2014,72：1-7.
[4] Yu, S., YM.Wei, J.Fan, X.Zhang, and K.Wang.Exploring the regional
characteristics of inter-provincial CO2emissions in China：An improved fuzzy
Clustering analysis based on particle swarm optimization [J] .Applied Energy,
2012,92：552-562.
[5] Saracli, Sinan.Performance of Rand ' s C statistics in clustering
analysis：an application to clustering the regions of Turkey[J].Journal of
Inequalities and Applications, 2013：142.
[6] Iyigun, C., M.Batmaz, C.Yozgatligil, V.E.K.and
M.Z.2013.Clustering current climate regions of Turkey by using a
multivariate statistical method.Theoretical and applied climatology 114(1)：
95-106。
Claims (10)
1. a kind of region clustering method based on air-conditioning reliability effect factor, it is characterised in that its step is as follows：
Step A, systematic analysiss affect the regional differentiation of the Working Environments and user's use habit factor of air-conditioning reliability,
And two class reliability key influence factor of its working environment and user's use habit is extracted respectively；
Step B, using the corresponding region clustering analysis indexes obtained by step A, build based on air-conditioning reliability effect because
The comprehensive evaluation model of the region clustering analysis of element；
Step C, the decision criteria for making air-conditioning start cooling and warming, accurately quantify air-conditioning average propensity to consume index；
Step D, using the weight of each distance of sum of deviation square method objective computation, then with weighting Ward methods respectively from building ring
Border, user's use habit this two class influence factor carry out cluster analyses, obtain two kinds of classification region clustering results, then using two
Secondary clustering procedure carries out synthesis to two kinds of classification region clustering results, so as to obtain final area distribution.
2. the region clustering method based on air-conditioning reliability effect factor according to claim 1, it is characterised in that described
The key influence factor of working environment is referred to chooses the factor of natural environment for including temperature, humidity, sunshine and precipitation；And use
Clustering target and measure be defined, wherein：
(1) temperature (DEG C)：With each province (city) all of region administrative unit mean temperature metric calculation monthly：
In formula：TMiptFor the p provinces t mean temperatures of i-th month, Np represents that p provinces include region administrative unit number, i
Represent month, the t expressions of years, xjiptFor j-th of the p provinces region administrative unit t mean temperatures of i-th month；
(2) relative humidity (%)：With each province (city) all of region administrative unit medial humidity metric calculation monthly：
In formula：HMiptFor the p provinces t medial humidities of i-th month, yjiptFor j-th of p provinces region administrative unit t
The medial humidity of i month；
(3) sunshine (h)：With each province (city) all of region administrative unit average sunshine time metric calculation monthly：
In formula：SMiptFor the p provinces t average sunshine times of i-th month, ZjiptFor j-th of p provinces region administrative unit t
The average sunshine time in i-th month year；
(4) precipitation (mm)：Using each province (city) all of region administrative unit monthly (evening 8：00- evenings next day 8：00) it is flat when
Equal precipitation metric calculation：
In formula：RMiptFor i-th month (evening 8 of p provinces t：00- evenings next day 8：00) average precipitation, UjiptFor p provinces jth
I-th month (evening 8 of individual region administrative unit t：00- evenings next day 8：00) average precipitation.
3. the region clustering method based on air-conditioning reliability effect factor according to claim 1, it is characterised in that described
The key influence factor of user's use habit is converted into analysis natural environment, economic condition and user psychology adaptability to user
The impact of use habit.
4. the region clustering method based on air-conditioning reliability effect factor according to claim 3, it is characterised in that described
Natural environment includes temperature, humidity, sunshine and wind speed environments factor；And using clustering target and measure be defined,
Wherein：
Temperature, humidity and sunshine are affected on human thermal comfort using season meansigma methodss quantitative measurement, computing formula and (1)-(3) formula
Unanimously, now the unit of i be season, season meansigma methodss can by correspondence month value statistical average be calculated；
Mean wind speed metric calculation of the wind speed (m/s) using each province (city) all of region administrative unit quarterly：
In formula：VJkptFor the mean wind speed in p provinces t k seasons, WjiptFor j-th of p provinces region administrative unit t k
The mean wind speed in individual season.
5. the region clustering method based on air-conditioning reliability effect factor according to claim 3, it is characterised in that described
Economic condition includes two economic indicators of the total retail sales of consumer goods and average propensity to consume, is weighed by above-mentioned economic indicator
Influence degree of the level of economic development to user's use habit；And using clustering target and measure be defined, wherein：
Average propensity to consume tolerance of the average propensity to consume (%) using average each household quarterly, i.e., average each household quarterly air-conditioning
Overall consumption expenditure account for the ratio of average each household quarterly disposable income；Average propensity to consume is defined as：
In formula：APCkptFor air-conditioning average propensity to consume of the p provinces t annuals per user's kth season；ACkptIt is average for p provinces
Disposable income per subscriber household t kth seasons；GCkptIt is solid for air-conditioning of the p provinces t annuals per user's kth season
Determine the consumption expenditure, the price of one-time payment is converted in each season plus disposable mounting cost when being bought by user；
FCkptAir-conditioning for p provinces average per-user t kth seasons uses the consumption expenditure, exempts within 6 years or so as producer typically has
Take guarantee, air-conditioning kth season cost of use can be simplified to kth season using the electricity charge for producing, i.e.,
FCkpt=Mpt×Tkpt×Pk×Fpt (7)
In formula：MptIt is the every subscriber household air-conditioning owning amount of p provinces t annuals；PkRefrigeration for air-conditioning kth season heats consumption
Electrical power；FptFor the electricity price of p provinces t；TkptFor the available machine time of p provinces t kth season air-conditionings：
Tkpt=dkpt×tp (8)
In formula：dkptFor the start natural law in p provinces t kth seasons；tpAveragely heat for p provinces air-conditioning or freeze start daily
Duration.
6. the region clustering method based on air-conditioning reliability effect factor according to claim 3, it is characterised in that described
Impact of the user psychology adaptability to user's use habit, that is, refer to that NORTH CHINA area winter is cold, carry out mostly and concentrate system
One heating mode, does not partly have central heating but the subscriber household in cold district tends to using solar heating, burns coal
Other heating systems such as heating replace warming by air conditioner；SOUTHERN CHINA, the air conditioner user at middle part then generally tend to take using air-conditioning
Warm mode.
7. the region clustering method based on air-conditioning reliability effect factor according to claim 1, it is characterised in that step
Make described in C air-conditioning start shooting cooling and warming decision criteria, the cooling and warming date needed in referring to judge 1 year,
Calculate each cooling and warming natural law in season, and determine the air conditioner user of different regions averagely daily start refrigeration, heat when
Long, so as to the air-conditioning available machine time of different regions various quarters in accurate calculating formula (8), the air-conditioning average consumption for obtaining quantifying is inclined
To.
8. the region clustering method based on air-conditioning reliability effect factor according to claim 1, it is characterised in that described
Gathered from working environment, user's use habit this two class influence factor with the Ward clustering algorithms of weighting respectively in step D
During alanysis, the similarity measure function of multi objective panel data covers absolute magnitude apart from dij(AQED), speedup is apart from dij
(ISED) and fluctuation apart from dij(VCED) comprehensive distance between individuality, i.e. D are represented using the weighted array of three kinds of distances,ij
(CED)=α*·dij(AQED)+β*·dij(ISED)+γ*·dij(VCED), wherein α*,β*And γ*Three kinds of distances are represented respectively
Corresponding weight coefficient, and meet α*+β*+γ*=1.
9. the region clustering method based on air-conditioning reliability effect factor according to claim 8, it is characterised in that described
Three kinds, apart from corresponding weight coefficient, calculate according to the sum of deviation square of three kinds of distances, specific as follows：
In formula：R is the root-mean-square standard deviation of the l time and class, and ν is the individual dimension of observation, and p is class LlIn observation it is individual
Number, L (AQED), L (ISED), L (VCED) is represented respectively individually carries out cluster analyses to sample to be analyzed with three kinds of distances
Whole class in sum of deviation square.
10. the region clustering method based on air-conditioning reliability effect factor according to claim 1, it is characterised in that institute
State and synthesis is carried out to two kinds of region clustering result using secondary clustering procedure described in step D, concrete grammar is：First will be from
So environmental impact factor region clustering result, user's use habit influence factor's region clustering result are effectively combined, and are under the jurisdiction of
Of a sort multiple geographic areas be considered initial classes, cannot determine the geographic area of belonging kinds be considered not sort out, it is independent
Groups of geographic area is considered separate class；After category label is finished, the classification comprising two and above geographic area is solved
All categories are clustered again by the barycenter of cluster centre, i.e. classification with the sum of deviation square method of weighting, are analyzed and are weighed both
Between class distance obtain the general area distribution situation after secondary cluster.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610996880.1A CN106570729A (en) | 2016-11-14 | 2016-11-14 | Air conditioner reliability influence factor-based regional clustering method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610996880.1A CN106570729A (en) | 2016-11-14 | 2016-11-14 | Air conditioner reliability influence factor-based regional clustering method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106570729A true CN106570729A (en) | 2017-04-19 |
Family
ID=58542523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610996880.1A Pending CN106570729A (en) | 2016-11-14 | 2016-11-14 | Air conditioner reliability influence factor-based regional clustering method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106570729A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063769A (en) * | 2018-08-01 | 2018-12-21 | 济南大学 | Clustering method, system and the medium of number of clusters amount are automatically confirmed that based on the coefficient of variation |
| CN111322716A (en) * | 2020-02-24 | 2020-06-23 | 青岛海尔工业智能研究院有限公司 | Air conditioner temperature automatic setting method, air conditioner, equipment and storage medium |
| CN112948524A (en) * | 2021-04-21 | 2021-06-11 | 广东电网有限责任公司计量中心 | Intelligent electric meter operation area grouping method and system based on environment and geographic characteristics |
| CN113283056A (en) * | 2021-04-28 | 2021-08-20 | 长安大学 | Method for calculating adaptability of evaporative cooling air conditioning technology in different regions |
| CN113505823A (en) * | 2021-07-02 | 2021-10-15 | 中国联合网络通信集团有限公司 | Supply chain security analysis method and computer-readable storage medium |
| CN115978720A (en) * | 2022-12-30 | 2023-04-18 | 北京创今智能科技有限公司 | Non-equivalent grouping method for air source heat pump units |
-
2016
- 2016-11-14 CN CN201610996880.1A patent/CN106570729A/en active Pending
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063769B (en) * | 2018-08-01 | 2021-04-09 | 济南大学 | Clustering method, system and medium for automatically determining cluster number based on coefficient of variation |
| CN109063769A (en) * | 2018-08-01 | 2018-12-21 | 济南大学 | Clustering method, system and the medium of number of clusters amount are automatically confirmed that based on the coefficient of variation |
| CN111322716A (en) * | 2020-02-24 | 2020-06-23 | 青岛海尔工业智能研究院有限公司 | Air conditioner temperature automatic setting method, air conditioner, equipment and storage medium |
| CN111322716B (en) * | 2020-02-24 | 2021-08-03 | 青岛海尔工业智能研究院有限公司 | Air conditioner temperature automatic setting method, air conditioner, equipment and storage medium |
| CN112948524A (en) * | 2021-04-21 | 2021-06-11 | 广东电网有限责任公司计量中心 | Intelligent electric meter operation area grouping method and system based on environment and geographic characteristics |
| CN112948524B (en) * | 2021-04-21 | 2024-04-26 | 广东电网有限责任公司计量中心 | Intelligent ammeter operation area grouping method and system based on environment and geographic features |
| CN113283056B (en) * | 2021-04-28 | 2023-12-01 | 长安大学 | Method for calculating adaptability of evaporative cooling air conditioning technology in different areas |
| CN113283056A (en) * | 2021-04-28 | 2021-08-20 | 长安大学 | Method for calculating adaptability of evaporative cooling air conditioning technology in different regions |
| CN113283056B8 (en) * | 2021-04-28 | 2024-02-23 | 长安大学 | Method for calculating adaptability of evaporative cooling air conditioning technology in different areas |
| CN113505823A (en) * | 2021-07-02 | 2021-10-15 | 中国联合网络通信集团有限公司 | Supply chain security analysis method and computer-readable storage medium |
| CN113505823B (en) * | 2021-07-02 | 2023-06-23 | 中国联合网络通信集团有限公司 | Supply chain security analysis method and computer readable storage medium |
| CN115978720B (en) * | 2022-12-30 | 2023-07-04 | 北京创今智能科技有限公司 | Unequal grouping method for air source heat pump units |
| CN115978720A (en) * | 2022-12-30 | 2023-04-18 | 北京创今智能科技有限公司 | Non-equivalent grouping method for air source heat pump units |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106570729A (en) | Air conditioner reliability influence factor-based regional clustering method | |
| Yu et al. | A systematic procedure to study the influence of occupant behavior on building energy consumption | |
| Hu et al. | Using bottom-up model to analyze cooling energy consumption in China's urban residential building | |
| Chen et al. | Statistical analyses on winter energy consumption characteristics of residential buildings in some cities of China | |
| Roberts et al. | Characterisation of Australian apartment electricity demand and its implications for low-carbon cities | |
| CN107169606A (en) | A kind of Forecasting Methodology of office building refrigeration duty | |
| Chen et al. | Statistical analyses on summer energy consumption characteristics of residential buildings in some cities of China | |
| Sun et al. | Improved energy performance evaluating and ranking approach for office buildings using Simple-normalization, Entropy-based TOPSIS and K-means method | |
| CN106681964A (en) | Public building energy consumption auditing statistics splitting method | |
| Swan | Residential sector energy and GHG emissions model for the assessment of new technologies | |
| Lei et al. | Prediction method of energy consumption for high building based on LMBP neural network | |
| Li et al. | Relations of total electricity consumption to climate change in Nanjing | |
| Lu et al. | The performance and analysis of office building energy consumption in the west of Inner Mongolia Autonomous Region, China | |
| Sha et al. | Development of a key-variable-based parallel HVAC energy predictive model | |
| Lam et al. | An analysis of climatic influences on chiller plant electricity consumption | |
| Chen et al. | Contrastive analyses on annual energy consumption characteristics and the influence mechanism between new and old residential buildings in Shanghai, China, by the statistical methods | |
| Xia et al. | Technological adaptation zone of passive evaporative cooling of China, based on a clustering analysis | |
| Yu et al. | The optimization research on the coupled of active and passive energy supplying in public institutions in China | |
| Liu et al. | Evaluation on the energy consumption and thermal performance in different residential building types during mid-season in hot-summer and cold-winter zone in China | |
| Li et al. | The impacts of temperature on residential electricity consumption in Anhui, China: does the electricity price matter? | |
| CN109961181A (en) | Method and system for the prediction of building energy hourly load | |
| Jie et al. | Regional cluster analysis of mainland china based on factors of the reliability of air-conditioning systems | |
| Vihola et al. | Heat loss rate of the Finnish building stock | |
| Lan et al. | Applicability analysis of solar heating system in China based on a reliability-based optimization method for auxiliary heater capacity | |
| CN115271168A (en) | Method and device for predicting response potential of electrical load and storage medium |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170419 |
|
| WD01 | Invention patent application deemed withdrawn after publication |