CN106651425A - User electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data - Google Patents
User electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data Download PDFInfo
- Publication number
- CN106651425A CN106651425A CN201610861397.2A CN201610861397A CN106651425A CN 106651425 A CN106651425 A CN 106651425A CN 201610861397 A CN201610861397 A CN 201610861397A CN 106651425 A CN106651425 A CN 106651425A
- Authority
- CN
- China
- Prior art keywords
- electricity
- user
- variable
- business process
- process system
- 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
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- 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
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The invention discloses a user electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data. The method comprises the following steps: the newly-added user number and the newly-added electricity consumption equipment capacity are determined, and a capacity growth rate and a ring growth rate are calculated; a Von Bertanlanffy growth curve is used for constructing an electricity consumption trend after business expanding installation, a principal component analysis is used to obtain a business expanding-used electricity consumption growth curve, an instantaneous growth rate serves as a base value and an inflection point serves as a recognition basis, and staged interpretation on the curve is carried out; the electricity consumption and the business expanding installation are converted, a census X12 algorithm is used for seasonal decomposition on an explanatory variable and a dependent variable, and the original variable is decomposed into a trend cycle term, a seasonal factor term and a random term; and the relevance between two variable trends is analyzed, linear regression is then carried out, and after seasonal factors are considered, the prediction value of the electricity consumption is determined. The actual electricity consumption of each user and the prediction value of the electricity consumption are compared, if the difference exceeds an estimated threshold, the user is judged to consume the electricity abnormally, and the user is marked.
Description
Technical field
The present invention relates to a kind of user's power stealing for considering Business Process System data, electric leakage behavior monitoring method.
Background technology
Due to being affected by national policy, production cost, sales situations, market prospect, industry expansion trend, it is impossible to expand new to industry
Increase-volume amount, electricity, load are made accurate prediction and are judged.Longer time process is needed as industry expands new clothes, it is impossible to accurately
The release rule of newly-increased capacity is held, when the rule and electricity of electricity growth point reach normal level, the increasing of capacity
Long impact to the whole province's electricity etc..That is, the growth of the actual electricity sales amount of power grid enterprises often to be lagged behind the growth of Business Process System,
It is newly-increased to apply to install the growth that capacity also absolutely be converted into electricity sales amount.Historical data is counted, newly-increased report is found out
Dress and electricity sales amount increase between relation, contribute to power grid enterprises the growth of following electricity sales amount change is carried out it is more accurately pre-
Meter.
But at present the correlational study principal concern of Business Process System is the perfect of policy making and management system, seldom
In-depth study is done from measurement technology, mathematical model, algorithm rehearsal, such as load forecasting method based on S- curves is based on
Business expands the accuracy for improving the prediction of future 3-5 internal loadings of region large electricity consumer's simple, intuitive of information.But to not
Carry out the prediction of several years, if thinking finer being predicted, in addition it is also necessary to carry out monthly analysis.
And accurate predictive value cannot be obtained, cannot just judge electricity consumption user whether normal electricity consumption, if there is power stealing, leakage
The behavior of electricity, normally runs to electric power maintenance and power system and brings larger harm.
The content of the invention
The present invention is in order to solve the above problems, it is proposed that a kind of user's power stealing for considering Business Process System data, electric leakage behavior
Monitoring method, this method proposes the analysis theory and overall framework of power sales analyses and prediction index system, and here basis
On developed market share analysis, Market Concentration Ratio, typical user's analysis, apply to install tracking etc. and have distinct marketing work special
The New Set of color, to comprehensively, in depth evaluating the overall development situation of power sales.The index system is to realize marketing work
Standardization, scientific, efficient there is provided support.
To achieve these goals, the present invention is adopted the following technical scheme that：
A kind of user's power stealing for considering Business Process System data, electric leakage behavior monitoring method, comprise the following steps：
(1) determine Add User amount and newly-increased power consumption equipment capacity, calculate capacity rate of increase and sequential growth rate；
(2) the electricity consumption trend after Business Process System is built using Von Bertanlanffy growth curves, using pivot analysis
Method obtains industry and expands electricity consumption growth curve, and with transient growth rate as base value, flex point is basis of characterization, carries out the solution stage by stage of curve
Read；
(3) power consumption and Business Process System are changed, explanatory variable and dependent variable is entered using census X12 algorithms
Row is seasonal to be decomposed, and original variable is decomposed into trend circulation item (TC), seasonal factor (SF) and random entry (IR)；
(4) dependency analyzed between two variable trends items carries out linear regression again, it is considered to after seasonal factor, confirms electricity consumption
The predictive value of amount；
(5) predictive value by the actual power consumption of each user with its power consumption is compared, if its difference exceedes estimated
Threshold value, then judge that the user has abnormal electricity consumption, examined accordingly and early warning.
In the step (1), by calculating newly-increased quantity and capacity and its rate of increase, to confirm that following electrity market increases
Trend.
In the step (1), total numerical quantity is held, while the composition situation of total amount is grasped, from different regions, different electricity consumptions
The multiple angle analysis total amounts of classification, different industries, each electric pressure and each industry and its constitute the relation of component.
In the step (2), power consumption is confirmed based on maximum electricity, parameter, transient growth speed and time scale
Growth tendency.
In the step (2), the flex point in curve is exactly the turning point of curvilinear motion speed, and second dervative is zero
Point, which characterizes the variation tendency of curve, when second dervative is zero, reaches respective flex point month, and its corresponding electricity is
Flex point electricity.
In the step (3), using the electricity consumption trend curve after Business Process System, with reference to instantaneous growth rate and relative growth
Rate, emphasizes the trend characteristic of electricity consumption from growth change rate aspect, with transient growth rate as base value, weighs the wave characteristic of curve.
In the step (3), using principle component analysis by different angle analysis variables, using linear combination by original
Beginning information aggregate gets up, and forms aggregate variable that is orthogonal and covering most data information.
In the step (3), it is applied to industry and expands growth curve using the electricity of different clients as original variable, carry out timesharing
Between, the two-way extraction of point client, obtain pivot and the typical customers chosen lifted in the overall aspect representated by client, with
, used as explanatory variable, power consumption is used as dependent variable for Business Process System data.
In the step (3), it is assumed that the sample space of data is s, the index number of the observation of each sample is m, pivot
The mathematical model of analytic process is represented by：
Wherein output matrix y1, y2... ymFor the key message of initial data, it is exactly pivot, coefficient matrices A refers to original
Coefficient of association between variable and pivot, state matrix x1, x2... xmFactor variable, that is, original variable are referred to, is herein referred to
The power consumption of different industries client.
In the step (4), power consumption and apply to install all there is certain seasonality, using month degree as time observation unit
Time serieses generally with the cyclically-varying in units of year, this affects to cause due to seasonal factor, referred to as season
Property change, analyze objectivity influence factor when, season key element is rejected from former sequence, seasonal adjustment is carried out.
In the step (5), calculate actual power consumption and estimate the difference of power consumption and the ratio of actual power consumption, and estimate
Meter threshold range [- 10% ,+10%].
Beneficial effects of the present invention are：
(1) present invention is from numerical quantization, trend deciphering, three kinds of aspects of association analysiss, bent using typical " S " type growth
Electricity growth trend curve after line fitting Business Process System, and the method analysis industry industry using principle component analysis expands electricity trend；
(2) present invention is changed and is decomposed with applying to install to Analyzing Total Electricity Consumption using census X12 algorithms, analysis report
Dressing amount and the association analysiss of Analyzing Total Electricity Consumption, are respectively used to prefectures and cities, different electricity consumption classifications, different industries, each voltage etc.
Level and each industry carry out the analysis of Multi-angle omnibearing, can obtain the capacity release rule of quantization and apply to install capacity and use
The relation of family electricity, so as to accurately carry out electricity anticipation；
(3) present invention can be compared with actual power consumption according to the anticipation result of electricity, and reasonable judgement is
It is no to have user to carry out the dangerous electricity consumption behaviors such as electric leakage of sneaking current.
Description of the drawings
Fig. 1 is by the newly-increased number percent pie chart of industry statistic；
Fig. 2 is by the newly-increased number percent pie chart of electricity consumption classification statistics；
Fig. 3 is by the newly-increased number percent pie chart of industrial statistics；
Fig. 4 is electricity consumption trend after big commercial power high pressure new clothes；
Fig. 5 (a) is big commercial power transient growth rate schematic diagram；
Fig. 5 (b) is big commercial power relative growth rate schematic diagram；
Fig. 6 is that big commercial power applies to install growth curve；
Fig. 7 is whole industry power consumption and apply to install capacity scatterplot；
Fig. 8 is whole industry power consumption exploded view；
Fig. 9 is that the whole industry applies to install capacity exploded view.
Specific embodiment：
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
Feature analysiss：
1) Add User amount and newly-increased power consumption equipment capacity.Newly-increased quantity and its proportion of total quantity will be accounted for predictive of future
The impetus that power sales increase, also reflects the effect of sales department market development to a certain extent.
Index 1：Newly-increased number percent UR=increases quantity/always newly and increases quantity (1.1) newly
Index 2：Newly-increased capacity rate of increase CR=newly added equipments capacity/total capacity (1.2)
Index 3：New volume reduction amount rate of increase DR=newly subtracts place capacity/total capacity (1.3)
Index 4：Newly-increased capacity sequential growth rate CGR=(newly added equipment capacity-last year newly added equipment capacity)/last year increases newly
Place capacity (1.4)
Index 5：New volume reduction amount sequential growth rate DGR=(newly subtract place capacity-last year newly subtract place capacity)/last year newly subtracts
Place capacity (1.5)
2) total numerical quantity is held, while grasping the composition situation of total amount.Can be from different regions, different electricity consumption classification, no
Electric pressure of the same trade, each and the multiple angle analysis total amounts of each industry and its constitute the relation of component.By such analysis,
Sales department can be expected that the variation tendency of the key indexs such as power sales future income, average price, profit.
1.2.2 trend is understood：
It is widely used in describing vegeto-animal growth course before growth curve, when its correlation analysis is substantially difference
The informix of phase is into a few parameters.Growth curve can be divided three classes：One is the equation for representing diminishing returns performance, is such as referred to
Number function；Another is the smooth S types curve of description, has equation such as Logistic, Gompertz of a fixed flex point, also a class
Smooth S types curve, but the variable equation of flex point, such as Von Bertanlanffy are also described.
Business Process System (based on new clothes) is completed by the visible client of trend analysiss, and electricity consumption trend presents S sigmoid growth curves afterwards
Characteristic, obtains industry using principle component analysis and expands electricity consumption growth curve, and with transient growth rate as base value, flex point is basis of characterization, real
The deciphering stage by stage of existing curve, such that it is able to analyze apply to install after the completion of capacity release rule.
1st, S sigmoid growth curves
S type curves are widely used in the analysis of vegeto-animal growth rhythm, are the curves for describing biological growth trend,
Growth curve is, wherein most representative there are 3 kinds of curve models：Logistic, Gompertz and Von Bertanlanffy,
Every kind of curve is all Richard curve Qt=α (1- β e-kt)1/(1-γ)Special expression form, its mathematical model and reference index
It is shown in Table 1.1.
Three kinds of nonlinear models of 1.1 Curve of growth fitting of table and its characteristic
Wherein, QtFor electricity；α is maximum electricity；β is parameter；K is transient growth speed；T is time scale, the present invention
With month as cycle.In typical " S " type, the flex point in curve is exactly the turning point of curvilinear motion speed to curvilinear trend, and
In mathematical model, second dervative is zero point, can characterize the variation tendency of curve, when second dervative is zero, is reached respective
Flex point month, its corresponding electricity are flex point electricity.The flex point electricity of each model, flex point month and maximum cycle increment
Mathematical model is as shown in table 1.1.Analyze Logistic points of inflexion on a curve electricity be α/2, be the half of maximum electricity, and occur
In (ln β)/k months；Gompertz knee of curve months are consistent with Logistic curves, and flex point electricity is α/e, equivalent to
36.8% α；Von Bertanlanffy curvilinear trends are relatively gentle, flex point month in (3 β of ln)/k, at that time charge value be 8 α/
27, equivalent to 29.6% α.
Expand electricity consumption growth curve characteristic to further analyze industry, except common counter flex point time and flex point value it
Outward, it is also contemplated that other indexs：
1) transient growth rate, embodies emphatically the variation tendency of the speed of growth, withFor basis；
2) relative growth rate, in the unit interval, initial value accounts for the ratio of last value, embodies effective growth of net power consumption.
The trend characteristic of electricity consumption can be emphasized by above index from growth change rate aspect, with transient growth rate as base
Value, weighs the wave characteristic of curve, compares the flex point that curve model is base value and judges, it is easier to finds the inherent law of curve.
2nd, principle component analysis
Principle component analysis are a kind of Multielement statistical analysis methods, multi objective are converted into several aggregative indicatores, by difference
Raw information is gathered by angle analysis variable using linear combination, is formed orthogonal and is covered most numbers
It is believed that the aggregate variable of breath.It is applied to industry and expands growth curve using the electricity of different clients as original variable, carries out between timesharing, divides
The two-way extraction of client, so obtaining pivot preferably could lift the typical customers chosen to the integral layer representated by client
Come on face.
The sample space for assuming data is s, and the index number of the observation of each sample is m, the mathematical modulo of principle component analysis
Type is represented by：
Wherein output matrix y1, y2... ymFor the key message of initial data, it is exactly pivot, coefficient matrices A refers to original
Coefficient of association between variable and pivot, state matrix x1, x2... xmFactor variable, that is, original variable are referred to, is herein referred to
The power consumption of different industries client.
The thought of pivot is exactly to turn parts into the whole, and is simplified, and multidimensional, polytomy variable are integrated according to particular demands,
Realize dimensionality reduction, depression of order.During the asking for of pivot, population variance is constant, and the first pivot refers to the larger unit of variance ratio, and which deviates flat
Weighing apparatus point distance it is bigger, dispersion degree is bigger, comprising data message it is more；Second pivot refers to time unit of big variance；Etc.
The like.Then one threshold value (such as 90%) of setting is extracted to pivot, only to accounting for n pivot of whole variance proportions 90%
Analysis, thus m original aggregation be n index.
1.3.3 association analysiss
The prediction of electricity is the one of which content of load forecast, and careful accurate prediction is only carried out to electricity
The prediction of other indexs such as electricity, load can be called to provide reference frame for system.Business Process System has for the prediction work of electricity
Leading meaning, is direct acting factor, the relation that both a difinite quality is closed for a long time, it is fixed that no effective method is realized
Amount is analyzed and sets up model.
Due to power consumption and apply to install all there is certain seasonality, power consumption and applying to install is changed, the present invention is utilized
Census X12 algorithms carry out seasonal decomposition to explanatory variable and dependent variable.Original variable is decomposed into into trend circulation item
(TC), seasonal factor (SF) and random entry (IR).Linear regression is carried out again by analyzing the dependency between two variable trends items.
After seasonal factor is considered, final predictive value is obtained.Capacity will be applied to install and user's electricity will be connected, and complete to align
True prediction.
Using month degree as time observation unit time serieses generally with the cyclically-varying in units of year, this is
Caused due to seasonal factor impact, referred to as seasonal variety.Seasonal fluctuations are not only due to directly affecting for climatic factor, together
When social system and social mores there is also seasonal move.Due to seasonal fluctuation highly significant, it will usually cover developing
Objective law, impacts to our analysis and prediction work, it is therefore necessary to remove the impact of seasonal fluctuation, by season key element
Reject from former sequence, carry out seasonal adjustment.
1.3 instance analysis
1.3.1 feature analysiss
1.3.1.1 terrain analysis
A in 2013 saved into 17 districts and cities apply to install situation and counted, including it is newly-increased apply to install amount, existing capacity, newly-increased capacity,
New volume reduction amount.Calculate the five indices introduced in 1.2.1 according to statistical conditions respectively, count in table 1.2.
1.2 A of table saves 17 districts and cities' Business Process System situation statistical tables
Index 1 increases number percent newly：
The quantity that Adds User of five districts and cities of A7, A13, A2, A14 and A6 is most as can be seen from the table, account for total newly-increased
More than half of amount, it is seen that economic growth rate is very fast.
Index 2 increases capacity rate of increase newly：
In the capacity of increase, A6, A2, A17 comprehensively account for existing capacity large percentage for 3 years, and the capacity of increase is more.
3 new volume reduction amount rate of increase of index：
It can be seen that the volume reduction amount of A6, A4, A12 is relatively more in new volume reduction amount.It can be seen that the newly-increased volume reduction amounts of A6 all compare many.
Index 4, index 5 increase/volume reduction amount sequential growth rate newly：
As can be seen from the table, the newly-increased capacity of A1, A2, A3, A4, A5 is increased continuously.
1.3.1.2 category of employment
1.3 A of table saves eight big industry Business Process System situation statistical tables
Index 1 increases number percent newly：
Fig. 1 presses the newly-increased number percent pie chart of industry statistic
As can be seen from the figure industry accounts for half of total newly-increased amount or so, it is seen that industry saves whole electricity consumption situation in A
Proportion is larger.
Index 2 increases capacity rate of increase newly：
It can be seen that information transfer, Computer Service and software industry increase quickly in the capacity of increase, existing capacity ratio
Larger, the capacity of increase is more.
3 new volume reduction amount rate of increase of index：
It can be seen that agriculture, forestry, animal husbandry, fisheries and industry are more in the reduction Capacity Ratio of 3 years in new volume reduction amount, it is seen that the 3rd
Industry development develops more rapid compared with agro-industry.
Index 4, index 5 increase/volume reduction amount sequential growth rate newly：
As can be seen from the table, every profession and trade volume change does not continuously increase and reduces.
1.3.1.3 electricity consumption classification
1.4 A of table saves electricity consumption classification Business Process System situation statistical table
Index 1 increases number percent newly：
Fig. 2 is by the newly-increased number percent pie chart of electricity consumption classification statistics
As can be seen from the figure the new clothes situation of general industry and commerce accounts for more than 70 the percent of all electricity consumption classifications, from
Electric situation is it can also be seen that general industry and commerce is developed rapidly.
Index 2 increases capacity rate of increase newly：
It is it can be seen that the agricultural production electricity consumption new clothes proportion that accounts for existing capacity is increasing in the capacity of increase, general industrial and commercial
Industry sustainable growth.
3 new volume reduction amount rate of increase of index：
It can be seen that the capacity that reduces every year of agricultural drainage and irrigation and agricultural production electricity consumption is relative with original total appearance in new volume reduction amount
Amount is a lot.But the newly-increased capacity of agricultural production electricity consumption is also a lot, therefore simply user changes more frequent.
Index 4, index 5 increase/volume reduction amount sequential growth rate newly：
As can be seen from the table, only agricultural production was with newly-increased capacity sustainable growth in electrically continuous 2 years.
1.3.1.4 electric pressure
1.5 A of table saves each electric pressure Business Process System situation statistical table
Index 1 increases number percent newly：
10kv new clothes account for the overwhelming majority of all new clothes quantity.
Index 2 increases capacity rate of increase newly：
It can be seen that the equal sustainable growth of newly-increased capacity of 220kv, 10kv in the capacity of increase.
3 new volume reduction amount rate of increase of index：
It can be seen that new volume reduction amount 10kv electric pressure is also relatively most in new volume reduction amount, in quantity, feelings are applied to install
In condition it can be seen that 10kv be it is most active in all electric pressures be also most complicated.
Index 4, index 5 increase/volume reduction amount sequential growth rate newly：
As can be seen from the table, continuous 2 years of no electric pressure it is new/subtract increase-volume amount sustainable growth.Notice 220kv rings
Exist than index and be mutated, be due in data 11 years new volume reduction amounts compare 12,13 years it is much smaller, it may be possible to because of situation of applying to install by
Prefectures and cities are manually entered has very big random factor.
1.3.1.5 industry is divided
According to industry and the membership relation of industry, primary, secondary and tertiary industries are divided.
The division relation of 1.6 industry of table and every profession and trade
| The primary industry | Agriculture, forestry, animal husbandry and fishery |
| Secondary industry | Industry |
| The tertiary industry | All industries in addition to the industry that the first secondary industry is included |
1.7 A of table is saved with product Business Process System situation statistical table
Index 1 increases number percent newly：
As can be seen from Figure 3 secondary industry new clothes account for all new clothes quantity half in quantity.
It can be seen that the capacity of each industry reduces regular obvious in new volume reduction amount, but on the whole it can be seen that capacity
Reduction situation slowed down.
Index 4, index 5 increase/volume reduction amount sequential growth rate newly：
As can be seen from the table, the newly-increased volume reduction amount of the tertiary industry from the point of view of the situation of nearest 3 years is all continuously increased,
It can be seen that with economic development, the tertiary industry plays more and more active role in the whole industry.
1.3.2 trend is understood -- apply to install rear power consumption trend analysiss
1st, typical user applies to install rear power consumption trend analysiss
Due to being affected by national policy, production cost, sales situations, market prospect, industry expansion trend, it is impossible to expand new to industry
Increase-volume amount, electricity, load are made accurate prediction and are judged.Again, industry expands new clothes needs longer time process, current technology
The restriction of means and decision method, it is impossible to accurately hold the release rule of newly-increased capacity, the rule of electricity growth point, capacity
The impact of growth to the whole province's electricity etc..For new clothes business, client needs to carry out each side such as electrical equipment after completing to apply to install
The debugging in face, can not reach stable use electricity condition at once, and the power consumption of this section of limber up period client is our weights to be analyzed
Point.
For this purpose, the present invention saves the Business Process System data of 2011 to parts of in August, 2014 as foundation with A, with high pressure new clothes industry
Three typical enterprises of big commercial power in data as a example by business, are chosen, the power consumption rule in its new clothes latter year is analyzed.By right
The trend analysiss of typical enterprise, as shown in Figure 4：3 typical enterprises complete the substantially S-type growth characteristics of electricity consumption after new clothes business,
Also comply with the trend of biological growth.
2nd, typical customers industry expands Curve of growth fitting
The present invention chooses the new clothes business that electricity consumption classification is big commercial power, sieve with sales department of A provinces metric data as comedy
The power consumption of 12 months after power transmission is summarized in choosing, is sorted by electricity, removes the data that customer electricity curve does not meet rule, such as because
It is that the Market Reasons underproduction or policy implication are carried out energy-saving and emission-reduction and cause what electricity declined, is more than on the basis of zero, most by electricity chain rate
After filter out ten typical customers, represent electricity consumption trend after big commercial power new clothes.
Electricity after applying to install to ten clients adopts Logistic, Gompertz and Von Bertanlanffy models, profit
Iteration is circulated with SPSS16.0 statistical analysis softwares, the Fitting Calculation goes out optimal estimation value A, B of each model parameter, K, is restrained
Standard is 10-8, and flex point month, flex point electricity and the degree of fitting R2 of model are extrapolated according to estimates of parameters, is shown in Table 1.8.
1.8 growth curves model estimates of parameters of table and degree of fitting
It is excessive it can be seen that part garbled data still suffers from certain mistake, possible cause by part flex point month in form
Mistake or enterprise are manually entered when being due to statistics while carrying out the industry expansion project of other classifications, power consumption are affected, is made part point
Analysis data have obvious careless mistake.But it can be seen that some useful conclusions：In ten clients, there are six clients to Von
Bertanlanffy models fittings degree preferably, removes flex point month excessive enterprise three and enterprise seven, although each enterprise it is inclined
Good model is not quite similar, but flex point month is substantially at 3.3 months.
3rd, pivot is extracted
Expand growth curve to Von Bertanlanffy models fittings industry using SPSS16.0 statistical analysis softwares to lead
Unit extracts.Principal component scores value table can be obtained, 1.9 are shown in Table.Then the pivot electricity of the 1-12 months after new clothes is worth to according to fitting
Amount, is shown in Table 1.10.
1.9 principal component scores value table of table
1.10 pivot electricity of table
4th, big industrial electric industry expands Curve of growth fitting
Logistic, Gompertz and Von Bertanlanffy models fittings, three kinds of models pair are carried out to pivot electricity
The fitting effect of pivot electricity is all fabulous, and Logistic model-fitting degrees are 0.998, Gompertz and Von Bertanlanffy
Model-fitting degree is 1.Von Bertanlanffy model flex point months are April, and flex point electricity is 623.45 ten thousand kwh.For
Von Bertanlanffy models carry out the deciphering of indices, draw transient growth rate and relative growth rate change under the model
Situation, such as Fig. 5 (a), Fig. 5 (b).
Analysis increment percent model understands that Von Bertanlanffy models were in accelerated growth phase before May, located afterwards
In deceleration trophophase.Can be seen which gathers way within accelerated growth phase constantly slowing down, nearby instantaneous rate of increase reaches flex point
Peak value, is not further added by.Constantly decline in deceleration trophophase instantaneous rate of increase and finally may tend to zero, power consumption tends towards stability.
For the Von Bertanlanffy models that degree of fitting is 1, the overall growth curve of big commercial power is shown in Fig. 6.Plus
Fast-growing is the 1-5 months for a long time, and deceleration trophophase is the 6-12 months, finally stable in 13,340,000 kwh.Early stage monthly puts into electricity and is respectively
7.94%th, 22.39%, 38.81%, 55.98%, 69.42%.It can be seen that big commercial power just can be realized substantially at five months newly
The release of dressing amount, in terms of electrical equipment debugging, progress is very fast, into steady statue.
1.3.3 association analysiss -- apply to install capacity and electricity relation
Due to power consumption and apply to install all there is certain seasonality, power consumption and applying to install is changed, that is, is utilized
Census X12 algorithms carry out seasonal decomposition to explanatory variable and dependent variable.Original variable is decomposed into into trend circulation item
(TC), seasonal factor (SF) and random entry (IR).Linear regression is carried out again by analyzing the dependency between two variable trends items.
After seasonal factor is considered, final predictive value is obtained.Capacity will be applied to install and user's electricity will be connected, and complete to align
True prediction.
Analyses and prediction based on Business Process System whole industry power consumption
Statistics A saves each monthly whole industry of -2014 years first half of the year in 2011 and applies to install capacity and power consumption situation, observes two fingers
Target scatterplot simultaneously finds corresponding relation.
1. pair dependent variable decomposes
By the whole industry power consumption of -2014 years first half of the year in 2011 (Power usage of the whole industry)
Dependent variable QHY is decomposed into trend circulation item QHY_TC, seasonal factor QHY_SF and random using the X12 algorithms in E-view softwares
Item QHY_IR.As shown in Figure 8.Trend circulation item grows steadily, it can be seen that the expanding economy over time of whole industry power consumption
Constantly it is continuously increased.Seasonal factor is presented low ebb in 2 months in regular change, and this is to be stopped due to the first month of the lunar year whole industry on a large scale
Produce, cause power consumption to decline.
2. pair independent variable decomposes
The whole industry of -2014 years first half of the year in 2011 is applied to install into capacity (expansion of the whole industry)
Independent variable EWI is decomposed into trend circulation item EWI_TC, seasonal factor EWI_SF and random using the X12 algorithms in E-view softwares
Item EWI_IR.As shown in Figure 9.There is slight downward trend the second half year in 2012 from the visible capacity of applying to install of cyclical trend item, can
To learn that the whole industry integrally remained basically stable the trend that do not rise appreciably in the second half year in 2012 to new clothes capacity in 2013.At random
Item is followed substantially applies to install curve, and this is that have very strong randomness due to Business Process System itself, adds each electric company's logging data
When there is very big unstable factor.
3. trend circulates item correlation analysiss and linear regression
Independent variable trend circulation item (EWI_TC) and dependent variable trend circulation item (QWI_TC) are carried out into correlation analysiss.Meter
Calculate result such as table 1.11.All result of calculations are all significantly correlated under unilateral 0.01 test level.And dependency is preferable, because
When variable delayed 3 cycles, the dependency of two variables is most strong.
When dependency is most strong be in delayed 6 months of independent variable trend term afterwards.
1.11 liang of variable trends of table circulate item correlation coefficient
Using EWI_TC (- 3) as independent variable, QWI_TC carries out one-variable linear regression as dependent variable.Obtain formula
4.8.1.There is no serial correlation and Singular variance in the residual error of the regression model for obtaining, R2 is higher, can be very good fitted trend
The relation of cyclic variable.In formula, 0.216 can represent increasing quantity coefficient, and constant is represented and can deposit electricity value.
YTC=0.216XTC(-S)+2144598.512
R2=0.885 (1.7)
4. power quantity predicting
The whole industry electricity trend circulation item match value that above-mentioned Equation for Calculating is obtained with decompose the seasonal factor that obtains and
Random entry is summed up, and obtains whole industry electricity match value, and actual comparison, and two curve co-insides degree are very high, and turning point also may be used
Substantially it coincide.
The electricity consumption of in May, 2014 for obtaining is fitted for 2986527.214 ten thousand kilowatt hours by above method, actual power consumption is
3067010.534 ten thousand kilowatt hours, error are 2.62%；June predicts 3019921.841 ten thousand kilowatt hour of power consumption, actual electricity consumption
Measure 3079281.457 kilowatt hours, error 1.93%.As the monthly prediction of short-term forecast, future can be accurately predicted
The electricity of 2-3 month, prediction afterwards should re-start modeling work.
Although the above-mentioned accompanying drawing that combines is described to the specific embodiment of the present invention, not to present invention protection model
The restriction enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not
The various modifications made by needing to pay creative work or deformation are still within protection scope of the present invention.
Claims (10)
1. a kind of user's power stealing for considering Business Process System data, electric leakage behavior monitoring method, is characterized in that：Comprise the following steps：
(1) determine Add User amount and newly-increased power consumption equipment capacity, calculate capacity rate of increase and sequential growth rate；
(2) the electricity consumption trend after Business Process System is built using Von Bertanlanffy growth curves, obtained using principle component analysis
Expand electricity consumption growth curve to industry, with transient growth rate as base value, flex point is basis of characterization, carries out the deciphering stage by stage of curve；
(3) power consumption and Business Process System are changed, is carried out to explanatory variable and dependent variable using census X12 algorithms season
Section property is decomposed, and original variable is decomposed into trend circulation item, seasonal factor and random entry；
(4) dependency analyzed between two variable trends items carries out linear regression again, it is considered to after seasonal factor, confirms power consumption
Predictive value；
(5) predictive value by the actual power consumption of each user with its power consumption is compared, if its difference exceedes estimates threshold value,
Then judge that the user has abnormal electricity consumption, examined accordingly and early warning.
2. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (1), by calculating newly-increased quantity and capacity and its rate of increase, to confirm that what following electrity market increased becomes
Gesture.
3. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (1), hold total numerical quantity, while grasp the composition situation of total amount, from different regions, different electricity consumption classifications,
The multiple angle analysis total amounts of different industries, each electric pressure and each industry and its constitute the relation of component.
4. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (2), the growth for confirming power consumption based on maximum electricity, parameter, transient growth speed and time scale becomes
Gesture.
5. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (2), the flex point in curve is exactly the turning point of curvilinear motion speed, and the point that second dervative is zero, its
The variation tendency of curve is characterized, when second dervative is zero, respective flex point month is reached, its corresponding electricity is flex point electricity
Amount.
6. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (3), using the electricity consumption trend curve after Business Process System, with reference to instantaneous growth rate and relative growth rate, from life
Long rate of change aspect emphasizes the trend characteristic of electricity consumption, with transient growth rate as base value, weighs the wave characteristic of curve.
7. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (3), using principle component analysis by different angle analysis variables, using linear combination by original letter
Breath is gathered, and forms aggregate variable that is orthogonal and covering most data information；
In the step (3), it is applied to industry and expands growth curve using the electricity of different clients as original variable, carry out between timesharing,
Divide the two-way extraction of client, obtain pivot and the typical customers chosen are lifted in the overall aspect representated by client, with industry
Expansion applies to install data as explanatory variable, and power consumption is used as dependent variable.
8. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (3), it is assumed that the sample space of data is s, the index number of the observation of each sample is m, principle component analysis
Mathematical model be represented by：
Wherein output matrix y1, y2... ymFor the key message of initial data, it is exactly pivot, coefficient matrices A refers to original variable
Coefficient of association between pivot, state matrix x1, x2... xmFactor variable, that is, original variable are referred to, difference is herein referred to
The power consumption of industry customer.
9. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, its feature
It is：In the step (4), power consumption and apply to install all there is certain seasonality, using month degree as time observation unit when
Between sequence generally with the cyclically-varying in units of year, this affects to cause due to seasonal factor, referred to as seasonal to become
Change, when objectivity influence factor is analyzed, season key element is rejected from former sequence, seasonal adjustment is carried out.
10. a kind of user's power stealing for considering Business Process System data as claimed in claim 1, electric leakage behavior monitoring method, which is special
Levying is：Calculate actual power consumption and estimate the difference of power consumption and the ratio of actual power consumption, and estimation threshold range [- 10%,
+ 10%].
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610861397.2A CN106651425A (en) | 2016-09-28 | 2016-09-28 | User electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610861397.2A CN106651425A (en) | 2016-09-28 | 2016-09-28 | User electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106651425A true CN106651425A (en) | 2017-05-10 |
Family
ID=58854753
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610861397.2A Pending CN106651425A (en) | 2016-09-28 | 2016-09-28 | User electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106651425A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111784379A (en) * | 2020-05-19 | 2020-10-16 | 北京中电普华信息技术有限公司 | Estimation method and device for additional payment electric charge and screening method and device for abnormal cases |
| CN113610409A (en) * | 2021-08-12 | 2021-11-05 | 北京中电普华信息技术有限公司 | Early warning method and device for electric charge recovery risk |
| CN114676896A (en) * | 2022-03-16 | 2022-06-28 | 佰聆数据股份有限公司 | Electric power operation monitoring method and system based on time decomposition improved algorithm |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005182528A (en) * | 2003-12-19 | 2005-07-07 | Tokyo Electric Power Co Inc:The | Economic analysis device, method, and program, and storage medium storing program |
| CN104537433A (en) * | 2014-12-18 | 2015-04-22 | 国网冀北电力有限公司 | Sold electricity quantity prediction method based on inventory capacities and business expansion characteristics |
| KR20150064448A (en) * | 2013-12-03 | 2015-06-11 | 에너지관리공단 | Method for calculating expected amount of greenhouse gas emission or energy consumption per facility and computer readable recording media storing program for executing method thereof |
| CN105023066A (en) * | 2015-07-31 | 2015-11-04 | 山东大学 | Business expansion analytical prediction system and method based on seasonal adjustment |
| CN105260802A (en) * | 2015-11-06 | 2016-01-20 | 国网冀北电力有限公司 | Monthly electric quantity prediction method based on correction of business expansion growth curve and season adjustment |
| CN105404935A (en) * | 2015-11-11 | 2016-03-16 | 国网浙江省电力公司经济技术研究院 | Electric power system monthly load prediction method considering business expansion increment |
-
2016
- 2016-09-28 CN CN201610861397.2A patent/CN106651425A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005182528A (en) * | 2003-12-19 | 2005-07-07 | Tokyo Electric Power Co Inc:The | Economic analysis device, method, and program, and storage medium storing program |
| KR20150064448A (en) * | 2013-12-03 | 2015-06-11 | 에너지관리공단 | Method for calculating expected amount of greenhouse gas emission or energy consumption per facility and computer readable recording media storing program for executing method thereof |
| CN104537433A (en) * | 2014-12-18 | 2015-04-22 | 国网冀北电力有限公司 | Sold electricity quantity prediction method based on inventory capacities and business expansion characteristics |
| CN105023066A (en) * | 2015-07-31 | 2015-11-04 | 山东大学 | Business expansion analytical prediction system and method based on seasonal adjustment |
| CN105260802A (en) * | 2015-11-06 | 2016-01-20 | 国网冀北电力有限公司 | Monthly electric quantity prediction method based on correction of business expansion growth curve and season adjustment |
| CN105404935A (en) * | 2015-11-11 | 2016-03-16 | 国网浙江省电力公司经济技术研究院 | Electric power system monthly load prediction method considering business expansion increment |
Non-Patent Citations (2)
| Title |
|---|
| 余向前 等: "基于生长曲线的行业业扩用电趋势研究", 《电力需求侧管理》 * |
| 葛斐 等: "基于业扩报装的全社会电量预测方法研究", 《安徽电气工程职业技术学院学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111784379A (en) * | 2020-05-19 | 2020-10-16 | 北京中电普华信息技术有限公司 | Estimation method and device for additional payment electric charge and screening method and device for abnormal cases |
| CN111784379B (en) * | 2020-05-19 | 2023-09-15 | 北京中电普华信息技术有限公司 | Estimation method and device for electric charge after-payment and screening method and device for abnormal cases |
| CN113610409A (en) * | 2021-08-12 | 2021-11-05 | 北京中电普华信息技术有限公司 | Early warning method and device for electric charge recovery risk |
| CN114676896A (en) * | 2022-03-16 | 2022-06-28 | 佰聆数据股份有限公司 | Electric power operation monitoring method and system based on time decomposition improved algorithm |
| CN114676896B (en) * | 2022-03-16 | 2023-06-16 | 佰聆数据股份有限公司 | Power operation monitoring method and system based on time decomposition improvement algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106447198A (en) | Power consumption checking method based on business expanding installation data | |
| CN106447108A (en) | Power utilization demand analysis prediction method taking business-expansion installation data into consideration | |
| CN104123600B (en) | A kind of electric power manager's exponential trend Forecasting Methodology towards representative row sparetime university data | |
| Galán et al. | Inefficiency persistence and heterogeneity in Colombian electricity utilities | |
| CN110689279A (en) | System and method for analyzing potential safety hazard of residential electricity consumption based on power load data | |
| Tang et al. | GM (1, 1) based improved seasonal index model for monthly electricity consumption forecasting | |
| CN104537433A (en) | Sold electricity quantity prediction method based on inventory capacities and business expansion characteristics | |
| CN106651425A (en) | User electricity stealing and electricity leakage behavior monitoring method considering business expanding installation data | |
| CN111798333A (en) | Energy utilization evaluation and electricity utilization safety analysis method and system | |
| CN111126499A (en) | Secondary clustering-based power consumption behavior pattern classification method | |
| CN111127099A (en) | E-commerce user analysis system based on big data and analysis method thereof | |
| Xie et al. | The scale effect in China's power grid sector from the perspective of malmquist total factor productivity analysis | |
| Yang et al. | Innovation and Market Value in Newly‐Industrialized Countries: The Case of Taiwanese Electronics Firms | |
| CN105825290A (en) | Electric quantity prediction method based on industrial chain product output | |
| Sari et al. | The effectiveness of hybrid backpropagation Neural Network model and TSK Fuzzy Inference System for inflation forecasting | |
| CN109118250A (en) | Electricity market main body act of unfair competition evaluation method and device | |
| CN111724049B (en) | Research and judgment method for potential electric power energy efficiency service clients | |
| CN114169802A (en) | Power grid user demand response potential analysis method, system and storage medium | |
| Popeangă | Data mining smart energy time series | |
| Davarzani et al. | Study of missing meter data impact on domestic load profiles clustering and characterization | |
| Liu et al. | Efficient electricity sales forecasting based on curve decomposition and factor regression | |
| Fadillah et al. | Application of The Sequential Hot-deck Imputation Method for Identification of Indonesian Standard Classification of Business Fields (KBLI) | |
| Tao et al. | Power consumption behavior analysis for customer side flexible resources based on data mining | |
| Chunshan et al. | Study and application of data mining and NARX neural networks in load forecasting | |
| Fu et al. | On a class of estimation and test for long memory |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170510 |
|
| RJ01 | Rejection of invention patent application after publication |