CN103487682A - Method for early warning of sensitive client electric energy experience quality under voltage dip disturbance - Google Patents

Abstract

The invention provides a method for early warning of sensitive client electric energy experience quality under the voltage dip disturbance. The method comprises the steps that based on the S conversion rapid algorithm and an increment SVM classifier, voltage dip disturbances of sensitive clients are automatically identified; based on identification results of the voltage dip disturbances, voltage tolerance curves of devices corresponding to multiple types of sensitive clients at different load levels are determined; historical monitoring data of the voltage dip disturbances serve as samples, the samples are converted into sample values of a voltage dip amplitude ponderance index MSI and a lasting time ponderance index DSI, a probability density function of the MSI and the DSI is determined on the basis of the maximum entropy principle, the sensitive device fault probability is evaluated, and the probabilities of the sensitive devices corresponding to the sensitive clients at the voltage dip level are obtained. By the adoption of the method for early warning of sensitive client electric energy experience quality under the voltage dip disturbance, the electric energy quality disturbance condition can be accurately monitored, whether a client load is influenced by the disturbance or not is determined according to the load sensitivity degree of each client, and potential risks of load operation are found.

Description

The method for early warning of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance

Technical field

The present invention relates to the power technology field, relate in particular to the method for early warning of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance.

Background technology

In recent years, be accompanied by modern industry and expanding economy, the sensitive load such as computing machine, power electronic equipment is in the industry-by-industry widespread use, cause power customer very responsive to voltage dip, voltage swells, interruption (interruption) in short-term etc., the fault of individual equipment or element may cause great economic loss.On the other hand, the undulatory property such as some electric arc furnaces, rectifier, single-phase load, high power motor or impact load are connected to the grid, and the particularly injection of a large amount of harmonic waves and order harmonic components causes the serious distortion of voltage and current waveform in electrical network.In addition, some external factor also can be disturbed the normal operation of electric system as thunder and lightning, outside destroy, branch impact, grid equipment fault etc., cause power quality problem to happen occasionally.Load to the voltage dip sensitivity in electric system at present is more and more, voltage dip has become the main cause that causes voltage-sensitive equipment cisco unity malfunction, and investigation shows that the client's loss caused by voltage dip in all kinds of power quality problems accounts for more than 80% of quality of power supply loss.

In actual applications, the different characteristic and to requirement and the susceptibility of the quality of power supply according to power load, generally be divided into power load common load, sensitive load.Some electric system clients have used a large amount of sensitive loads, are referred to as the responsive client of the quality of power supply.The outwardness of power quality problem and the sensitivity characteristic of sensitive load cause responsive client's electricity usage to have great risk.For this class client, even slight power quality problem occurs, also can cause serious economic loss.

Electrical energy power quality disturbance (PQD) signal form numerous and complicated, how correctly extracting the characteristic quantity of disturbing signal and how identifying accurately the disturbing signal type becomes the matter of utmost importance solved with the raising quality of power supply.In addition, client's electric energy Quality of experience not merely depends on power supply quality, also and client's sensitivity be closely related.

Summary of the invention

The present invention is based on S conversion fast algorithm and increment svm classifier device and realize the automatic identification to voltage Sag Disturbance, and use the failure rate of maximum entropy method estimating user sensitive load under voltage Sag Disturbance.

Concrete, the method for early warning of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention comprises:

Based on S conversion fast algorithm and increment svm classifier device, responsive client's voltage Sag Disturbance is identified automatically;

Result based on described voltage Sag Disturbance identification, determine the voltage-tolerance curve of equipment under the different loads level corresponding to the responsive client of several classes;

According to the Historical Monitoring data of voltage Sag Disturbance as sample, be converted into the sample value of amplitude seriousness index MSI and the duration seriousness index DSI of voltage dip, and determine the probability density function of MSI and DSI based on principle of maximum entropy, assessment sensitive equipment probability of malfunction, show that sensitive equipment corresponding to responsive client is at the probability broken down under the voltage dip rank.

Wherein, describedly based on S conversion fast algorithm and increment svm classifier device, responsive client's voltage Sag Disturbance is identified automatically, being comprised:

Determine that responsive client, as monitoring point, carries out Real-Time Monitoring to its electrical energy power quality disturbance, obtain sensitive equipment voltage signal, current signal as monitoring point;

For voltage signal and the current signal of monitoring point, carry out feature extraction based on S conversion fast algorithm, extract the proper vector of mode coefficient corresponding to standard deviation, maximum norm coefficient and the rated frequency of each frequency band mode coefficient as disturbing signal;

Described proper vector is inputed in increment svm classifier device and carries out Classification of Power Quality Disturbances, automatically identify voltage Sag Disturbance.

Wherein, described S conversion fast algorithm is as follows:

wherein, the broad sense window function that ω (τ-t, σ) is all frequency v in unit area.

Wherein, described proper vector is inputed in increment svm classifier device and carries out Classification of Power Quality Disturbances, automatically identify voltage Sag Disturbance, comprising:

Proper vector is carried out to the time window division, use incremental learning mechanism, the proper vector arrived in a time window is carried out to study in batch;

In nearest n time window, if the number of times that proper vector sample does not become support vector is deleted it over ξ from training set.

Wherein, described sensitive equipment comprises FPGA (Field Programmable Gate Array) control PLC, variable speed drivP ASD, computer PC, A.C. contactor ACC.

Wherein, described maximum entropy model is:

$\begin{array}{c}\max H\left(x\right)=-\underset{R}{\∫}f\left(x\right)\ln f\left(x\right)\mathrm{dx}\\ s.t.\\ \underset{R}{\∫}f\left(x\right)\mathrm{dx}=1\\ \underset{R}{\∫}\mathrm{xf}\left(x\right)\mathrm{dx}=E_1\\ \underset{R}{\∫}{(x-E_1)}^{h}f\left(x\right)\mathrm{dx}=E_h,h=\mathrm{2,3},...,N\end{array}$

Wherein, the stochastic variable that x is sensitive equipment voltage dip seriousness index MSI or DSI, the value border that R is variable x, the entropy that H (x) is stochastic variable, the probability density function that f (x) is stochastic variable x, E ₁and E _h1 rank unit point distance and the h rank centre distance for voltage dip seriousness index.

Wherein, the described probability density function of determining MSI and DSI based on principle of maximum entropy, assessment sensitive equipment probability of malfunction show that sensitive equipment corresponding to responsive client, at the probability broken down under the voltage dip rank, comprising:

Introduce Lagrangian in described maximum entropy model, and obtain the probability density function analytic expression by classical partial differentiation;

When voltage dip occurs in uncertain region (i, j), the failure rate P (i, j) of equipment is:

$P(i,j)=\underset{x_{\mathrm{MSI}}}{\overset{{\mathrm{\γ}}_{\stackrel{\‾}{i}}}{\∫}}f\left(x_{\mathrm{MSI}}\right){\mathrm{dx}}_{\mathrm{MSI}}\underset{x_{\mathrm{DSI}}}{\overset{{\mathrm{\γ}}_{\stackrel{\‾}{j}}}{\∫}}f\left(x_{\mathrm{DSI}}\right){\mathrm{dx}}_{\mathrm{DSI}}$

Wherein, $f\left(x\right)=\exp ({\mathrm{\λ}}_0+{\mathrm{\λ}}_{1}x+\underset{h=2}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\λ}}_h{(x-E_1)}^h),h=\mathrm{2,3,4,5}.$

with

respectively the residual voltage amplitude in the intermediate value of regional i and duration the intermediate value at regional j.

Implement the present invention, can more accurately monitor the electrical energy power quality disturbance situation, and determine that according to client's load-sensitive degree whether this disturbance likely has influence on client's load, finds the potential danger of load operation.On this basis, propose the concept that client's electric energy is experienced, to load operation, exist the client of potential danger to carry out early warning.This method is conducive to electric power enterprise and responsive client reduces for electricity consumption risk, technological transformation, and significant to differentiation customization electric power.

The accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, below will the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.

The schematic flow sheet that Fig. 1 is the method for early warning embodiment mono-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention;

The schematic flow sheet that Fig. 2 is the method for early warning embodiment bis-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention;

The schematic flow sheet that Fig. 3 is the method for early warning embodiment tri-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention;

The schematic flow sheet that Fig. 4 is the method for early warning embodiment tetra-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention;

The schematic flow sheet that Fig. 5 is the method for early warning embodiment five of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention;

The schematic flow sheet that Fig. 6 is the method for early warning embodiment six of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention;

The schematic flow sheet that Fig. 7 is the method for early warning embodiment seven of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance provided by the invention.

Embodiment

Below in conjunction with accompanying drawing, specific embodiments of the invention are elaborated.

As shown in Figure 1, the schematic flow sheet for the method for early warning embodiment mono-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance of the present invention mainly comprises the following steps:

Step 100, based on S conversion fast algorithm and increment svm classifier device, identify automatically to responsive client's voltage Sag Disturbance;

Step 101, the result based on described voltage Sag Disturbance identification, determine the voltage-tolerance curve of equipment under the different loads level corresponding to the responsive client of several classes;

Step 102, according to the Historical Monitoring data of voltage Sag Disturbance as sample, be converted into the sample value of amplitude seriousness index MSI and the duration seriousness index DSI of voltage dip, and determine the probability density function of MSI and DSI based on principle of maximum entropy, assessment sensitive equipment probability of malfunction, show that sensitive equipment corresponding to responsive client is at the probability broken down under the voltage dip rank.

As shown in Figure 2, be the schematic flow sheet of the method for early warning embodiment bis-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance of the present invention, the specific implementation process that in the present embodiment, the main temporal aspect of describing based on S conversion fast algorithm extracts comprises:

Step 200, this converts fast algorithm by S and introduces the Power Quality Disturbance feature extraction.Wherein, S conversion fast algorithm as the formula (1).

$S(\mathrm{\τ},v)=\underset{-\∞}{\overset{+\∞}{\∫}}g\left(t\right)\·\mathrm{\ω}(\mathrm{\τ}-t,\mathrm{\σ})\·e^{(-j2\mathrm{\πvt})}\mathrm{dt}---\left(1\right)$

In formula (1), the broad sense window function that ω (τ-t, σ) is all frequency v in unit area.

By this conversion, in one dimension S-territory, frequency and the time of its representative is determined in the position of certain value, it becomes one-dimensional vector by the two-dimentional time-frequency matrix compression of S conversion, the clock signal that is about to N point obtains only having the time-frequency vector of N point by fast algorithm, this method can effectively be avoided the bulk redundancy information in S conversion time-frequency matrix, for the characteristic component that accurately extracts signal is laid a good foundation, and saved computing time.

Step 201, processed quality of power supply Historical Monitoring data, specific as follows:

Sequential continues 8 nominal period 0.16s altogether, and sampling number is 1024, and sample frequency is 6.4kHz, and the maximum frequency that can detect is 3.2kHz.By two-dimentional relationship between frequency and time corresponding to S conversion fast algorithm gained one dimension mode coefficient, can obtain 1024 sequential points according to formula (2) and can form 11 frequency bands.In one dimension S-territory, frequency and the time of its representative is determined in the position of certain value, and a upper frequency range and the next frequency range relation of synchronization, and the division of concrete frequency band is suc as formula shown in (2).

Wherein, the time-domain signal sequence that is N for a length, n _indexfor a measuring point in S conversion one dimension mode coefficient, and meet 2<n _index<n-1, n _lfor the number of plies of current place frequency band, n ₁and n ₂for the starting and ending position of current frequency band at the one dimension mode coefficient.The initial position with end of the one dimension mode coefficient that interior each frequency of each frequency band is corresponding is identical.

Step 202, set up a proper vector that comprises sequential point, extracts the characteristic component of mode coefficient corresponding to standard deviation, maximum norm coefficient and the rated frequency of each frequency band mode coefficient as disturbing signal.

Concrete, each seasonal effect in time series time window is set to 8 nominal period usually, and the S Transform Module Matrixes has retained the amplitude information of signal, and in the S transformation matrix, the amplitude of element is with corresponding with the mould value of the S of frequency place conversion sometime.The amplitude that its column vector is a certain sampling instant of signal is with the distribution of frequency change, the time dependent distribution of amplitude that the row vector is a certain frequency of signal.

Step 203, be electrical energy power quality disturbance classification under its mark according to proper vector, is designated as T={ (n ₁(t _i), y ₁) ..., (n _n(t _i), y _m).This temporal aspect vector is as the training set of incremental learning in the SVM algorithm.

As shown in Figure 3, be the schematic flow sheet of the method for early warning embodiment tri-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance of the present invention, the specific implementation process that the voltage dip based on increment SVM algorithm is identified is automatically mainly described in the present embodiment, comprising:

Step 300 arranges the replacement policy of training sample in increment SVM algorithm, crosses threshold xi and eliminates the low sample of classification contribution degree.Its basic thought is, in n time window recently, if the number of times that sample does not become support vector surpasses ξ, it concentrated from training sample to deletion.

Wherein, the SVM algorithm is actual is by the Largrange function $L(w,b,a)=\frac{1}{2}{\left|\right|w\left|\right|}^2-\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{a}_j(y_i((w\&CenterDot;n_j\left(t_i\right)+b)-1),$ A quadratic programming problem just is converted into to dual problem as the formula (3).

$\begin{array}{c}\min \underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{a}_j-\frac{1}{2}\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}\underset{j\&prime;=1}{\overset{n}{\mathrm{\&Sigma;}}}{a}_j{a}_{j\&prime;}{y}_j{y}_{j\&prime;}(n_j\left(t_i\right)\&CenterDot;n_{j\&prime;}\left(t_i\right))\\ s.t.\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{y}_j{a}_j=0\\ a_j\&GreaterEqual;0,j=\mathrm{1,2},...,n\end{array}---\left(3\right)$

Step 301, for time window t _iinterior electric energy quality signal sequence T={ (n ₁(t _i), y ₁) ..., (n _n(t _i), y _m) as training set, trained, obtain new classification lineoid and classification function w _newx+b _new=0, its arthmetic statement is as follows:

Input: timeslice t _i-1the time classification lineoid w ₀x+b ₀=0 and support vector collection SV _i-1,

Timeslice t _iin sample T={ (n ₁(t _i), y ₁) ..., (n _n(t _i), y _m), training set sample T _i-1

Output: the classification lineoid w after batch incremental learning _newx+b _new=0, SV _new

Step 302, according to above-mentioned increment SVM algorithm, obtain a time window t _iafter characteristic hyperplane w _ix+b _i=0, by this characteristic hyperplane to time window t _i+1middle electric energy quality signal is classified, by time window t _i+1signal classification and mark after add again training set, re-start above-mentioned batch of incremental learning, the rest may be inferred.

As shown in Figure 4, be the schematic flow sheet of the method for early warning embodiment tetra-of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance of the present invention, the main specific implementation process of describing computing client sensitive equipment voltage dip tolerance level in the present embodiment comprises:

Step 400, for client's sensitive equipment, under the guidance of ITIC curve and SEMI curve, the method based on the voltage Sag Disturbance actual measurement, determine the voltage-tolerance curve of a few class sensitive equipments under the different loads level.

Step 401, determine according to the voltage dip tolerance curve fault zone, normal region and the uncertain region that sensitive equipment moves.

Concrete, according to ITIC curve and SEMI curve, can obtain the voltage Sag Disturbance sensitizing range of various kinds of equipment as shown in table 1.Wherein, a-quadrant represents that general sensitive equipment can affected normal region, the affected uncertain region of B Regional Representative semiconductor production enterprise, the computer equipment of C Regional Representative, PLC, the affected uncertain region of AC relay and semiconductor production enterprise, D Regional Representative semiconductor production enterprise and motor driver and the affected uncertain region of metal sodium vapor lamp, the E zone is the fault zone of all sensitive equipments.The sensitizing range that obtains thus one 6 * 7 is divided.

All kinds of sensitive equipment voltage Sag Disturbance of table 1 sensitizing range

Load for the unknown of client's sensitivity characteristic, sensitive features difference under different running environment, operating mode and different electrical energy power quality disturbance, be subject to the impact of the numerous uncertain factors of electric power system and load itself, its voltage dip sensitivity characteristic can't be known, need be assessed by the method for test of many times.

By the load voltage dip feature of supply terminals compare to determine the load-sensitive degree with equipment voltage tolerance level of monitoring.Embodiment five as shown in Figure 5, its concrete steps are as follows:

Step 500, preparation can be carried out the sensitive equipment of experimental test, simulation load operation scene.

Step 501, supply terminals and metering points of electric energy are set to power supply department and user's points of common connection.

Step 502, according to table 1, divide the voltage Sag Disturbance sensitizing range, and the medium power supply state by this grade on each zone is divided, and at points of common connection, simulates the voltage dip scene that these test samples are corresponding.On different voltage dip grades, the number of test sample and probability of malfunction gauging table are as shown in 2.

Table 2 voltage dip equipment susceptibility testing scheme

Step 503, in the actual measurement process, by slightly being tested to serious order, according to the zone of breaking down first, determine the uncertain region of sensitive load by the voltage Sag Disturbance degree.

As shown in Figure 6, schematic flow sheet for the method for early warning embodiment six of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance of the present invention, mainly describe the specific implementation process of the sensitive equipment failure rate assessment based on the maximum entropy theory in the present embodiment, comprising:

Step 600, for the voltage Sag Disturbance identified, if it falls temporarily residual voltage amplitude and duration and is positioned at the uncertain region of sensitive load, it is converted into to corresponding range of decrease value seriousness index (magnitude severity index temporarily according to formula (3) and formula (4), MSI) and the sample value of duration seriousness index (duration severity index, DSI).

${\mathrm{\&gamma;}}_{\mathrm{MSI}}=\left\{\begin{array}{cc}0& U>U_{\max }\\ (U_{\max }-U)\frac{100}{U_{\max }-U_{\min }}& U_{\min }\&le;U\&le;U_{\max }\\ 100& U<U_{\min }\end{array}\right.---\left(3\right)$

${\mathrm{\&gamma;}}_{\mathrm{DSI}}=\left\{\begin{array}{cc}0& T<T_{\min }\\ (T-T_{\min })\frac{100}{T_{\max }-T_{\min }}& T_{\min }\&le;T\&le;T_{\max }\\ 100& T>T_{\max }\end{array}\right.---\left(4\right)$

Step 601, determine sample value substitution maximum entropy model the probability density function of voltage dip seriousness index (MSI or DSI).

Determine the probability density function of voltage dip seriousness index (MSI or DSI) based on principle of maximum entropy, its advantage is directly according to sample data, to ask the probability density function of stochastic variable.When the stochastic variable continuous distribution, its maximum entropy model is:

$\begin{array}{c}\max H\left(x\right)=-\underset{R}{\&Integral;}f\left(x\right)\ln f\left(x\right)\mathrm{dx}\\ s.t.\\ \underset{R}{\&Integral;}f\left(x\right)\mathrm{dx}=1\\ \underset{R}{\&Integral;}\mathrm{xf}\left(x\right)\mathrm{dx}=E_1\\ \underset{R}{\&Integral;}{(x-E_1)}^{h}f\left(x\right)\mathrm{dx}=E_h,h=\mathrm{2,3},...,N\end{array}---\left(5\right)$

In formula (5), the stochastic variable that x is sensitive equipment voltage dip seriousness index MSI or DSI, the value border that R is variable x (depend on the residual voltage amplitude of sensitive load uncertain region and fall temporarily the border of duration), the entropy that H (x) is stochastic variable, the probability density function that f (x) is stochastic variable x, E ₁and E _hfor 1 rank unit point distance and h rank centre distance of voltage dip seriousness index, and huge exponent number is defined as to 5 to meet the requirement of Evaluation accuracy, i.e. N=5.

Step 602, introduce Lagrangian to maximum entropy model, and obtain the probability density function analytic expression by classical partial differentiation.When voltage dip occurs in uncertain region (i, j), the failure rate P (i, j) of equipment is:

$P(i,j)=\underset{x_{\mathrm{MSI}}}{\overset{{\mathrm{\&gamma;}}_{\stackrel{\&OverBar;}{i}}}{\&Integral;}}f\left(x_{\mathrm{MSI}}\right){\mathrm{dx}}_{\mathrm{MSI}}\underset{x_{\mathrm{DSI}}}{\overset{{\mathrm{\&gamma;}}_{\stackrel{\&OverBar;}{j}}}{\&Integral;}}f\left(x_{\mathrm{DSI}}\right){\mathrm{dx}}_{\mathrm{DSI}}$

Wherein, $f\left(x\right)=\exp ({\mathrm{\&lambda;}}_0+{\mathrm{\&lambda;}}_{1}x+\underset{h=2}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_h{(x-E_1)}^h),h=\mathrm{2,3,4,5}.$

with

respectively the residual voltage amplitude in the intermediate value of regional i and duration the intermediate value at regional j.

As shown in Figure 7, be the schematic flow sheet of the method for early warning embodiment seven of responsive client's electric energy Quality of experience under a kind of voltage Sag Disturbance of the present invention, the main specific implementation process of describing the analysis of client's electric energy Quality of experience and early warning in the present embodiment comprises:

Step 700, set up client's electric energy Quality of experience monitor supervision platform;

Step 701, according to the result of electrical energy power quality disturbance on-line real time monitoring, in conjunction with client's sensitivity levels, may threaten client's power quality to detect current electrical energy power quality disturbance.

Implement the present invention, can more accurately monitor the electrical energy power quality disturbance situation, and determine that according to client's load-sensitive degree whether this disturbance likely has influence on client's load, finds the potential danger of load operation.On this basis, propose the concept that client's electric energy is experienced, to load operation, exist the client of potential danger to carry out early warning.This method is conducive to electric power enterprise and responsive client reduces for electricity consumption risk, technological transformation, and significant to differentiation customization electric power.

One of ordinary skill in the art will appreciate that all or part of flow process realized in above-described embodiment method, to come the hardware that instruction is relevant to complete by computer program, described program can be stored in a computer read/write memory medium, this program, when carrying out, can comprise the flow process as the embodiment of above-mentioned each side method.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access Memory, RAM) etc.

Above content is in conjunction with concrete preferred implementation further description made for the present invention, can not assert that specific embodiment of the invention is confined to these explanations.For the general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, can also make some simple deduction or replace, all should be considered as belonging to protection scope of the present invention.

Claims (7)

1. the method for early warning of responsive client's electric energy Quality of experience under a voltage Sag Disturbance, is characterized in that, comprising:

Based on S conversion fast algorithm and increment svm classifier device, responsive client's voltage Sag Disturbance is identified automatically;

Result based on described voltage Sag Disturbance identification, determine the voltage-tolerance curve of equipment under the different loads level corresponding to the responsive client of several classes;

According to the Historical Monitoring data of voltage Sag Disturbance as sample, be converted into the sample value of amplitude seriousness index MSI and the duration seriousness index DSI of voltage dip, and determine the probability density function of MSI and DSI based on principle of maximum entropy, assessment sensitive equipment probability of malfunction, show that sensitive equipment corresponding to responsive client is at the probability broken down under the voltage dip rank.

2. as the method for early warning of responsive client's electric energy Quality of experience under the claim voltage Sag Disturbance, it is characterized in that, describedly based on S conversion fast algorithm and increment svm classifier device, responsive client's voltage Sag Disturbance identified automatically, comprising:

Determine that responsive client, as monitoring point, carries out Real-Time Monitoring to its electrical energy power quality disturbance, obtain sensitive equipment voltage signal, current signal as monitoring point;

For voltage signal and the current signal of monitoring point, carry out feature extraction based on S conversion fast algorithm, extract the proper vector of mode coefficient corresponding to standard deviation, maximum norm coefficient and the rated frequency of each frequency band mode coefficient as disturbing signal;

Described proper vector is inputed in increment svm classifier device and carries out Classification of Power Quality Disturbances, automatically identify voltage Sag Disturbance.

3. the method for early warning of responsive client's electric energy Quality of experience under voltage Sag Disturbance as claimed in claim 2, is characterized in that, described S conversion fast algorithm is as follows:

wherein, the broad sense window function that ω (τ-t, σ) is all frequency v in unit area.

4. the method for early warning of responsive client's electric energy Quality of experience under voltage Sag Disturbance as claimed in claim 3, is characterized in that, described proper vector is inputed in increment svm classifier device and carries out Classification of Power Quality Disturbances, automatically identifies voltage Sag Disturbance, comprising:

Proper vector is carried out to the time window division, use incremental learning mechanism, the proper vector arrived in a time window is carried out to study in batch;

In nearest n time window, if the number of times that proper vector sample does not become support vector is deleted it over ξ from training set.

5. the method for early warning of responsive client's electric energy Quality of experience under voltage Sag Disturbance as claimed in claim 4, is characterized in that, described sensitive equipment comprises FPGA (Field Programmable Gate Array) control PLC, variable speed drivP ASD, computer PC, A.C. contactor ACC.

6. the method for early warning of responsive client's electric energy Quality of experience under voltage Sag Disturbance as claimed in claim 5, is characterized in that, described maximum entropy model is:

$\begin{array}{c}\max H\left(x\right)=-\underset{R}{\&Integral;}f\left(x\right)\ln f\left(x\right)\mathrm{dx}\\ s.t.\\ \underset{R}{\&Integral;}f\left(x\right)\mathrm{dx}=1\\ \underset{R}{\&Integral;}\mathrm{xf}\left(x\right)\mathrm{dx}=E_1\\ \underset{R}{\&Integral;}{(x-E_1)}^{h}f\left(x\right)\mathrm{dx}=E_h,h=\mathrm{2,3},...,N\end{array}$

Wherein, the stochastic variable that x is sensitive equipment voltage dip seriousness index MSI or DSI, the value border that R is variable x, the entropy that H (x) is stochastic variable, the probability density function that f (x) is stochastic variable x, E ₁and E _h1 rank unit point distance and the h rank centre distance for voltage dip seriousness index.

7. the method for early warning of responsive client's electric energy Quality of experience under voltage Sag Disturbance as claimed in claim 6, it is characterized in that, the described probability density function of determining MSI and DSI based on principle of maximum entropy, assessment sensitive equipment probability of malfunction, show that sensitive equipment corresponding to responsive client, at the probability broken down, comprising under the voltage dip rank:

Introduce Lagrangian in described maximum entropy model, and obtain the probability density function analytic expression by classical partial differentiation;

When voltage dip occurs in uncertain region (i, j), the failure rate P (i, j) of equipment is:

$P(i,j)=\underset{x_{\mathrm{MSI}}}{\overset{{\mathrm{\&gamma;}}_{\stackrel{\&OverBar;}{i}}}{\&Integral;}}f\left(x_{\mathrm{MSI}}\right){\mathrm{dx}}_{\mathrm{MSI}}\underset{x_{\mathrm{DSI}}}{\overset{{\mathrm{\&gamma;}}_{\stackrel{\&OverBar;}{j}}}{\&Integral;}}f\left(x_{\mathrm{DSI}}\right){\mathrm{dx}}_{\mathrm{DSI}}$

Wherein, $f\left(x\right)=\exp ({\mathrm{\&lambda;}}_0+{\mathrm{\&lambda;}}_{1}x+\underset{h=2}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_h{(x-E_1)}^h),h=\mathrm{2,3,4,5}.$

with

respectively the residual voltage amplitude in the intermediate value of regional i and duration the intermediate value at regional j.

Images