CN104239971B - Spatial load forecasting error evaluation method based on multi-scale spatial resolution - Google Patents

Abstract

The invention discloses a spatial load forecasting error evaluation method based on the multi-scale spatial resolution. The method is characterized by comprising the steps that a power geographic information system is set, two kinds of load cells are generated under power load spatial resolutions of different scales in the power geographic information system, and the load cells are called cells for short; a load density index method is applied to spatial load forecasting of an area to be measured, and a load forecasting error is calculated according to a predicted value and a real value; based on the multi-scale spatial resolution analysis technology, the affiliation membership between the two kinds of cells is judged by defining variables of the cells, and the spatial error influence value is calculated according to the distance of a certain cell and the other cells and the amplitude error of the load predicated value; the predicated error influence value is compared with a difference value between the capacity of corresponding power equipment or facility and the actual load, an error evaluation index is set, and the spatial load forecasting error is evaluated.

Description

A kind of Spatial Load Forecasting error assessment method based on multiscale space resolution

Technical field

The present invention relates to the Spatial Load Forecasting field in urban distribution network planning, it is a kind of based on multiscale space resolution The Spatial Load Forecasting error assessment method of rate.

Background technology

Spatial Load Forecasting (Spatial Load Forecasting, SLF) as the basis of urban distribution network planning, no Only it needs to be determined that going out the future load size of each cell in urban distribution network planning region in addition it is also necessary to predict dividing of this load Cloth.Only it is improved the accuracy of Spatial Load Forecasting, could more accurately instruct building of transformer station, feeder line, switchgear etc. And if use, make the development of electrical network and run more economical.

Spatial Load Forecasting error refers to predict inclined between the numerical values recited of load and actual load and its spatial distribution Difference, has obvious duality.

If not considering to predict load place locus, and only calculate the relative error of prediction load amplitude, definitely miss Difference, absolute error meansigma methodss (Average Absolute Value, AAV), absolute error root-mean-square valve (Root Mean Square, RMS) to carry out evaluation and foreca effect be then not sufficient.Because traditional forecast error analysis generally believes that load is pre- The amplitude error surveying result is less, and the adverse effect that Power System Planning is caused is also less；But actual SLF result is to power train The impact of system planning is also relevant with the space error predicting the outcome, that is, the size of the error that predicts the outcome and its spatial distribution are to electric power Equipment, the capacity of facility, scale, placement position, the planning of time of putting into operation suffer from large effect.

Content of the invention

It is an object of the present invention to provide a kind of methodological science, rationally, simple, practical one kind is divided based on multiscale space The Spatial Load Forecasting error assessment method of resolution.

Realize the object of the invention be employed technical scheme comprise that, a kind of space load based on multiscale space resolution is pre- Survey error assessment method, it is characterized in that, it comprises the following steps：

1) set up GIS for electric power

In GIS platform, the satellite photo in region to be measured is carried out longitude and latitude registration and forms base figure, with base Figure is position and its supply district figure layer, 10kV feeder line and its supply district figure layer that background sets up secondary substation respectively, supplies Electric office and the supply district figure layer of power supply branch office；

2) electric load spatial resolution

(1) I space-like resolution

The gross area of power supply area is S, and the area of each power supply area is respectively S₁, S₂..., S_n, n is the total of power supply area Number；Make k₁=S₁/ S, k₂=S₂/ S ..., k_n=S_n/ S, has r_kn={ (1/ (k₁S), 1/ (k₂S) ..., 1/ (k_nS)) }, then claim r_knFor (k₁, k₂..., k_n) metric space resolution；Now r_knIt is an ordered series of numbers, be used for portraying one group of size, irregular such as not Cell, and it is defined as I space-like resolution；

Wherein：S is the gross area of power supply area, S₁, S₂..., S_nIt is respectively the area of each power supply area, n is power supply area Total number, k₁, k₂..., k_nIt is respectively the area of each power supply area and the ratio of the power supply area gross area, r_knFor (k₁, k₂..., k_n) metric space resolution；

But, for the identical power supply area of different sequences, corresponding I space-like resolution is diverse, this Occur as soon as the problem inconsistent to the spatial resolution statement in same region to be measured；Build during statement SREL for this The vertical uniform rules with regard to power supply area sequence：First, obtain the approximate geometric center (x, y) of each power supply area respectively, with a left side The approximate geometric center of inferior horn power supply area is initial point, sets up rectangular coordinate system；Then, by (x+y), order from small to large will Sort in power supply area；If certain several power supply area (x+y) is equal, by x order from small to large, sorted in power supply area；? Afterwards, ask for I space-like resolution by the sequence of gained power supply area；This method determining spatial resolution solves to same The spatial resolution in individual region to be measured states inconsistent problem, abbreviation WT-XP method；

Wherein：(x, y) is latitude and longitude coordinates in electric power GIS for each power supply area geometric center；

(2) II space-like resolution

The gross area of power supply area is S, and each power supply area such as is at the regular shape of size, square, rectangle or six sides Shape, area is S₀, make k=S₀/ S, has r_k=1/ (kS),

$C={\left[\begin{array}{cccc}{r}_k& r_k& \·\·\·& r_k\\ r_k& r_k& \·\·\·& r_k\\ \·\·\·& \·\·\·& \·\·\·& \·\·\·\\ r_k& r_k& \·\·\·& r_k\end{array}\right]}_{m\×n}$

Now r_kIt is a numerical value, C is used for portraying the size rule cell such as a group, and is defined as II space-like resolution；

Wherein：S is the gross area of power supply area, S₀For the area of the power supply area of the size regular shape such as each, k is each etc. The power supply area area of size regular shape and the ratio of the power supply area gross area, m is the line number of matrix, and n is matrix column number；

3) generate cellular

Divide region to be measured to generate cellular with the supply district of power equipment, facility, power supply unit respectively, and by these Not etc. the cellular of size is not referred to as I class cellular；

According to etc. size regular grid divide region to be measured to generate cellular, and by these cellulars be referred to as II class cellular；

4) ask for space error

With district load density index method, Spatial Load Forecasting is carried out to the II class cellular in region to be measured, and pre- according to its load Measured value and load actual value ask for space error,

E=L_y-L_s(1)

Wherein：E is the error of " region to be measured " cellular load prediction results, and L is cellular load, L_sFor region to be measured cellular Load actual value, the actual meaning of behalf, L_yFor the predicted load of region to be measured cellular, the meaning of y representative prediction；

5) set up flag variable

In order to judge that certain II class cellular (i, j) belongs to the supply district of which secondary substation, set up flag variable F (i, j), is shown in formula (2),

Wherein：(i, j) is the ranks that II class cellular is located, and i is the line number that II class cellular is located, and j is that II class cellular is located Columns,

F (i, j) is to judge that II class cellular (i, j) belongs to the supply district of which secondary substation and the flag set up Variable；

It is on multiple secondary substation supply districts boundaries in view of some II class cellulars, in order to judge these II classes units Born of the same parents belong to the supply district of which secondary substation, set up corresponding judgment rule：Complete to arrange in secondary substation supply district On the premise of sequence, II class cellular (i, j) is intersected with n secondary substation supply district, and the area of overlapping part is respectively s₁, s₂..., s_n, then

(1) in max { s₁, s₂..., s_n}=s_iWhen, this II class cellular belongs to the supply district of i-th secondary substation；

(2) in max { s₁, s₂..., s_n}=s_i=...=s_jWhen, this II class cellular belongs to secondary substation supply district row Sequence preceding secondary substation supply district；

Wherein：S is the area of II class cellular and secondary substation supply district overlapping part, s₁, s₂..., s_nIt is respectively II Class cellular (i, j) and the 1st, 2 ..., the area of n secondary substation supply district overlapping part, n is the total of secondary substation Number,

s_i、s_jIt is respectively II class cellular (i, j) and the i-th, area of the supply district overlapping part of j secondary substation；

6) build membership probability function

Judge each II class cellular is powered by which secondary substation respectively according to flag variable, statistics any two II class unit The probability whether born of the same parents are powered by same secondary substation, utilization index function pair statistical result is fitted, and then must take office The membership probability function that two II class cellulars of meaning are powered by same secondary substation, is shown in formula (3)；

$p=f\left(d\right)=a\·e^{{-(b\·d)}^2}---\left(3\right)$

Wherein：P is the probability powered by same secondary substation of cellular (m, n) and cellular (i, j), and f closes for Probability p In the function apart from d, a, b are relevant parameter, and d is the distance between cellular (m, n) and cellular (i, j), and m, i are respectively cellular The line number that (m, n), cellular (i, j) are located, the columns that n, j are respectively cellular (m, n), cellular (i, j) is located；

7) ask for error influence value

According to the space error value of membership probability function f (d) and each II class cellular can obtain II class cellular (m, N) error influence value E (i, j) to any II class cellular (i, j) for the space error, the error influence value E of any cellular (i, j) (i, j) refers to the impact to cellular (i, j) forecast error of the error of all II class cellulars；

$E(i,j)=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}[f\left(\sqrt{{(m-i)}^2+{(n-j)}^2}\right)\·e(m,n)]/T---\left(4\right)$

$T=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}f\left(\sqrt{m^2+n^2}\right)---\left(5\right)$

Wherein：E is Spatial Load Forecasting error influence value, and E (i, j) is the error influence value of any II class cellular (i, j), E (m, n) is the space error of II class cellular (m, n), and T is probability function sum, and M is total line number of II class cellular, and N is II class unit Total columns of born of the same parents；

8) Plan mode based on error influence value

Under the resolution of II space-like, secondary substation transformator specified active for P_N, actual load maximum is P_max, the error influence value of i-th I class cellular is E_i, specified active difference DELTA P with actual load maximum is

Δ P=P_N-P_max(6)

(1) if E_i＜ Δ P, then illustrate that the error of Spatial Load Forecasting result has exceeded the anti-of secondary substation and has been subject to ability, The Spatial Load Forecasting resultant error of secondary substation will not produce impact to original planning；

(2) if E_i＞ Δ P, then illustrate that the error of Spatial Load Forecasting result has exceeded the anti-of secondary substation and has been subject to ability, The Spatial Load Forecasting resultant error of this secondary substation will produce considerable influence to original planning, needs to reformulate planning Scheme；

Wherein：P_NSpecified active, P for secondary substation transformator_maxFor actual load maximum, Δ P is secondary power transformation The specified active difference with actual load maximum of transformator of standing, E_iFor the error influence value of i-th I class cellular, i=1, 2 ..., n, n are the sum of secondary substation.

A kind of Spatial Load Forecasting error assessment method based on multiscale space resolution of the present invention, by setting up WT-XP rule it is achieved that to spatial resolution statement concordance；By defining flag variable, building membership probability function, With multiscale space resolution technique it is achieved that the quantitation that the error of a cellular produces impact to other cellular loads is divided Analysis, has methodological science, rationally, the advantages of simple, practical.

Brief description

Fig. 1 is I class electric load spatial resolution schematic diagram；

Fig. 2 is each I class cellular and its Geometric center coordinates schematic diagram；

Fig. 3 is the II class electric load spatial resolution schematic diagram under large scale；

Fig. 4 is the II class electric load spatial resolution schematic diagram under little yardstick；

Fig. 5 is membership probability function figure；

Fig. 6 is the spatial distribution schematic diagram of error influence value in 2009.

Specific embodiment

Below with drawings and Examples, the invention will be further described.

Reference picture 1- Fig. 6, shows Jilin Chuanying District in Fig. 1, according to secondary substation supply district, with WT-XP Rule statement I space-like resolution：The process realizing WT-XP rule is shown in Fig. 2；II class electric load is shown in Fig. 3 The yardstick of spatial resolution is larger, and resolution is relatively low；Show in Fig. 4 that the yardstick of II class electric load spatial resolution is less, Resolution is higher；Fig. 5 gives distance between two II class cellulars and they are under the jurisdiction of between the probability of an I cellular simultaneously Relation；Fig. 6 gives under the resolution of specified II space-like, SLF resultant error influence values in 2009 of Jilin Chuanying District scope Distribution situation.

A kind of Spatial Load Forecasting error assessment method based on multiscale space resolution of the present invention, walks including following Suddenly：

1) set up GIS for electric power, its English name is Geographic Information System, GIS

In GIS software platform, the satellite photo of Jilin is carried out longitude and latitude registration and forms base figure, with Base figure is position and its supply district figure layer, 10kV feeder line and its supply district figure layer that background sets up secondary substation respectively, Power supply administration and the supply district figure layer of power supply branch office, the powering area of each secondary substation in Jilin Chuanying District, are shown in Table 1；

The powering area of table 1 secondary substation

2) electric load spatial resolution, its English name is Spatial Resolution of Electric Load, SREL

(1) I space-like resolution

First, obtain the approximate geometric center (x, y) of each power supply area respectively, with the approximate geometry of lower left corner power supply area Center is initial point, sets up rectangular coordinate system；Then, by (x+y), sorted in power supply area by order from small to large；If certain several confession Electric cell (x+y) is equal, then by x, sorted in power supply area by order from small to large；It is shown in Table 2；Finally, by gained power supply area I class SREL, r are asked in sequence_k7={ 1/7.098,1/4.372,1/4.196,1/2.128,1/2.151,1/2.014,1/ 1.385}；

The latitude and longitude coordinates of secondary substation and sequence in table 2 electric power GIS

(2) II space-like resolution

The square net of length of side 0.3km according to the sizes such as 792, asks for II class SREL, that is,

$C={\left[\begin{array}{cccc}1/0.09& 1/0.09& \·\·\·& 1/0.09\\ 1/0.09& 1/0.09& \·\·\·& 1/0.09\\ \·\·\·& \·\·\·& \·\·\·& \·\·\·\\ 1/0.09& 1/0.09& \·\·\·& 1/0.09\end{array}\right]}_{33\×24}$

3) generate cellular

Divide region to be measured to generate I class cellular by the supply district of secondary substation；

Square net according to sizes such as length of sides 0.3km divides region to be measured to generate II class cellular；

4) ask for space error

With district load density index method, Spatial Load Forecasting is carried out to " region to be measured ", and according to its predicted load and bear Lotus actual value asks for space error；

E=L_y-L_s(7)

Wherein：E is the error of " region to be measured " cellular load prediction results, and L is cellular load, L_sFor " region to be measured " unit The load actual value of born of the same parents, the actual meaning of behalf, L_yFor the predicted load of " region to be measured " cellular, the meaning of y representative prediction Think；Associated loadings data is shown in Table 3～5；

The load actual value (MW) of table 3 Chuanying District II class cellular in 2009

The predicted load (MW) of table 4 Chuanying District II class cellular in 2009

The space error (MW) of table 5 Chuanying District II class cellular in 2009

5) set up flag variable

In order to judge that certain II class cellular (i, j) belongs to the supply district of which secondary substation, set up flag variable F (i,j)：

Wherein：(i, j) is the ranks that II class cellular is located, and i is the line number that II class cellular is located, and j is that II class cellular is located Columns,

F (i, j) is to judge that II class cellular (i, j) belongs to the supply district of which secondary substation and the flag set up Variable；

It is on multiple secondary substation supply districts boundaries in view of some II class cellulars, in order to judge these II classes units Born of the same parents belong to the supply district of which secondary substation, set up corresponding judgment rule：Complete to arrange in secondary substation supply district On the premise of sequence, arbitrary II class cellular is intersected with n secondary substation supply district, and the area of overlapping part is respectively S₁, S₂..., S_n, then

(1) in max { S₁, S₂..., S_n}=S_iWhen, this II class cellular belongs to the supply district of i-th secondary substation；

(2) in max { S₁, S₂..., S_n}=S_i=...=S_jWhen, this II class cellular belongs to secondary substation supply district row Sequence preceding secondary substation supply district；

Wherein：S is the area of II class cellular and secondary substation supply district overlapping part,

s₁, s₂..., s_nIt is respectively II class cellular (i, j) and 1,2 ..., n secondary substation supply district overlapping part Area, n is the sum of secondary substation；

6) build membership probability function

Judge each II class cellular is powered by which secondary substation respectively according to flag variable, statistics any two II class unit The probability whether born of the same parents are powered by same secondary substation, utilization index function pair statistical result is fitted, and then must take office The membership probability function that two II class cellulars of meaning are powered by same secondary substation, is shown in formula (9)；

$p=f\left(d\right)=a\·e^{{-(b\·d)}^2}---\left(9\right)$

Wherein：P is the probability powered by same secondary substation of cellular (m, n) and cellular (i, j), and f closes for Probability p In the function apart from d, a, b are relevant parameter, and d is cellular (m, n) and the distance between (i, j), m, i be respectively cellular (m, n), The line number that cellular (i, j) is located, the columns that n, j are respectively cellular (m, n), cellular (i, j) is located；

The relevant parameter of Chuanying District secondary substation probability function in 2009 is shown in Table 6；

The relevant parameter of table 6 probability function

7) ask for error influence value

According to the space error value of membership probability function f (d) and each II class cellular can obtain II class cellular (m, N) II class error influence value E (i, j) to any II class cellular (i, j) for the space error, the error impact of any cellular (i, j) Value E (i, j) refers to the impact to cellular (i, j) forecast error of the error of all II class cellulars；

$E(i,j)=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}[f\left(\sqrt{{(m-i)}^2+{(n-j)}^2}\right)\·e(m,n)]/T---\left(10\right)$

$T=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}f\left(\sqrt{m^2+n^2}\right)---\left(11\right)$

Wherein：E is Spatial Load Forecasting error influence value, and E (i, j) is the error influence value of any II class cellular (i, j), E (m, n) is the space error of II class cellular (m, n), and T is probability function sum, and M is total line number of II class cellular, and N is II class unit Total columns of born of the same parents；

Said method is equally applicable to asking for of 10kV feeder line rank SLF resultant error influence value；

Calculate the error influence value of Chuanying District II class cellular in 2009 using formula (7)～(11), the results are shown in Table 7；

The error influence value (MW) of table 7 Chuanying District II class cellular in 2009

8) Plan mode based on error influence value

Under the resolution of II space-like it is assumed that secondary substation transformator specified active for P_N, actual load maximum For P_max, the error influence value of i-th I class cellular is E_i, then specified active difference DELTA P with actual load maximum be

Δ P=P_N-P_max(12)

(1) if E_i＜ Δ P, then illustrate that the error of Spatial Load Forecasting result has exceeded the anti-of secondary substation and has been subject to ability, The Spatial Load Forecasting resultant error of secondary substation will not produce impact to original planning；

(2) if E_i＞ Δ P, then illustrate that the error of Spatial Load Forecasting result has exceeded the anti-of secondary substation and has been subject to ability, The Spatial Load Forecasting resultant error of this secondary substation will produce considerable influence to original planning, needs to reformulate planning Scheme；

Wherein：P_NSpecified active, P for secondary substation transformator_maxFor actual load maximum, Δ P is secondary power transformation The specified active difference with actual load maximum of transformator of standing, E_iFor the error influence value of i-th I class cellular, i=1, 2 ..., n, n are the sum of secondary substation；

The index of correlation of Chuanying District secondary substation in 2009 is shown in Table 8；

The index of correlation of table 8 Chuanying District secondary substation in 2009

By the interpretation of result to table 8, it can be found that the II class error influence value E of arctic change in 2009_iMore than corresponding Specified active difference DELTA P with load actual value, has illustrated the Spatial Load Forecasting resultant error of this secondary substation above-mentioned Anti- beyond corresponding secondary substation is subject to ability, and original planning will be unable to meet the demand of future load.

Only certain embodiments of the present invention has made detailed explanation to present disclosure, but does not limit to the present embodiment, Any obvious change that those skilled in the art are done according to the enlightenment of the present invention, broadly falls into rights protection of the present invention Scope.

Claims (1)

1. a kind of Spatial Load Forecasting error assessment method based on multiscale space resolution, is characterized in that, it includes following Step：

1) set up GIS for electric power

In GIS platform, the satellite photo in region to be measured is carried out longitude and latitude registration and forms base figure, with base figure be Background sets up position and its supply district figure layer, 10kV feeder line and its supply district figure layer of secondary substation, power supply administration respectively And the supply district figure layer of power supply branch office；

2) electric load spatial resolution

(1) I space-like resolution

The gross area of power supply area is S, and the area of each power supply area is respectively S₁, S₂..., S_n, n is the total number of power supply area； Make k₁=S₁/ S, k₂=S₂/ S ..., k_n=S_n/ S, has r_kn={ (1/ (k₁S), 1/ (k₂S) ..., 1/ (k_nS)) }, then claim r_knFor (k₁, k₂..., k_n) metric space resolution；Now r_knIt is an ordered series of numbers, be used for portraying one group of size, irregular cell such as not, And it is defined as I space-like resolution；

Wherein：S is the gross area of power supply area, S₁, S₂..., S_nIt is respectively the area of each power supply area, n is the total of power supply area Number, k₁, k₂..., k_nIt is respectively the area of each power supply area and the ratio of the power supply area gross area, r_knFor (k₁, k₂..., k_n) Metric space resolution；

But, for the identical power supply area of different sequences, corresponding I space-like resolution is diverse, and this just goes out Show the problem inconsistent to the spatial resolution statement in same region to be measured；For this during statement spatial resolution Set up the uniform rules with regard to power supply area sequence：First, obtain the approximate geometric center (x, y) of each power supply area respectively, with The approximate geometric center of lower left corner power supply area is initial point, sets up rectangular coordinate system；Then, the order from small to large by (x+y) Sorted in power supply area；If certain several power supply area (x+y) is equal, by x order from small to large, sorted in power supply area；? Afterwards, ask for I space-like resolution by the sequence of gained power supply area；This method determining spatial resolution solves to same The spatial resolution in individual region to be measured states inconsistent problem, abbreviation WT-XP method；

Wherein：(x, y) is latitude and longitude coordinates in electric power GIS for each power supply area geometric center；

(2) II space-like resolution

The gross area of power supply area is S, and each power supply area such as is at the regular shape of size, square, rectangle or hexagon, face Amass and be S₀, make k=S₀/ S, has r_k=1/ (kS),

$C={\left[\begin{array}{cccc}{r}_k& r_k& ...& r_k\\ r_k& r_k& ...& r_k\\ ...& ...& ...& ...\\ r_k& r_k& ...& r_k\end{array}\right]}_{m\×n}$

Now r_kIt is a numerical value, C is used for portraying the size rule cell such as a group, and is defined as II space-like resolution；

Wherein：S is the gross area of power supply area, S₀For the area of the power supply area of the size regular shape such as each, k is the size such as each The power supply area area of regular shape and the ratio of the power supply area gross area, m is the line number of matrix, and n is matrix column number；

3) generate cellular

Divide region to be measured to generate cellular with the supply district of power equipment, facility, power supply unit respectively, and by these The cellular of size is referred to as I class cellular；

According to etc. size regular grid divide region to be measured to generate cellular, and by these cellulars be referred to as II class cellular；

4) ask for space error

With district load density index method, Spatial Load Forecasting is carried out to the II class cellular in region to be measured, and according to its predicted load Ask for space error with load actual value,

E=L_y-L_s(1)

Wherein：E is the error of " region to be measured " cellular load prediction results, and L is cellular load, L_sNegative for region to be measured cellular Lotus actual value, the actual meaning of behalf, L_yFor the predicted load of region to be measured cellular, the meaning of y representative prediction；

5) set up flag variable

In order to judge that certain II class cellular (i, j) belongs to the supply district of which secondary substation, set up flag variable F (i, j), See formula (2),

Wherein：(i, j) is the ranks that II class cellular is located, and i is the line number that II class cellular is located, and j is the row that II class cellular is located Number,

F (i, j) is that the flag set up to judge II class cellular (i, j) to belong to the supply district of which secondary substation becomes Amount；

It is on multiple secondary substation supply districts boundaries in view of some II class cellulars, in order to judge these II class cellulars genus In the supply district of which secondary substation, set up corresponding judgment rule：Complete sequence in secondary substation supply district Under the premise of, II class cellular (i, j) is intersected with n secondary substation supply district, and the area of overlapping part is respectively s₁, s₂..., s_n, then

(1) in max { s₁, s₂..., s_n}=s_iWhen, this II class cellular belongs to the supply district of i-th secondary substation；

(2) in max { s₁, s₂..., s_n}=s_i=...=s_jWhen, this II class cellular belongs to the sequence of secondary substation supply district and exists Front secondary substation supply district；

Wherein：S is the area of II class cellular and secondary substation supply district overlapping part, s₁, s₂..., s_nIt is respectively II class unit Born of the same parents (i, j) and the 1st, 2 ..., the area of n secondary substation supply district overlapping part, n is the sum of secondary substation,

s_i、s_jIt is respectively II class cellular (i, j) and the i-th, area of the supply district overlapping part of j secondary substation；

6) build membership probability function

Judge each II class cellular is powered by which secondary substation respectively according to flag variable, statistics any two II class cellular is The no probability powered by same secondary substation, utilization index function pair statistical result is fitted, and then obtains any two The membership probability function that individual II class cellular is powered by same secondary substation, is shown in formula (3)；

$p=f\left(d\right)=a\·e^{-{(b\·d)}^2}---\left(3\right)$

Wherein：P is the probability powered by same secondary substation of cellular (m, n) and cellular (i, j), f be Probability p with regard to away from The distance between from the function of d, a, b are relevant parameter, and d is cellular (m, n) with cellular (i, j), m, i be respectively cellular (m, n), The line number that cellular (i, j) is located, the columns that n, j are respectively cellular (m, n), cellular (i, j) is located；

7) ask for error influence value

Can obtain II's class cellular (m, n) according to the space error value of membership probability function f (d) and each II class cellular Error influence value E (i, j) to any II class cellular (i, j) for the space error, error influence value E (i, j) of any cellular (i, j) Refer to the impact to cellular (i, j) forecast error of the error of all II class cellulars；

$E(i,j)=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}\[f\left(\sqrt{{(m-i)}^2+{(n-j)}^2}\right)\·e(m,n)\]/T---\left(4\right)$

$T=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}f\left(\sqrt{m^2+n^2}\right)---\left(5\right)$

Wherein：E is Spatial Load Forecasting error influence value, and E (i, j) is the error influence value of any II class cellular (i, j), e (m, N) space error for II class cellular (m, n), T is probability function sum, and M is total line number of II class cellular, and N is II class cellular Total columns；

8) Plan mode based on error influence value

Under the resolution of II space-like, secondary substation transformator specified active for P_N, actual load maximum is P_max, i-th The error influence value of individual I class cellular is E_i, specified active difference DELTA P with actual load maximum is

Δ P=P_N-P_max(6)

(1) if E_i<Δ P, then illustrate that the error of Spatial Load Forecasting result has exceeded the anti-of secondary substation and has been subject to ability, secondary change The Spatial Load Forecasting resultant error in power station will not produce impact to original planning；

(2) if E_i>Δ P, then illustrate that the error of Spatial Load Forecasting result has exceeded the anti-of secondary substation and has been subject to ability, this is secondary The Spatial Load Forecasting resultant error of transformer station will produce considerable influence to original planning, needs to reformulate programme；

Wherein：P_NSpecified active, P for secondary substation transformator_maxFor actual load maximum, Δ P is secondary substation change The specified active difference with actual load maximum of depressor, E_iFor the error influence value of i-th I class cellular, i=1,2 ..., n, n Sum for secondary substation.