CN104021501A - PRA (Probabilistic Risk Assessment) based partitioning method for coordinating power-system planning with operation - Google Patents

Abstract

A PRA-based partitioning method for coordinating power-system planning with operation comprises the steps that S1) basic data of the grid is read in, and an accident set is selected; S2) power flow is calculated based on BPA software, whether problems as overload or overvoltage of branches occur is determined, and the static safety analysis is carried out; and at the same time, models of component failure rate and failure event probability are established, and the component failure rate and failure event probability are calculated; S3) based on a sensitivity matrix, load loss caused by a fault event is calculated via a safety correction strategy; S4) according to the fault event probability and the load loss caused by the fault event, the probabilistic rick index of load loss is calculated, and quantitative risk assessment is carried out; S5) according to the accident grade determining criteria and system bearing capability, load loss values allowed by different accident grades of the system are determined, and different accident grades for which planning and operation are respectively responsible are clearly determined; and S6) the responsibility boundary value between planning and operation obtained by S5 is compared with the probabilistic risk index of load loss of the fault event to determine the responsibility area to which the fault event belongs.

Description

A kind of partition method of coordinating Power System Planning and operation based on Risk Theory

Technical field

The present invention relates to a kind of utilization and lose the partition method that Load Probability risk indicator is coordinated Power System Planning and operation, is a kind of partition method of coordinating Power System Planning and operation based on Risk Theory specifically.

Background technology

Power System Planning is constructing system with operation, maintain two importances of system normal operation.In planning, to be considered as on the one hand the probability of each factor (as load, trend etc.) that precondition used, when the necessity of installing power equipment is discussed, meet N-1 criterion on the other hand, or even N-2, N-3 criterion; And the consideration of system operation needs is the problems such as system limits situation, equipment ability to bear, the quality of power supply.Therefore, the emphasis of systems organization and operation is different, and planning is not identical with the angle that operation department pays close attention to yet.

Both at home and abroad to how coordinating Electric Power Network Planning and still planless discussion of operation, planning is probabilistic at present, and operation need to be considered ultimate value, thinks that operating protection, stability control equipment are supplementing facility planning.In WSCC (Western Systems Coordinating Council), applied probability method is planned as a whole systems organization, but it only limits to planning field, does not relate to operation of power networks.In addition, except Britain will plan with reliability of operation, require to separate, and require planning to observe outside more high-grade reliability standard, in the reliability standard of most countries, do not distinguish planning and operation, but be as the criterion with the requirement in planning.Though meet reliability requirement in the time of so can guaranteeing to move, can increase construction cost, can not realize maximization of economic benefit, easily cause unnecessary waste.Between planning and operation, how division limits, the relation of putting in order are the long-standing practical problemss of system, but existing document or pay close attention to power network planning scheme and formulate, or pay close attention to scheduling controlling, and the research of relation is but comparatively limited to planning and between moving.

Risk assessment (Probabilistic Risk Assessment, PRA) is now applied in a plurality of fields of electric system, comprises Electricity market analysis, Design of DSS, EMS system, cascading failure analysis etc. ^[5-9].Risk assessment is considered the various uncertain factors of electric system for science provides effective means, and also the relation for further coordinated planning and operation provides possibility.

Summary of the invention

Technical matters to be solved by this invention, just be to provide a kind of partition method of coordinating Power System Planning and operation based on Risk Theory, risk indicator is the important indicator of evaluating system reliability, has incorporated probability understanding, more meets the practical operation situation of system; According to Risk Theory, judge the influence degree of each accident to system, proposed to come dividing system planning and the partition method moving based on losing Load Probability risk indicator, can solve planning and the divisions of responsibility can problem of moving comparatively simple and effectively, the relation of coordinated planning and operation, balanced system reliability requirement and construction cost, stable, the reliable and safety of assurance system.

Solve the problems of the technologies described above, the technical solution used in the present invention is:

A partition method of coordinating Power System Planning and operation based on Risk Theory, is characterized in that comprising the following steps:

S1 reads in rack basic data, comprises that the operational factor of bus, generator, transformer, overhead transmission line and load distribute; And according to electric power safety, stablize the accident pattern that requires check in guide rule and electric power safety accident investigation regulation, choose accident collection;

S2 carries out trend calculating based on BPA software, has judged whether the problems such as branch road overload or voltage cross the border, and carries out static security analysis; Meanwhile, set up element failure rate and event of failure probability model, computing element failure rate and event of failure probability;

S3, based on sensitivity matrix, utilizes safe Correction Strategies to calculate the load loss that event of failure causes;

The load loss that S4 causes according to event of failure probability and event of failure, calculates and loses Load Probability risk indicator, carries out quantifying risk assessment;

S5 is according to incident classification criterion and system ability to bear, and the load loss value that each incident classification of judgement system allows, clearly plans and move the different incident classifications of being responsible for separately; In addition, in order to guarantee the safety and stability of electric system, event of failure is done to enough conservative region and divide, during the cut off value of zoning, adopt minimum element failure rate as reference element failure rate, in conjunction with the responsibility cut off value of calculating between planning and operation;

According in 599 commands for minor accident, ordinary accident with define compared with the load loss of major break down, minor accident and ordinary accident and ordinary accident and be respectively compared with the mistake Load Probability risk indicator cut off value of major break down: system total load is multiplied by 4% and is multiplied by reference element failure rate and system total load again and is multiplied by 7% and is multiplied by reference element failure rate again; Due to the minor accident impact weak on system, minor accident is included into operation and is responsible for region, other accidents are included into planning and are responsible for region, so region and mistake Load Probability risk cut off value between the responsible region of planning are responsible in operation, can be decided to be system total load and be multiplied by 4% and be multiplied by reference element failure rate again;

S6 compares the planning of step S5 gained and the mistake Load Probability risk indicator of operation responsibility cut off value and event of failure, determines the area of responsibility that event of failure is affiliated.

In described step S2, the computing method of element failure rate u are as follows:

$u=\frac{F_{\mathrm{outage}}\×T_{\mathrm{repair}}}{8760}=\frac{\mathrm{\λ}}{\mathrm{\λ}+\mathrm{\μ}}=1-\frac{D}{100}---\left(1\right);$

Wherein, F _outagethe average stoppage in transit frequency of indication equipment, unit is Failure count/year, T _repairmean time to repair after indication equipment generation forced outage, unit is h, and λ is crash rate, and unit is Failure count/year, and μ is repair rate, and unit is for repairing number of times/year, and D represents availability coefficient;

The computing method of availability coefficient D are as follows:

$D=\frac{T_1}{T_1+T_2+T_3}---\left(2\right)$

Wherein, T ₁represent pot life (h), T ₂represent the planned outage time (h), T ₃represent unplanned idle time (h); The planned outage time refers to the time of the maintenance of stopping transport according to original plan or stop transport, and unplanned idle time refers to not in inside the plan fault idle time;

In the time cannot obtaining the probability of malfunction of circuit based on statistics, by the proportionate relationship of each line impedance and reference impedance, estimate the length of each circuit, thus the failure rate of computational scheme; Concrete formula is as follows:

$L_{\mathrm{Length}}=\frac{X_L}{X_B}---\left(3\right)$

Wherein, L _lengththe length that represents required circuit, X _lrepresent the impedance of required circuit, X _brepresent circuit reference impedance;

The computing method of event of failure probability P are:

$P=\underset{i\∈U}{\mathrm{\Π}}{u}_i\underset{j\∈A}{\mathrm{\Π}}(1-u_j)=\frac{\underset{i\∈U}{\mathrm{\Π}}{u}_i\underset{j\∈S}{\mathrm{\Π}}(1-u_j)}{\underset{k\∈U}{\mathrm{\Π}}(1-u_k)}---\left(4\right)$

Wherein, P represents the probability of happening of event of failure, and S represents all devices collection, and U represents faulty equipment collection, and A represents normal device collection, and u is element failure rate.

Described step S3 comprises following sub-step:

The computing method of S3-1 sensitivity matrix are as follows:

The homeostasis condition of electric system is by n nonlinear network the Representation Equation, and under a certain running status, this equational compact form is:

f(x,u)＝0 (5)

Wherein, x represents by control variable (dependent variable) column vector; U represents control variable (independent variable) column vector;

Because equation number and the form of formula (5) depends on variables collection how to choose x and u, thereby concrete form may have larger difference; Equational form and number also can change to some extent along with coordinate form (polar coordinates or the rectangular coordinate) difference of selecting simultaneously; When system is moved in given stable operation situation, formula (5) becomes:

f(x ₀,u ₀)＝0 (6)

Wherein, x ₀and u ₀while representing respectively system stable operation situation, by control variable and control variable;

After running status changes, when x and u have respectively a departure Δ x and Δ u, the homeostasis equation of system becomes formula (7):

f(x ₀+Δx,u ₀+Δu)＝0 (7)

Formula (7) is at operating point x ₀and u ₀place carries out Taylor series expansion, and omits second order and above higher order term:

$f(x_0,u_0)+\frac{\∂f}{\∂x}\mathrm{\Δx}+\frac{\∂f}{\∂u}\mathrm{\Δu}=0---\left(8\right)$

By formula (6), above formula becomes:

$\frac{\∂f}{\∂x}\mathrm{\Δx}+\frac{\∂f}{\∂u}\mathrm{\Δu}=0---\left(9\right)$

The citation form that formula (9) is sensitivity equation;

Matrix of coefficients in formula with also referred to as Jacobian matrix;

By the linear model of formula (9), the linear relationship of controlled variable Δ u and controlled variable Δ x is:

$\mathrm{\Δx}=-{\left(\frac{\∂f}{\∂x}\right)}^{-1}\frac{\∂f}{\∂u}\mathrm{\Δu}=\mathrm{S\Δu}---\left(10\right);$

Wherein, be called sensitivity matrix;

The load loss method that S3-2 utilizes safe Correction Strategies calculating event of failure to cause is as follows:

(as shown in Figure 1) carry out cutting load operation: first according to overload branch road or the node that crosses the border, from sensitivity matrix, select the sensitivity coefficient vector of each node to overload branch road or the node that crosses the border, and the sensitivity coefficient vector extracting is sorted by order from small to large; Then from carrying out safe Correction Strategies adjusting on the occasion of being up to 0 direction by absolute value respectively with negative value two ends, on the occasion of the carry out PQ, PV node that locate, weaken the operation that generated power is exerted oneself, negative value place carries out the operation of PV node cutting load, in order to guarantee the active balance of system, the amount of attenuation that generated power is exerted oneself and cutting load amount will be consistent; Final load loss is eliminates the load that phenomenon of the failure (branch road overload or node cross the border) need to be cut away.

In described step S4, the computing method of losing Load Probability risk indicator are as follows:

R _i＝P _i×I _i (11)

Wherein, R _ithe mistake Load Probability risk indicator that represents event of failure i, P _iand I _irepresent respectively the probability that event of failure i occurs and the mistake load value causing.

In described step S5, adopt minimum element failure rate as reference element failure rate, according to defining for ordinary accident with compared with the load loss of major break down in 599 commands, the mistake Load Probability risk indicator cut off value of minor accident and ordinary accident and ordinary accident and be respectively compared with the mistake Load Probability risk indicator cut off value of major break down: system total load is multiplied by 4% and is multiplied by reference element failure rate and system total load again and is multiplied by 7% and is multiplied by reference element failure rate again; Due to the minor accident impact weak on system, minor accident is included into operation and is responsible for region, other accidents are included into plans responsible region, so the mistake Load Probability risk cut off value that operation is responsible between region and the responsible region of planning can be decided to be:

A＝S _Loadloss×4％×u _Base (12)

Wherein, A represents to move the mistake Load Probability risk cut off value of being responsible between region and the responsible region of planning, S _loadlossand u _baserepresent respectively system total load and reference element failure rate.

In described step S6, according to formula (11), can obtain the mistake Load Probability risk indicator R of each event of failure _i, according to step 5, can obtain planning and the mistake Load Probability risk cut off value A moving, by R _icompare with A, if R _i< A event of failure belongs to operation and is responsible for, if R _i> A event of failure belongs to planning and is responsible for.

Technical characterstic of the present invention and effect: the present invention takes into full account Power System Planning field and operation field feature separately, take into full account the impact of the operating uncertain factor of power system device on system, incorporated probability understanding, a kind of partition method that comes dividing system planning and operation based on mistake Load Probability risk indicator is proposed, the method result of calculation can objectively respond the influence degree of the various event of failures of electric system to system, can solve planning and the divisions of responsibility can problem of moving comparatively simple and effectively, auxiliary Electric Power Network Planning and the identification of operations staff's discrimination system weak link and unreasonable grid structure, the relation of coordinated planning and operation, balanced system reliability requirement and construction cost etc.

What by the present invention, proposed comes dividing system planning and the partition method moving based on losing Load Probability risk indicator, can truly reflect the influence degree of different faults event to system, and rationally determine planning and the responsibility subregion moving, solve targetedly the weak link of electric system; According to risk level, the present invention is divided into two classes by event of failure: need to be strengthened and planned the accident solving and can strengthen the accident that operation maintenance solves by traffic department by Fa Ce department.As rational appraisal procedure, the present invention can help system planning personnel and operations staff judge which measure which event of failure need to solve and take solve, effectively suit engineering reality, and with this coherent system planning and operation, define respective responsibilities and obligation, relation between balanced system reliability requirement and construction cost, compare technological means in the past, determine the responsibility of planning and operation department more accurate science, two departments that prevent scold mutually when electrical network exists potential safety hazard or has an accident, the situation of shift onto each other occurs, there is very strong practical value.

Accompanying drawing explanation

Fig. 1 is cutting load operation chart;

Fig. 2 is the subregion frame diagram based on Risk Theory coordinated planning and operation area;

Fig. 3 is the process flow diagram based on Risk Theory coordinated planning and operation;

Fig. 4 is the mistake load value histogram that the event of failure of instance system causes;

Fig. 5 is that schematic diagram is divided in the event of failure region of instance system.

Embodiment

Below by coming dividing system planning and the partition method of operation to be applied to, in Guangdong Power Grid 500kV planning system example in 2015, the application of partition method is described by of the present invention based on losing Load Probability risk indicator.

In the electric system of this embodiment, comprise the whole power transmission network equipment of Guangdong Power Grid, covered the Guangdong 23Ge prefecture-level cities such as Guangzhou, Shenzhen, Dongguan, Foshan Distribution Network Equipment and with Guangxi, Hong Kong, Macao's Power System Interconnection equipment etc., comprise altogether 2626 buses, article 1475, transmission line of electricity, 2064 transformers and 198 generators.

The present invention proposes to lose Load Probability risk indicator, be used for reflecting the actual influence degree of electric power system fault event to system, and take into full account the uncertainty of following electrical network, set up a set of partition method based on losing Load Probability risk indicator and come dividing system planning and operation.

Referring to Fig. 3, it is the process flow diagram based on Risk Theory coordinated planning and operation, and concrete steps of the present invention are as follows:

Step 1: read in rack basic data: the electric parameter of instance system 500kV circuit, bus, transformer, generator and load distribute; According to the requirement of Guangdong Power Grid safety and stability guide rule, access line, bus and transformer N-1 accident are as accident collection;

Step 2: according to accident collection, carry out trend calculating by BPA software, judged whether the problems such as branch road overload or voltage cross the border, instance system is carried out to static security analysis, record branch road overload values or the voltage value of crossing the border;

Meanwhile, according to the dependability parameter of the bus of instance system historical statistics, circuit and transformer, can try to achieve the element failure rate of bus, circuit and transformer and try to achieve event of failure probability.Specific as follows:

The historical reliability statistics of bus, circuit and transformer based on instance system statistics, can obtain 500kV circuit, bus, transformer average availability coefficient and the planned outage time of circuit, unplanned idle time and forced outage duration over the years, as shown in table 1.

Table 1 instance system 500kV component reliability parameter

According to formula (1) and formula (2), can obtain the failure rate of circuit, transformer, bus, as shown in table 2.

Table 2 instance system 500kV element failure rate

Element	Circuit	Bus	Transformer
				Failure rate	9.0E-04	5.4E-04	7.0E-04

According to formula (4), by the failure rate of each element, can be tried to achieve the probability of malfunction of event of failure, as shown in table 3.

The probability of malfunction of table 3 instance system 500kV event of failure

In addition, the reference impedance that adopts the impedance of 4*720 line style unit length to calculate as line length herein, and according to formula (3) computational scheme length.

Step 3: meter sensitivity matrix, and utilize safe Correction Strategies to calculate the load loss that event of failure causes based on sensitivity matrix; Result of calculation shows, has 28 event of failures and can cause system to lose load, and wherein transformer fault event is 23, and bus-bar fault event is 3, and line fault event is 2, and the mistake load value that these event of failures cause as shown in Figure 3.

Step 4: the load loss causing according to event of failure probability and event of failure, utilize formula (11) to calculate the risk indicator of this each event of failure of instance system, result of calculation is as shown in table 4.

The risk indicator of each event of failure of table 4 instance system

Event of failure is compiled

Event of failure is general

Lose load value

Losing Load Probability risk refers to

Number	Rate		Mark
				1	4.70E-03	954	4.48445
2	4.70E-03	500	2.35034
				3	4.70E-03	500	2.35034
4	4.70E-03	483	2.27043
				5	4.70E-03	483	2.27043
6	6.91E-04	2530	1.74917
				7	4.70E-03	369	1.73455
8	4.70E-03	369	1.73455
				9	4.70E-03	251	1.17987
10	4.70E-03	251	1.17987
				11	4.70E-03	251	1.17987
12	4.70E-03	246	1.15637
				13	4.70E-03	246	1.15637
14	4.70E-03	118	0.55468
				15	4.70E-03	118	0.55468
16	4.70E-03	118	0.55468
				17	4.70E-03	118	0.55468
18	4.70E-03	100	0.470068
				19	4.70E-03	100	0.470068
20	6.91E-04	581.1	0.401653
				21	6.91E-04	504	0.348398
22	4.70E-03	53	0.249136
				23	4.70E-03	53	0.249136
24	4.70E-03	29	0.13632
				25	4.70E-03	29	0.13632
26	4.70E-03	29	0.13632
				27	5.18E-05	138.5	0.007177
28	6.11E-05	117.1	0.007159

Step 5: State Council's 599 command regulations, in the power grid accident of provincial region, cause load loss to surpass 4% for ordinary accident, surpassing 7% is compared with major break down, in order to guarantee the safety and stability of electric system, the present invention adopts bus-bar fault rate that in instance system, element failure rate is minimum as reference element failure rate, event of failure is done to enough conservative region division.Because instance system total load is about 80000MW, the failure rate of 500kV bus N-1 accident is 6.91E-04, therefore according to defining for ordinary accident with compared with the load loss of major break down in 599 commands, minor accident, ordinary accident and be respectively 80000 * 4% * 0.000691=2.21 and 80000 * 7% * 0.000691=3.87 compared with the mistake Load Probability risk indicator cut off value of major break down in this instance system.Due to the minor accident impact weak on system, the present invention is included into operation by minor accident and is responsible for region, other accidents are included into planning and are responsible for region, so according to formula (12), operation is responsible for region and is planned that the mistake Load Probability risk cut off value A being responsible between region can be decided to be 2.21.

Step 6: according to the mistake Load Probability risk indicator of each event of failure in table 2, can analyze and obtain being positioned at the different event of failures of being responsible for region, subregion result as shown in Figure 5.In Fig. 5, put under event of failure that planning is responsible for region due to its probability of malfunction or fault effects larger, value-at-risk is higher, by traffic department, be responsible for processing and easily cause more serious load loss, safe and stable operation to electric system affects greatly, thereby should process by modes such as the planning of planning department strengthening system, newly-built grid equipment, adjustment grid structures.For being put under scheduling, be responsible for region; they have lower probability of malfunction or fault effects; risk indicator is lower; although these event of failures can cause impact to a certain degree to system; but by planning means, process often with high costs; economic benefit is poor, so Ke You traffic department passes through to protect, stability control equipment, and the modes such as plant maintenance, transformation and upgrading of strengthening are solved.

Claims (6)

1. based on Risk Theory, coordinate a partition method for Power System Planning and operation, it is characterized in that comprising the following steps:

S1 reads in rack basic data, comprises that the operational factor of bus, generator, transformer, overhead transmission line and load distribute; And stablize in guide rule and electric power safety accident investigation regulation and require the accident pattern of check to choose accident collection according to electric power safety;

S2 carries out trend calculating based on BPA software, has judged whether branch road overload or the voltage problem of crossing the border, and carries out static security analysis; Meanwhile, set up element failure rate and event of failure probability model, computing element failure rate and event of failure probability;

S3, based on sensitivity matrix, utilizes safe Correction Strategies to calculate the load loss that event of failure causes;

The load loss that S4 causes according to event of failure probability and event of failure, calculates and loses Load Probability risk indicator, carries out quantifying risk assessment;

S5 is according to incident classification criterion and system ability to bear, and the load loss value that each incident classification of judgement system allows, clearly plans and move the different incident classifications of being responsible for separately; Event of failure is done to region and divide, during the cut off value of zoning, adopt minimum element failure rate as reference element failure rate, in conjunction with the responsibility cut off value of calculating between planning and operation;

Minor accident and ordinary accident and ordinary accident and be respectively compared with the mistake Load Probability risk indicator cut off value of major break down: system total load is multiplied by 4% and is multiplied by reference element failure rate and system total load again and is multiplied by 7% and is multiplied by reference element failure rate again; Minor accident is included into operation and is responsible for region, and other accidents are included into planning and are responsible for region; Operation is responsible for region and mistake Load Probability risk cut off value between the responsible region of planning and is decided to be system total load and is multiplied by 4% and is multiplied by reference element failure rate again;

S6 compares the planning of step S5 gained and the mistake Load Probability risk indicator of operation responsibility cut off value and event of failure, determines the area of responsibility that event of failure is affiliated.

2. the partition method of coordinating Power System Planning and operation based on Risk Theory according to claim 1, is characterized in that: in described step S2, the computing method of element failure rate u are as follows:

$u=\frac{F_{\mathrm{outage}}\&times;T_{\mathrm{repair}}}{8760}=\frac{\mathrm{\&lambda;}}{\mathrm{\&lambda;}+\mathrm{\&mu;}}=1-\frac{D}{100}---\left(1\right);$

Wherein, F _outagethe average stoppage in transit frequency of indication equipment, unit is Failure count/year, T _repairmean time to repair after indication equipment generation forced outage, unit is h, and λ is crash rate, and unit is Failure count/year, and μ is repair rate, and unit is for repairing number of times/year, and D represents availability coefficient;

The computing method of availability coefficient D are as follows:

$D=\frac{T_1}{T_1+T_2+T_3}---\left(2\right);$

Wherein, T ₁represent pot life (h), T ₂represent the planned outage time (h), T ₃represent unplanned idle time (h); The planned outage time refers to the time of the maintenance of stopping transport according to original plan or stop transport, and unplanned idle time refers to not in inside the plan fault idle time;

In the time cannot obtaining the probability of malfunction of circuit based on statistics, by the proportionate relationship of each line impedance and reference impedance, estimate the length of each circuit, thus the failure rate of computational scheme; Concrete formula is as follows:

$L_{\mathrm{Length}}=\frac{X_L}{X_B}---\left(3\right);$

Wherein, L _lengththe length that represents required circuit, X _lrepresent the impedance of required circuit, X _brepresent circuit reference impedance;

The computing method of event of failure probability P are:

$P=\underset{i\&Element;U}{\mathrm{\&Pi;}}{u}_i\underset{j\&Element;A}{\mathrm{\&Pi;}}(1-u_j)=\frac{\underset{i\&Element;U}{\mathrm{\&Pi;}}{u}_i\underset{j\&Element;S}{\mathrm{\&Pi;}}(1-u_j)}{\underset{k\&Element;U}{\mathrm{\&Pi;}}(1-u_k)}---\left(4\right);$

Wherein, P represents the probability of happening of event of failure, and S represents all devices collection, and U represents faulty equipment collection, and A represents normal device collection, and u is element failure rate.

3. the partition method of coordinating Power System Planning and operation based on Risk Theory according to claim 1, is characterized in that: described step S3 comprises following sub-step:

S3-1 meter sensitivity matrix

The homeostasis condition of electric system is by n nonlinear network the Representation Equation, and under a certain running status, this equational compact form is:

f(x,u)＝0 (5)；

Wherein, x represents to be controlled variable column vector; U represents control variable column vector;

Equation number and the form of formula (5) depend on variables collection how to choose x and u, thereby concrete form can be had any different; Simultaneously equational form and number also can be different and change along with the coordinate form of selecting; When system is moved in given stable operation situation, formula (5) becomes:

f(x ₀,u ₀)＝0 (6)；

Wherein, x ₀and u ₀while representing respectively system stable operation situation, by control variable and control variable;

After running status changes, when x and u have respectively a departure Δ x and Δ u, the homeostasis equation of system becomes formula (7):

f(x ₀+Δx,u ₀+Δu)＝0 (7)；

Formula (7) is at operating point x ₀and u ₀place carries out Taylor series expansion, and omits second order and above higher order term:

$f(x_0,u_0)+\frac{\&PartialD;f}{\&PartialD;x}\mathrm{\&Delta;x}+\frac{\&PartialD;f}{\&PartialD;u}\mathrm{\&Delta;u}=0---\left(8\right);$

By formula (6), above formula becomes:

$\frac{\&PartialD;f}{\&PartialD;x}\mathrm{\&Delta;x}+\frac{\&PartialD;f}{\&PartialD;u}\mathrm{\&Delta;u}=0---\left(9\right);$

The citation form that formula (9) is sensitivity equation;

Matrix of coefficients in formula with also referred to as Jacobian matrix;

By the linear model of formula (9), the linear relationship of controlled variable Δ u and controlled variable Δ x is:

$\mathrm{\&Delta;x}=-{\left(\frac{\&PartialD;f}{\&PartialD;x}\right)}^{-1}\frac{\&PartialD;f}{\&PartialD;u}\mathrm{\&Delta;u}=\mathrm{S\&Delta;u}---\left(10\right);$

Wherein, be called sensitivity matrix;

S3-2 utilizes safe Correction Strategies to calculate the load loss that event of failure causes

Carry out cutting load operation: first according to overload branch road or the node that crosses the border, from sensitivity matrix, select the sensitivity coefficient vector of each node to overload branch road or the node that crosses the border, and the sensitivity coefficient vector extracting is sorted by order from small to large; Then from carrying out safe Correction Strategies adjusting on the occasion of being up to 0 direction by absolute value respectively with negative value two ends, on the occasion of the carry out PQ, PV node that locate, weaken the operation that generated power is exerted oneself, negative value place carries out the operation of PV node cutting load, in order to guarantee the active balance of system, the amount of attenuation that generated power is exerted oneself and cutting load amount will be consistent; Final load loss is eliminates the load that phenomenon of the failure need to be cut away.

4. the partition method of coordinating Power System Planning and operation based on Risk Theory according to claim 1, is characterized in that: in described step S4, the computing method of mistake Load Probability risk indicator are as follows:

R _i＝P _i×I _i (11)；

Wherein, R _ithe mistake Load Probability risk indicator that represents event of failure i, P _iand I _irepresent respectively the probability that event of failure i occurs and the mistake load value causing.

5. the partition method of coordinating Power System Planning and operation based on Risk Theory according to claim 1, is characterized in that: in described step S5, operation is responsible for region and can be decided to be with planning the mistake Load Probability risk cut off value between responsible region:

A＝S _Loadloss×4％×u _Base (12)；

Wherein, A represents to move the mistake Load Probability risk cut off value of being responsible between region and the responsible region of planning, S _loadlossand u _baserepresent respectively system total load and reference element failure rate.

6. the partition method of coordinating Power System Planning and operation based on Risk Theory according to claim 1, is characterized in that: in described step S6, obtain the mistake Load Probability risk indicator R of each event of failure according to formula (11) _i, according to step S5, can obtain planning and the mistake Load Probability risk cut off value A moving, by R _icompare with A, if R _i< A event of failure belongs to operation and is responsible for, if R _i> A event of failure belongs to planning and is responsible for.