CN102637289B - Containing the security value appraisal procedure of the Power System Planning scheme of large-scale wind power - Google Patents

Abstract

The present invention relates to Expansion Planning and the safety analysis field of electric system.Safety evaluation for programme provides a Systematization method and comparatively perfect estimation flow, the technical scheme that the present invention takes is, containing the security value appraisal procedure of the Power System Planning scheme of large wind-powered electricity generation, comprise the following steps: the first step, set up power supply, the electrical network unified planning model of taking into account static system and dynamic security value, second step, starts to carry out detailed working train family to candidate's programme; 3rd step, after Unit Combination, by the period safety evaluation is carried out to system; 4th step, to this contingency set, carries out the optimization of security Comprehensive Control by " active constraint relaxation method ", obtains optimum prevention and control contingency set and the division of emergency control contingency set; 5th step, draws the total cost of programme, in this, as the foundation of program evaluation scheme.The present invention is mainly used in Expansion Planning and the safety analysis of electric system.

Description

Containing the security value appraisal procedure of the Power System Planning scheme of large-scale wind power

Technical field

The present invention relates to Expansion Planning and the safety analysis field of electric system, specifically, relate to the security value appraisal procedure of the Power System Planning scheme containing large-scale wind power.

Background technology

THE WIND ENERGY RESOURCES IN CHINA is enriched, and multiple large wind-powered electricity generation construction of base has entered among the overall planning of national Wind Power Development.But because wind-powered electricity generation has intermittence and uncertainty, it concentrates access all can bring great impact to the planning of electric system and operation.Large-scale wind electricity base is mostly positioned at the tip of electric system in addition, power cannot on-site elimination, and wind-powered electricity generation has construction period short feature, causes the large-scale development of wind-powered electricity generation must carry out comprehensive coordination with other form power supplys and electrical network and plans [1].But the research at present, planned about wind-powered electricity generation and electric network coordination also also imperfection.Document [2] with large-scale wind power field group for research object propose consider its power converge characteristic send transmission line capability static optimization method outside.The uncertain factors such as the output power of wind energy turbine set, load variations describe in the mode of scene analysis and probability calculation by document [3], establish the Flexible planning model of Transmission Expansion Planning in Electric and use genetic algorithm for solving.Document [4] processes the uncertainty of wind power output by the Chance-Constrained Programming Model setting up power transmission network.Due to the time variation of wind speed, output of wind electric field is all the time all in continuous change, and be the system condition (lack and the emulation of safety economy Optimized Operation is carried out to the system of by the hour and even more short time interval annual 8760 hours of planning level year) that method based on many scenes or the method based on probability statistics are all difficult to simulate wind speed and change suddenly, be therefore also just difficult to adjustment (frequency modulation peak regulation) characteristic of accurately simulation system and take into account the margin capacity of system.Although existing document employs the reliability [5,6] of Monte Carlo simulation approach assessment wind power integration systems organization scheme, do not relate to the dynamic security of system after wind power integration, more do not consider power supply, power network planning scheme to the impact of environment.

List of references:

[1] white Jian Hua, Xin Songxu, Jia Dexiang. the planning that China's wind-powered electricity generation large-scale development faces and analysis of problem of operation. power technology economy, 2009,21 (2): 7-11.

[2] Mu Gang, Cui Yang, Yan Gangui. determine the static comprehensive optimization method of power gathering and output electric capacity of wind power station group. Proceedings of the CSEE, 2011,31 (1): 15-19.

[3] Yuan Yue, Wu Bowen, Li Zhenjie, etc. the Transmission Network Flexible containing Large Scale Wind Farm Integration based on many scenes probability is planned. Electric Power Automation Equipment, 2009,29 (10): 8-12.

[4]YuH，ChungCY，WongKP，etal.Achanceconstrainedtransmissionnetworkexpansionplanningmethodwithconsiderationofloadandwindfarmuncertainties.IEEETransactionsonPowerSystems，2009，24(3)：1568-1576.

[5]BillintonR，WangdeeW，Reliability-basedtransmissionreinforcementplanningassociatedwithlarge-scalewindfarms，IEEETransactionsonPowerSystems，2007，22(1)：34-41

[6]BillintonR，GaoY，KarkiR.CompositeSystemAdequacyAssessmentIncorporatingLarge-ScaleWindEnergyConversionSystemsConsideringWindSpeedCorrelation.IEEETransactionsonPowerSystems，2009，24(3)：1375-1382

Summary of the invention

The present invention is intended to solution and overcomes the deficiencies in the prior art, safety evaluation for programme provides a Systematization method and comparatively perfect estimation flow, assessment into social value is widened from traditional technological economics evaluation in the visual angle of Power System Planning, for the impact embedding chain accident in programme safety evaluation provides instrument, make the assessment of security of system more comprehensive, reasonable.For achieving the above object, the technical scheme that the present invention takes is, containing the security value appraisal procedure of the Power System Planning scheme of large-scale wind power, comprises the following steps:

The first step, sets up power supply, the electrical network unified planning model of taking into account static system and dynamic security value, and the cost wherein containing the Power System Planning scheme of large-scale wind power mainly comprises the content of following several aspect: the 1. cost of development C of wind-powered electricity generation _w; 2. the investment of other power supplys and operation expense C except wind-powered electricity generation _g; 3. the investment of power transmission and transforming equipment and operation expense C _n; 4. prevention and control cost C ^pre; 5. emergency control cost C ^emer; 6. the social loss cost C that control measure are invalid ^out; 7. Web-based exercise C ^loss; First can try to achieve 1. according to the given data of power generating facilities and power grids-3.;

Second step, start to carry out detailed working train family to candidate's programme, thus the security value of evaluation scheme: first according to prediction every day day part node load and wind power output carry out Unit Combination a few days ago, with determine load and wind power output fluctuation condition under, whether current power source planning scheme can meet the requirement of system frequency modulation peak regulation and day part margin capacity;

3rd step, after Unit Combination, by the period safety evaluation is carried out to system: in a certain period, first forecast accident scanning and chain accident forecast is carried out, filter out the contingency set of the single failure that this period needs are considered and the contingency set of the chain stoppage in transit of circuit may be caused, the common contingency set forming this period;

4th step, to this contingency set, the optimization of security Comprehensive Control is carried out by " active constraint relaxation method ", obtain optimum prevention and control contingency set and the division of emergency control contingency set, and in the prevention and control optimization and emergency control Optimized model of each period use safety territory method taking into account of achieving that dynamic security retrains, thus optimum control scheme and corresponding prevention and control cost, the emergency control cost of this period can be obtained and control invalid social loss, namely obtain the total safety cost of system in this period;

5th step, by the assessment to annual 8760 hours of planning level year, and in conjunction with the initial outlay of power supply, power network planning scheme, namely the annual total security cost of programme is drawn, in conjunction with the first step, the total cost of programme can be drawn, in this, as the foundation of program evaluation scheme.

4th step is further refined as:

(1) the Power System Planning model of considering security cost: the Power System Planning scheme that after considering security cost C, total cost is minimum, its mathematical description is as follows:

minC＝C _w+C _g+C _n+C ^pre+C ^emer+C ^out+C ^loss(1)

(2) security value estimation flow:

First according to prediction every day day part node load and wind power output carry out Unit Combination a few days ago, with determine load and wind power output fluctuation condition under, whether current power source planning scheme can meet the requirement of system frequency modulation peak regulation and day part margin capacity, the method of Unit Combination is the net load curve first drawing system, namely load deducts wind power output, then combines conventional power unit accordingly; Unit Combination herein does not consider that network constraint is interior;

After Unit Combination, by the period safety analysis is carried out to system, in a certain period, first forecast accident scanning and chain accident forecast is carried out, filter out the contingency set of the single failure that this period needs are considered and the contingency set of the chain stoppage in transit of circuit may be caused, the contingency set of common this period of formation, to this contingency set, optimized by security Comprehensive Control, obtain optimum prevention and control contingency set and the division of emergency control contingency set, thus the optimum control scheme of this period and corresponding prevention and control cost can be obtained, emergency control cost and control invalid social loss, namely total safety cost of system in this period is obtained, by the assessment to annual 8760 hours of planning level year, and in conjunction with power supply, the initial outlay of power network planning scheme, the annual total security cost of programme can be drawn.

The security method of cost accounting is refined as:

If grid is by n+1 node, n _bbar circuit forms, and its interior joint 0 is slack bus, with G:={1, and 2 ..., n _grepresent the set of generator node; Use L:={n _g+ 1 ..., n} represents the set of load bus; The set of whole node is represented, i.e. N=G ∪ L ∪ 0 with N; The set of whole circuit is represented with B; CTS is the contingency set of single failure, and CTS is by contingency set and the preventive control collection Γ that need take Control Measure _pwith emergency control measure need be taked to guarantee contingency set and the emergency control collection Γ of stability _ecomposition; CTS _cfor the contingency set of the chain stoppage in transit of circuit may be caused:

3.1 prevention and control cost C ^pre

Prevention and control cost is made up of three parts:

$C^{\mathrm{pre}}=C_f^{\mathrm{pre}}+C_e^{\mathrm{pre}}+C_c^{\mathrm{pre}}---\left(2\right)$

In formula be respectively prevention and control fuel cost, prevention and control Environmental costs and prevention and control load summate cost;

The probability that accident of getting when calculating prevention and control cost occurs is 1;

(1) prevention and control fuel cost

If the result of Unit Combination directly meets the various security constraints of system through verifying, then illustrative system is without the need to carrying out extra generation adjustment, and now prevention and control cost is 0; Otherwise, then need to solve the fuel cost that following two optimization problems try to achieve prevention and control:

First problem is the security constrained economic dispatch problem based on AC power flow, and its model is as follows:

$\min F_C=\frac{1}{2}{P}^T{H}_{P}P+f_P^{T}P---\left(3\right)$

$s.t.P_i=U_i\underset{j\∈i}{\mathrm{\Σ}}{U}_j[G_{\mathrm{ij}}\cos {\mathrm{\θ}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\θ}}_{\mathrm{ij}}]\∀i\∈N---\left(4\right)$

$Q_i=U_i\underset{j\∈i}{\mathrm{\Σ}}{U}_j[G_{\mathrm{ij}}\sin {\mathrm{\θ}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\θ}}_{\mathrm{ij}}]\∀i\∈N---\left(5\right)$

$\left(I\right)P_{g,i}^m\≤P_{g,i}\≤P_{g,i}^M\∀i\∈G---\left(6\right)$

$Q_{g,i}^m\≤Q_{g,i}\≤Q_{g,i}^M\∀i\∈G---\left(7\right)$

$-P_l^M\≤P_l\≤P_l^M\∀l\∈B---\left(8\right)$

$U_i^m\≤U_i\≤U_i^M\∀i\∈N---\left(9\right)$

$\underset{\&ForAll;i\&Element;G\&cup;L}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}{P}_i<1\&ForAll;l\&Element;{\mathrm{\&Gamma;}}_p\&Subset;\mathrm{CTS}---\left(10\right)$

Wherein, F _cfor considering Active Generation cost during security constraint, H _p, f _pbe respectively cost of electricity-generating quadratic term matrix of coefficients and Monomial coefficient vector, P is the meritorious injection vector of node, P _i, Q _ibe respectively the meritorious injection of node i and idle injection, G _ij, B _ijbe respectively real part and the imaginary part of bus admittance matrix i-th j element, θ _ijfor branch road ij two ends node voltage phase angle difference; P _{g, i}, Q _{g, i}what be respectively generator i meritoriously exerts oneself and idlely to exert oneself, and its upper and lower limit is respectively p _lfor the effective power flow of branch road l, its maximum permission through-put power is u _ifor the magnitude of voltage of an i, its upper and lower limit is respectively α _{(l, i)}for i-th component of the Practical Dynamic Security Region lineoid coefficient for forecast accident l, subscript T represents transposition, wherein Practical Dynamic Security Region and PDSR, PracticalDynamicSecurityRegion;

Second Problem is the Economic Dispatch Problem without security constraint, and its model is as follows:

$\min F_U=\frac{1}{2}{P}^T{H}_{P}P+f_P^{T}P---\left(11\right)$

$\left(\mathrm{II}\right)s.t.\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_i=0---\left(12\right)$

$P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M\&ForAll;i\&Element;G---\left(13\right)$

Wherein, F _ufor disregarding Active Generation cost during security constraint;

The prevention and control fuel cost of t period of system is security constrained economic dispatch and the difference without security constrained economic dispatch objective function

$C_{f,t}^{\mathrm{pre}}=F_{C,t}-F_{U,t}---\left(14\right)$

In formula be the prevention and control fuel cost of t period, F _{c, t}, F _{u, t}be respectively t time safety constrained optimum trend and the target function value without security constraint optimal load flow;

(2) prevention and control Environmental costs

Second component of prevention and control is prevention and control Environmental costs, is made up of two parts: prevention and control environmental loss cost and prevention and control environmental countermeasure cost, is located at the meritorious vector that injects of node corresponding to the optimum solution of segment model I and model II during t and is respectively with the then prevention and control Environmental costs of t period for

$C_{e,t}^{\mathrm{pre}}=[C_{e,t}^{\mathrm{loss}}\left(P_{C,t}^{*}\right)-C_{e,t}^{\mathrm{loss}}\left(P_{U,t}^{*}\right)]+[C_{e,t}^{\mathrm{meas}}\left(P_{C,t}^{*}\right)-C_{e,t}^{\mathrm{meas}}\left(P_{U,t}^{*}\right)]---\left(15\right)$

Part I in formula for prevention and control environmental loss cost, be count security constraint and disregard security constraint environmental loss cost difference; Part II for prevention and control environmental countermeasure cost, be count security constraint and the difference disregarded with generated output relevant portion in security constraint environmental countermeasure cost, refer to second component of environmental countermeasure cost, namely relevant with generated output size environmental countermeasure cost component; Should be noted that, environmental countermeasure cost is former should be made up of two components, one-component is the cost of investment of environment protection equipment, for fixed value, second component is the operation and maintenance cost of equipment, relevant to generated output, due to environment protection equipment once purchase, its cost of investment is in operation irrelevant with Prevention and control strategy;

In formula (15), the environmental loss cost of t period for

$C_{e,t}^{\mathrm{loss}}\left(P_t^{*}\right)=\underset{h=1}{\overset{H}{\mathrm{\&Sigma;}}}{c}_h{Q}_{h,t}\left(P_t^{*}\right)---\left(16\right)$

Wherein, c _hbe the specific emissions cost of h kind pollutant, be the discharge capacity of t period h kind pollutant, or the optimum solution of the corresponding segment model I and model II when t respectively, H is the pollution species number considered;

The environmental countermeasure cost that t period is relevant with generated output for

$C_{e,t}^{\mathrm{meas}}\left(P_t^{*}\right)=\underset{h=1}{\overset{H}{\mathrm{\&Sigma;}}}{c}_{h,t}^o\left(P_t^{*}\right)---\left(17\right)$

Wherein, for the t period is for the operation and maintenance cost of the environment protection equipment of h kind pollutant;

(3) prevention and control load summate cost

As follows with the minimum system compensation Controlling model for target of load summate total amount:

$\min \underset{j\&Element;L}{\mathrm{\&Sigma;}}{P}_j^{\mathrm{cut}}---\left(18\right)$

$s.t.P_i^{\&prime;}=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}]\&ForAll;i\&Element;N---\left(19\right)$

$Q_i^{\&prime;}=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}]\&ForAll;i\&Element;N---\left(20\right)$

$P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M\&ForAll;i\&Element;G---\left(21\right)$

$\left(\mathrm{III}\right)Q_{g,i}^m\&le;Q_{g,i}\&le;Q_{g,i}^M\&ForAll;i\&Element;G---\left(22\right)$

$U_i^m\&le;U_i\&le;U_i^M\&ForAll;i\&Element;N---\left(23\right)$

$-P_l^M\&le;P_l\&le;P_l^M\&ForAll;l\&Element;B---\left(24\right)$

$0\&le;P_j^{\mathrm{cut}}\&le;P_{d,j}\&ForAll;j\&Element;L---\left(25\right)$

$0\&le;Q_j^{\mathrm{cut}}\&le;Q_{d,j}\&ForAll;j\&Element;L---\left(26\right)$

$\underset{\&ForAll;i\&Element;G}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}{P}_{g,i}+\underset{\&ForAll;i\&Element;L}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}(P_{d,i}-P_i^{\mathrm{cut}})<1\&ForAll;l\&Element;{\mathrm{\&Gamma;}}_p\&Subset;\mathrm{CTS}---\left(27\right)$

In formula be respectively meritorious, the load or burden without work reduction on a jth load bus, P _{d, j}, Q _{d, j}be respectively a jth node to gain merit and load or burden without work; α _{(l, i)}for the Practical Dynamic Security Region for forecast accident l, i.e. i-th component of PDSR, PracticalDynamicSecurityRegion lineoid coefficient; for the power factor of node i reduction plans, and be constant for node i, the burden with power reduction of t period node i can be obtained by above load summate model then the load summate cost of this period is

$C_{c,t}^{\mathrm{pre}}=\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_{i,t}^{\mathrm{cut}}\&times;{\mathrm{IEAR}}_i---\left(28\right)$

IEAR in formula _ifor the load loss Assessment Rate of node i, namely cut down the loss of outage cost of specific load;

Like this, whole system is made up of prevention and control fuel cost, prevention and control Environmental costs and prevention and control load summate cost three part at the prevention and control cost of period t

$C_t^{\mathrm{pre}}=C_{f,t}^{\mathrm{pre}}+C_{e,t}^{\mathrm{pre}}+C_{c,t}^{\mathrm{pre}}---\left(29\right)$

Corresponding with formula (2), the prevention and control cost of system whole year is

$C^{\mathrm{pre}}=C_f^{\mathrm{pre}}+C_e^{\mathrm{pre}}+C_c^{\mathrm{pre}}=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}(C_{f,t}^{\mathrm{pre}}+C_{e,t}^{\mathrm{pre}}+C_{c,t}^{\mathrm{pre}})---\left(30\right)$

3.2 emergency control cost C ^emer

Namely the cost of emergency control measure constitutes the Part II of security cost, is expressed as follows:

C ^emer＝C ^repl+C ^repair+C ^load(31)

C in formula ^repl, C ^repair, C ^loadbe respectively alternative cost, shutdown and start-up cost and load rejection cost, controlling cost of emergency control measure e is taked to comprise shutdown and start-up cost to a certain accident l that forecast accident is concentrated, alternative cost and load rejection cost three part, to t the period in planning level year, emergency control cost can be expressed as

$C_t^{\mathrm{emer}}(l,e)=C_t^{\mathrm{repair}}(l,e)+C_t^{\mathrm{repl}}(l,e)+C_t^{\mathrm{load}}(l,e)\&ForAll;l\&Element;{\mathrm{\&Gamma;}}_e---\left(32\right)$

Section 1 in formula for shutdown and start-up cost, be set to constant value herein; Section 2 for alternative cost, refer to that genset cut in emergency control its generated energy during stopping transport should be born by emergency power supply, the difference of the cost of electricity-generating caused therefrom, its computing formula is as follows:

$C_t^{\mathrm{repl}}(l,e)=(c_{\mathrm{emerg}}-c_{\mathrm{orig}})\&times;P_G\&times;h---\left(33\right)$

C in formula _emergfor the unit cost of electricity-generating of emergency power supply, c _origfor unit cost of electricity-generating when cut unit normally runs, P _gfor the generating capacity of loss, h is the idle time of cut unit,

To a certain forecast accident, if take to cut machine measure still can not ensure its transient stability, then must adopt cutting load control measure, the load rejection loss caused thus is embodied in the Section 3 of formula (32) in; Load rejection cost computing formula as follows:

$C_t^{\mathrm{load}}(l,e)=\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_{i,t}^{\mathrm{load}}\&times;h_i^{\&prime;}\&times;{\mathrm{IEAR}}_i---\left(34\right)$

In formula for the load of node i excision, h ' is the power off time of cut load;

The emergency control cost of system whole year is

$C^{\mathrm{emer}}=C^{\mathrm{repair}}+C^{\mathrm{repl}}+C^{\mathrm{load}}=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}(C_t^{\mathrm{repair}}+C_t^{\mathrm{repl}}+C_t^{\mathrm{load}})---\left(35\right)$

The social loss cost that 3.3 control measure are invalid

At period t, the invalid social loss cost caused is controlled to a certain forecast accident l for

$C_t^{\mathrm{out}}\left(l\right)=B\&times;P\left(l\right)\&times;h_l\&times;(1+C\left(l\right))---\left(36\right)$

In formula, B is that the average loss of unit loss electricity is worth, and can be obtained by electric system to the statistical data of accident in the past, also can by gross domestic product (GDP) and year power consumption be divided by and obtain; The power failure load general power that P (l) causes for control measure are invalid; Hl is power off time; C (l) is the society that middle reflection system security controls measure is invalid to be caused, the political fallout factor;

According to above analysis, after taking into account load and the probabilistic impact of wind power output, system is calculated as follows at the safety cost of t period of planning level year

$C_t^{\sec }=C_t^{\mathrm{pre}}+\underset{l\&Element;{\mathrm{\&Gamma;}}_e}{\mathrm{\&Sigma;}}\mathrm{Pr}\left(l\right)C_t^{\mathrm{emer}}(l,e_l)+\underset{l\&Element;\mathrm{\&Gamma;}}{\mathrm{\&Sigma;}}\mathrm{Pr}\left(l\right)\mathrm{Pr}\left(\mathrm{DIS}\right|l)C_t^{\mathrm{out}}\left(l\right)---\left(37\right)$

The probability that in formula, Pr (l) occurs for forecast accident l, Pr (DIS|l) is the dynamic unsafe probability of system after generation forecast accident l; Γ is the contingency set of present period, Γ _efor emergency control measure need be taked to guarantee its stable contingency set, and

${\mathrm{\&Gamma;}}_e\&Subset;\mathrm{\&Gamma;};$

System at the safety cost that planning level year is total is

$C_s=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}{C}_t^{\sec }---\left(38\right)$

" active constraint relaxation method " is, first allow prevention and control be responsible for the security of whole forecast accident, at this moment prevention and control cost is maximum, and emergency control cost is zero; Then active constraint part or all of in prevention and control subproblem, the accident namely corresponding to " active constraint " is from prevention and control contingency set Γ _pin " relaxing " out, be incorporated to Γ _ein, re-start the internal layer optimization under new constraint, and observe the change of total cost; If total cost becomes large, then the result that last optimization obtains is exactly optimum solution; If total cost diminishes, then still need the accident corresponding to the part or all of active constraint in this step to be also incorporated to Γ _e, proceed new internal layer optimization; So repeatedly carry out, until total cost no longer reduces.

The optimization of security Comprehensive Control carrys out the security of all forecast accidents of common guarantee, and the total cost simultaneously making security control is minimum, and model is as follows:

$\underset{\&ForAll;{\mathrm{\&Gamma;}}_p,{\mathrm{\&Gamma;}}_e\&Subset;\mathrm{\&Gamma;}}{\min }\left\{\underset{{\mathrm{\&Gamma;}}_p,{\mathrm{\&Gamma;}}_e}{\min }F(x,y,e)\right\}---\left(39\right)$

s.t.f(x)≤z(40)

g(x)＝h(41)

$y\&Element;\left(\underset{\&ForAll;l_i\&Element;{\mathrm{\&Gamma;}}_p}{\&cap;}{\mathrm{\&Omega;}}_d\left(l_i\right)\right)\&cap;\left(\underset{\&ForAll;l_j\&Element;{\mathrm{\&Gamma;}}_e}{\&cap;}{\mathrm{\&Omega;}}_d(l_j,e_j)\right)---\left(42\right)$

Wherein, x, y, e represent the node complex voltage vector as system state variables respectively, node injecting power vector and the emergency control measure taked; And objective function F (x, y, e) represents is then the total cost that security controls, realizing security control total cost by internal layer and outer field dual-layer optimization minimizes; Formula (40) is corresponding to inequality constrain formula (6)-(10) in Preventive Control Model (I), they are used multidimensional function f (x)≤z unified representation, formula (41) is corresponding to equality constraint formula (4)-(5), and formula (42) is the Transient Stability Constraints represented with PDSR; Z, h are the constant column vectors being merged into right side in above-mentioned simplification process; Ω _d(l _i) be accident l _ibasic Practical Dynamic Security Region, Ω _d(l _j, e _j) for considering emergency control measure e _jexpansion Practical Dynamic Security Region, they all represent by lineoid form used in formula (10), node injecting power vector need the common factor space being arranged in above-mentioned two security domains.

Technical characterstic of the present invention and effect:

The present invention devises the security value appraisal procedure of a set of Power System Planning scheme containing large wind-powered electricity generation first, gives the available model of overall flow figure and each composition frame.Example show whole society's Value Comprehensive Assessment result of programme realized on this basis more comprehensively, more science, more reasonable.

In addition, present invention achieves the synergistic application innovation of the following achievement of the first inventor:

(1) the present invention with Unit Combination a few days ago for instrument, by predict planning level year sequential load and wind power output data based on, achieve the comprehensive assessment of power supply, power network planning scheme security value.

(2) the present invention obtains by " active constraint relaxation method " the optimum control scheme and minimum Comprehensive Control cost that prevention and control and emergency control coordinate mutually in the assessment of security value, makes the assessment of security of system more comprehensive, scientifical, reasonable.

(3) up to the present, in electric power system optimization plan model, science takes into account the security of scheme, especially dynamic security, is still a stubborn problem.And the programme security value assessment models that the present invention carries is solve this thorny problem to provide scientific and reasonable solution by the lineoid mathematical description of Dynamic Security Region.

Accompanying drawing explanation

Fig. 1 is security value estimation flow figure.

Fig. 2 is that security Comprehensive Control optimizes calculation flow chart.

Tu3Shi New England system schematic.This system has 39 nodes, is numbered node 1-39 respectively in figure 3, and ten generators of system are G1 to G10, and their position and place node refer to Fig. 3.

Embodiment

For Problems existing in current techniques, the present invention concentrates access for background with large-scale wind power, power source planning and Electric Power Network Planning are considered as a whole, by power system security territory method, the emulation of safety economy Optimized Operation is carried out to the system of by the hour and even more short time interval annual 8760 hours of planning level year, establish power supply, the electrical network unified planning model of having taken into account static system and dynamic security value, achieve wind-powered electricity generation prediction, Unit Combination, security Comprehensive Control, the collaborative innovation of the Multiple techniques such as cost optimization.In the assessment of security value, with Unit Combination a few days ago for instrument, by prediction planning level year sequential load and wind power output data based on, emulated by the safety economy Optimized Operation of the whole year by the period, assess and the fuel used to generate electricity cost that prevention and control and emergency control bring is carried out to contingent forecast accident, Environmental costs and cut machine, cutting load cost, count load and the social loss of wind power output uncertainty when causing control measure invalid simultaneously, thus provide a Systematization method and comparatively perfect estimation flow for the safety evaluation of programme, also assessment into social value is widened from traditional technological economics evaluation in the visual angle of Power System Planning.In prevention and control and emergency control, strong instrument is provided on the impact embedding chain accident in programme safety evaluation that is thought of as of chain accident, also make the assessment of security of system more comprehensive, reasonable.The significance of the method that the present invention puies forward is demonstrated by the assessment of two programmes based on New England's system (in the world widely used model system) figure.

The first step, sets up power supply, the electrical network unified planning model of taking into account static system and dynamic security value, and the cost wherein containing the Power System Planning scheme of large-scale wind power mainly comprises the content of following several aspect: the 1. cost of development C of wind-powered electricity generation _w; 2. the investment of other power supplys and operation expense C except wind-powered electricity generation _g; 3. the investment of power transmission and transforming equipment and operation expense C _n; 4. prevention and control cost C ^pre; 5. emergency control cost C ^emer; 6. the social loss cost C that control measure are invalid ^out; 7. Web-based exercise C ^loss; First can try to achieve 1. according to the given data of power generating facilities and power grids-3..

Second step, starts to carry out detailed working train family to candidate's programme, thus the security value of evaluation scheme.First according to prediction every day day part node load and wind power output carry out Unit Combination a few days ago, with determine load and wind power output fluctuation condition under, whether current power source planning scheme can meet the requirement of system frequency modulation peak regulation and day part margin capacity.

3rd step, after Unit Combination, by the period safety evaluation is carried out to system.In a certain period, first carry out forecast accident scanning and chain accident forecast, filter out the contingency set of the single failure that this period needs are considered and the contingency set of the chain stoppage in transit of circuit may be caused, the common contingency set forming this period.Wherein, the method that single failure contingency set is formed refers to document [9].

4th step, to this contingency set, optimized by " active constraint relaxation method " security Comprehensive Control, obtain optimum prevention and control contingency set and the division of emergency control contingency set, and in the prevention and control optimization and emergency control control and optimize model of each period use safety territory method taking into account of achieving that dynamic security retrains, thus optimum control scheme and corresponding prevention and control cost, the emergency control cost of this period can be obtained and control invalid social loss, namely obtain the total safety cost of system in this period.The difficulty in computation of this step is very large, and it is core content of the present invention.

5th step, by the assessment to annual 8760 hours of planning level year, and in conjunction with the initial outlay of power supply, power network planning scheme, can draw the annual total security cost of programme.In conjunction with the first step, the total cost of programme can be drawn, in this, as the foundation of program evaluation scheme.

Only describe in detail with regard to the content of the 4th step below.

The Power System Planning model of 1 considering security cost

Cost containing the Power System Planning scheme of large-scale wind power mainly comprises the content of following several aspect: the 1. cost of development C of wind-powered electricity generation _w; 2. the investment of other power supplys and operation expense C except wind-powered electricity generation _g; 3. the investment of power transmission and transforming equipment and operation expense C _n; 4. prevention and control cost C ^pre; 5. emergency control cost C ^emer; 6. the social loss cost C that control measure are invalid ^out; 7. Web-based exercise C ^loss.Wherein, the security cost that 6. item is collectively referred to as programme is 4. arrived.Above every cost all refers to annual value cost.

The object of the invention is to find the considering security cost Power System Planning scheme that total cost is minimum later, its mathematical description is as follows:

minC＝C _w+C _g+C _n+C ^pre+C ^emer+C ^out+C ^loss(1)

Core of the present invention is to provide the security cost C calculating programme ^pre+ C ^emer+ C ^outmethod.

2 security value estimation flows

In order to overcome the deficiency of existing method to system cloud gray model shortage accurate analog, the present invention is based on the planning level year load per hour and wind power output data of prediction, with hour for assessment unit, detailed working train family is carried out to candidate's programme, thus the security value of evaluation scheme.Detailed estimation flow as shown in Figure 1.

In FIG, first according to prediction every day day part node load and wind power output carry out Unit Combination a few days ago, with determine load and wind power output fluctuation condition under, whether current power source planning scheme can meet the requirement of system frequency modulation peak regulation and day part margin capacity.It is noted that the method for Unit Combination is the net load curve (load deducts wind power output) first drawing system, more accordingly conventional power unit is combined.Like this, the operating cost increase that wind power output undulatory property causes is embodied among the result of Unit Combination, is finally also embodied among total cost.Note also, Unit Combination herein does not consider that network constraint is interior, and the result therefore combined is the real ruuning situation of representative system present period not, also needs to adopt security constraint generating economic load dispatching instrument to carry out the adjustment of generated output further.

After Unit Combination, by the period safety analysis is carried out to system.In a certain period, first carry out forecast accident scanning and chain accident forecast, filter out the contingency set of the single failure that this period needs are considered and the contingency set of the chain stoppage in transit of circuit may be caused, the common contingency set forming this period.Wherein, the method that single failure contingency set is formed refers to document [9].To this contingency set, optimized by security Comprehensive Control, obtain optimum prevention and control contingency set and the division of emergency control contingency set, thus optimum control scheme and corresponding prevention and control cost, the emergency control cost of this period can be obtained and control invalid social loss, namely obtain the total safety cost of system in this period.By the assessment to annual 8760 hours of planning level year, and in conjunction with the initial outlay of power supply, power network planning scheme, the annual total security cost of programme can be drawn.

3 security cost C _scomputing method

According to the division to forecast accident, the safety cost of system is correspondingly divided into prevention and control cost, social loss cost three part that emergency control cost is invalid with control.The computing method of safety cost will be introduced in detail below.

If grid is by n+1 node, n _bbar circuit forms, and its interior joint 0 is slack bus, with G:={1, and 2 ..., n _grepresent the set of generator node; Use L:={n _g+ 1 ..., n} represents the set of load bus; The set of whole node is represented, i.e. N=G ∪ L ∪ 0 with N; The set of whole circuit is represented with B; CTS is the contingency set of single failure, and it is by preventive control collection Γ _pwith emergency control collection Γ _ecomposition; CTS _cfor the contingency set of the chain stoppage in transit of circuit may be caused.

3.1 prevention and control cost C ^pre

The Part I of security of system cost is prevention and control cost.To some accident that forecast accident is concentrated, the accident that normally probability of happening is larger, the security domain that the Control Measure such as generated output adjustment must be taked to be withdrawn into by the operating point of system corresponding to these forecast accidents is inner, even if some like this accidents really occur, system without the need to take any emergency control measure just can accident occur after a certain operating point stable operation.The prevention and control cost of whole system is made up of three parts:

$C^{\mathrm{pre}}=C_f^{\mathrm{pre}}+C_e^{\mathrm{pre}}+C_c^{\mathrm{pre}}---\left(2\right)$

In formula be respectively prevention and control fuel cost, prevention and control Environmental costs and prevention and control load summate cost.

Because no matter whether accident really occurs all to need to take Control Measure, the probability that when therefore calculating prevention and control cost, accident of getting occurs is 1.

(1) prevention and control fuel cost

If the result of Unit Combination directly meets the various security constraints of system through verifying, then illustrative system is without the need to carrying out extra generation adjustment, and now prevention and control cost is 0; Otherwise, then need to solve the fuel cost that following two optimization problems try to achieve prevention and control.

First problem is the security constrained economic dispatch problem based on AC power flow, and its model is as follows:

$\min F_C=\frac{1}{2}{P}^T{H}_{P}P+f_P^{T}P---\left(3\right)$

$s.t.P_i=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}]\&ForAll;i\&Element;N---\left(4\right)$

$Q_i=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}]\&ForAll;i\&Element;N---\left(5\right)$

$\left(\mathrm{III}\right)P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M\&ForAll;i\&Element;G---\left(6\right)$

$Q_{g,i}^m\&le;Q_{g,i}\&le;Q_{g,i}^M\&ForAll;i\&Element;G---\left(7\right)$

$-P_l^M\&le;P_l\&le;P_l^M\&ForAll;l\&Element;B---\left(8\right)$

$U_i^m\&le;U_i\&le;U_i^M\&ForAll;i\&Element;N---\left(9\right)$

$\underset{\&ForAll;i\&Element;G\&cup;L}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}{P}_i<1\&ForAll;l\&Element;{\mathrm{\&Gamma;}}_p\&Subset;\mathrm{CTS}---\left(10\right)$

Wherein, F _cfor considering Active Generation cost during security constraint, H _p, f _pbe respectively cost of electricity-generating quadratic term matrix of coefficients and Monomial coefficient vector, P is the meritorious injection vector of node, P _i, Q _ibe respectively the meritorious injection of node i and idle injection, G _ij, B _ijbe respectively real part and the imaginary part of bus admittance matrix i-th j element, θ _ijfor branch road ij two ends node voltage phase angle difference; P _{g, i}, Q _{g, i}what be respectively generator i meritoriously exerts oneself and idlely to exert oneself, and its upper and lower limit is respectively p _lfor the effective power flow of branch road l, its maximum permission through-put power is u _ifor the voltage magnitude of node i, its upper and lower limit is respectively α _{(l, i)}for i-th component of the Practical Dynamic Security Region lineoid coefficient for forecast accident l, subscript T represents transposition, wherein Practical Dynamic Security Region and PDSR, PracticalDynamicSecurityRegion.

In above model, formula (6) and formula (7) give generated output constraint, formula (8) and formula (9) sets forth trend and node voltage constraint, and formula (10) is then further used the form of Practical Dynamic Security Region (list of references [7]) to give and ensured that the dynamic security of stability retrains.Here should be noted that constraint equation (10), this formula ensure that solution to model, i.e. system current point of operation, to arbitrary forecast accident system is all dynamic security, thus has embodied the effect of prevention and control.

Second Problem is the Economic Dispatch Problem without security constraint, and its model is as follows:

$\min F_U=\frac{1}{2}{P}^T{H}_{P}P+f_P^{T}P---\left(11\right)$

$\left(\mathrm{IV}\right)s.t.\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_i=0---\left(12\right)$

$P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M\&ForAll;i\&Element;G---\left(13\right)$

Wherein, F _ufor disregarding Active Generation cost during security constraint.

To each period in planning level year, all need to solve above-mentioned security constraint and without security constrained economic dispatch problem.The prevention and control fuel cost of t period of system is security constrained economic dispatch and the difference without security constrained economic dispatch objective function

$C_{f,t}^{\mathrm{pre}}=F_{C,t}-F_{U,t}---\left(14\right)$

In formula be the prevention and control fuel cost of t period, F _{c, t}, F _{u, t}be respectively t time safety constrained optimum trend and the target function value without security constraint optimal load flow.

(2) prevention and control Environmental costs

Second component of prevention and control is prevention and control Environmental costs, is made up of two parts: prevention and control environmental loss cost and prevention and control environmental countermeasure cost.When being located at t, the optimum solution of segment model (I) and model (II) is respectively with then the prevention and control Environmental costs of this period are

$C_{e,t}^{\mathrm{pre}}=[C_{e,t}^{\mathrm{loss}}\left(P_{C,t}^{*}\right)-C_{e,t}^{\mathrm{loss}}\left(P_{U,t}^{*}\right)]+[C_{e,t}^{\mathrm{meas}}\left(P_{C,t}^{*}\right)-C_{e,t}^{\mathrm{meas}}\left(P_{U,t}^{*}\right)]---\left(15\right)$

Part I in formula for prevention and control environmental loss cost, be count security constraint and disregard security constraint environmental loss cost difference; Part II for prevention and control environmental countermeasure cost, be count security constraint and the difference disregarded with generated output relevant portion in security constraint environmental countermeasure cost, refer to second component of environmental countermeasure cost, namely relevant with generated output size environmental countermeasure cost component.Should be noted that; environmental countermeasure cost is former should be made up of two components; one-component is the cost of investment of environment protection equipment; for fixed value; second component is the operation and maintenance cost of equipment; relevant to generated output, due to environment protection equipment once purchase, its cost of investment is in operation irrelevant with Prevention and control strategy.

In formula (15), the environmental loss cost of t period is

$C_{e,t}^{\mathrm{loss}}\left(P_t^{*}\right)=\underset{h=1}{\overset{H}{\mathrm{\&Sigma;}}}{c}_h{Q}_{h,t}\left(P_t^{*}\right)---\left(16\right)$

Wherein, c _hbe the specific emissions cost of h kind pollutant, be the discharge capacity of t period h kind pollutant, h is the pollution species number considered.

T period environmental countermeasure cost relevant with generated output is

$C_{e,t}^{\mathrm{meas}}\left(P_t^{*}\right)=\underset{h=1}{\overset{H}{\mathrm{\&Sigma;}}}{c}_{h,t}^o\left(P_t^{*}\right)---\left(17\right)$

Wherein, for the t period is for the operation and maintenance cost of the environment protection equipment of h kind pollutant.

(3) prevention and control load summate cost

If in a certain period, for certain or some accident that forecast accident is concentrated, even if system is adjusted by generated output also cannot meet static security (Branch Power Flow and node voltage) constraint and dynamic security (as transient stability) constraint, then the Corrective control measures such as load summate must be taked to ensure the security of system.Minimum for target with load summate total amount, system compensation Controlling model is as follows:

$\min \underset{j\&Element;L}{\mathrm{\&Sigma;}}{P}_j^{\mathrm{cut}}---\left(18\right)$

$s.t.P_i^{\&prime;}=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}]\&ForAll;i\&Element;N---\left(19\right)$

$Q_i^{\&prime;}=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}]\&ForAll;i\&Element;N---\left(20\right)$

$P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M\&ForAll;i\&Element;G---\left(21\right)$

$\left(\mathrm{III}\right)Q_{g,i}^m\&le;Q_{g,i}\&le;Q_{g,i}^M\&ForAll;i\&Element;G---\left(22\right)$

$U_i^m\&le;U_i\&le;U_i^M\&ForAll;i\&Element;N---\left(23\right)$

$-P_l^M\&le;P_l\&le;P_l^M\&ForAll;l\&Element;B---\left(24\right)$

$0\&le;P_j^{\mathrm{cut}}\&le;P_{d,j}\&ForAll;j\&Element;L---\left(25\right)$

$0\&le;Q_j^{\mathrm{cut}}\&le;Q_{d,j}\&ForAll;j\&Element;L---\left(26\right)$

$\underset{\&ForAll;i\&Element;G}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}{P}_{g,i}+\underset{\&ForAll;i\&Element;L}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}(P_{d,i}-P_i^{\mathrm{cut}})<1\&ForAll;l\&Element;{\mathrm{\&Gamma;}}_p\&Subset;\mathrm{CTS}---\left(27\right)$

In formula be respectively meritorious, the load or burden without work reduction on a jth load bus, P _{d, j}, Q _{d, j}be respectively a jth node to gain merit and load or burden without work. $P_i^{\&prime;}=P_{g,i}-P_{d,i}+P_i^{\mathrm{cut}},$ $Q_i^{\&prime;}=Q_{g,i}-Q_{d,i}+Q_i^{\mathrm{cut}};$ for the power factor of node i reduction plans, and be constant for node i.Owing to present invention employs the load summate model based on AC power flow, if therefore cut down the burden with power of a certain node, correspondingly load or burden without work generally should be cut down according to the principle that power factor is constant.

The burden with power reduction of t period node i can be obtained by above load summate model then the load summate cost of this period is

$C_{c,t}^{\mathrm{pre}}=\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_{i,t}^{\mathrm{cut}}\&times;{\mathrm{IEAR}}_i---\left(28\right)$

IEAR in formula _ifor the load loss Assessment Rate of node i, namely cut down the loss of outage cost of specific load.

Like this, whole system is made up of prevention and control fuel cost, prevention and control Environmental costs and prevention and control load summate cost three part at the prevention and control cost of period t

$C_t^{\mathrm{pre}}=C_{f,t}^{\mathrm{pre}}+C_{e,t}^{\mathrm{pre}}+C_{c,t}^{\mathrm{pre}}---\left(29\right)$

Corresponding with formula (2), the prevention and control cost of system whole year is

$C^{\mathrm{pre}}=C_f^{\mathrm{pre}}+C_e^{\mathrm{pre}}+C_c^{\mathrm{pre}}=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}(C_{f,t}^{\mathrm{pre}}+C_{e,t}^{\mathrm{pre}}+C_{c,t}^{\mathrm{pre}})---\left(30\right)$

3.2 emergency control cost C ^emer

For some accident that forecast accident is concentrated, the normally less and accident that consequence is comparatively serious of probability of happening, if just taked Control Measure before accident occurs, then can cause sizable prevention and control cost.And in fact for some accident that probability of happening is less, only need by emergency control measure process after accident occurs.Namely the cost of these emergency control measures constitutes the Part II of security cost, is expressed as follows:

C ^emer＝C ^repl+C ^repair+C ^load(31)

C in formula ^repl, C ^repair, C ^loadbe respectively alternative cost, shutdown and start-up cost and load rejection cost.Their concrete meaning and computing method can be introduced hereinafter in detail.Controlling cost of emergency control measure e is taked to comprise shutdown and start-up cost, alternative cost and load rejection cost three part to a certain accident l that forecast accident is concentrated.To t the period in planning level year, emergency control cost can be expressed as

$C_t^{\mathrm{emer}}(l,e)=C_t^{\mathrm{repair}}(l,e)+C_t^{\mathrm{repl}}(l,e)+C_t^{\mathrm{load}}(l,e)\&ForAll;l\&Element;{\mathrm{\&Gamma;}}_e---\left(32\right)$

Section 1 in formula for shutdown and start-up cost, be set to constant value herein.Section 2 for alternative cost, refer to that genset cut in emergency control its generated energy during stopping transport should be born by emergency power supply, the difference of the cost of electricity-generating caused therefrom.Its computing formula is as follows:

$C_t^{\mathrm{repl}}(l,e)=(c_{\mathrm{emerg}}-c_{\mathrm{orig}})\&times;P_G\&times;h---\left(33\right)$

C in formula _emergfor the unit cost of electricity-generating of emergency power supply, c _origfor unit cost of electricity-generating when cut unit normally runs, P _gfor the generating capacity of loss, h is the idle time of cut unit.

To a certain forecast accident, if take to cut machine measure still can not ensure its transient stability, then must adopt cutting load control measure, the load rejection loss caused thus is embodied in the Section 3 of formula (32) in.The computing formula of load rejection cost is as follows:

$C_t^{\mathrm{load}}(l,e)=\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_{i,t}^{\mathrm{load}}\&times;h_i^{\&prime;}\&times;{\mathrm{IEAR}}_i---\left(34\right)$

In formula for the load of node i excision, h ' is the power off time of cut load.

The emergency control cost of system whole year is

$C^{\mathrm{emer}}=C^{\mathrm{repair}}+C^{\mathrm{repl}}+C^{\mathrm{load}}=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}(C_t^{\mathrm{repair}}+C_t^{\mathrm{repl}}+C_t^{\mathrm{load}})---\left(35\right)$

The social loss cost that 3.3 control measure are invalid

When not considering node injecting power uncertain, if the system operating point after prevention and control is positioned at the basic Practical Dynamic Security Region of certain accident, or the system operating point after emergency control is positioned at the expansion Practical Dynamic Security Region [8] of this accident, then these preventions and emergency control measure must be effective, the social loss cost C that now control measure are invalid ^out=0.When taking into account node injecting power uncertain, due to the current point of operation of system accurately cannot be predicted, after taking prevention or emergency control measure, system still has certain dynamic dangerous probability, therefore for the accident that forecast accident is concentrated, control measure are invalid may be impacted society, and this part impact is exactly the invalid social loss costs of control measure.

The impact that power outage causes society is in fact extremely difficult to be weighed, with loss electricity number, interruption duration, power failure load type and significance level etc. have direct relation.Along with the arrival of digital society, an obvious trend is that the loss that unit power off time causes is increasing, therefore in security value assessment, takes into account social loss cost also more and more important.When the uncertain factor of such as wind power output and the load prediction etc. in consideration system, may be invalid to the safety control measures of a certain forecast accident, thus cause system local or large-scale power outage, all trades and professions and resident living are impacted, brings larger social loss.Here need to consider following three aspect factor: the loss of non-energizing quantity, society and political fallout, and load and wind power output uncertainty in traffic.The impact that the present invention adopts the method for factor of influence to count power outage to cause society and politics, economy.At period t, the invalid social loss cost caused is controlled to a certain forecast accident l for

$C_t^{\mathrm{out}}\left(l\right)=B\&times;P\left(l\right)\&times;h_l\&times;(1+C\left(l\right))---\left(36\right)$

In formula, B is that the average loss of unit loss electricity is worth, and can be obtained by electric system to the statistical data of accident in the past, also can by gross domestic product (GDP) and year power consumption be divided by and obtain; The power failure load general power that P (l) causes for control measure are invalid; Hl is power off time; C (l) is the society that middle reflection system security controls measure is invalid to be caused, the political fallout factor.

According to above analysis, after taking into account load and the probabilistic impact of wind power output, system is calculated as follows at the safety cost of t period of planning level year

$C_t^{\sec }=C_t^{\mathrm{pre}}+\underset{l\&Element;{\mathrm{\&Gamma;}}_e}{\mathrm{\&Sigma;}}\mathrm{Pr}\left(l\right)C_t^{\mathrm{emer}}(l,e_l)+\underset{l\&Element;\mathrm{\&Gamma;}}{\mathrm{\&Sigma;}}\mathrm{Pr}\left(l\right)\mathrm{Pr}\left(\mathrm{DIS}\right|l)C_t^{\mathrm{out}}\left(l\right)---\left(37\right)$

The probability that in formula, Pr (l) occurs for forecast accident l, Pr (DIS|l) is the dynamic unsafe probability of system after generation forecast accident l.Γ is the contingency set of present period, Γ _efor emergency control measure need be taked to guarantee its stable contingency set, and

${\mathrm{\&Gamma;}}_e\&Subset;\mathrm{\&Gamma;}.$

System at the safety cost that planning level year is total is

$C_s=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}{C}_t^{\sec }---\left(38\right)$

The determination of 4 security Comprehensive Control optimal cases

If with the forecast accident that prevention and control process is all, often making the cost of prevention and control very large, even occurring the situation in order to prevent the generation of certain accident to carry out cutting load; If all forecast accidents all transfer to emergency control process, the cost of emergency control can be made again to become very large, be sometimes even difficult to the security ensureing system cloud gray model.Best control program should be prevention and control and the cooperatively interacting of emergency control, and contingency set is divided into two parts, adopts prevention and control for a part of accident, another part accident employing emergency control.Here, the method implementing prevention and control describes in 3.1 joints, and the implementation method of emergency control refers to document [8].The target of optimum security Comprehensive Control is exactly carry out the security of all forecast accidents of common guarantee by prevention and control and emergency control, and the total cost simultaneously making security control is minimum.Its model is as follows:

$\underset{\&ForAll;{\mathrm{\&Gamma;}}_p,{\mathrm{\&Gamma;}}_e\&Subset;\mathrm{\&Gamma;}}{\min }\left\{\underset{{\mathrm{\&Gamma;}}_p,{\mathrm{\&Gamma;}}_e}{\min }F(x,y,e)\right\}---\left(39\right)$

s.t.f(x)≤z(40)

g(x)＝h(41)

$y\&Element;\left(\underset{\&ForAll;l_i\&Element;{\mathrm{\&Gamma;}}_p}{\&cap;}{\mathrm{\&Omega;}}_d\left(l_i\right)\right)\&cap;\left(\underset{\&ForAll;l_j\&Element;{\mathrm{\&Gamma;}}_e}{\&cap;}{\mathrm{\&Omega;}}_d(l_j,e_j)\right)---\left(42\right)$

Wherein, x, y, e represent the node complex voltage vector as system state variables respectively, node injecting power vector and the emergency control measure taked; And objective function F (x, y, e) represents is then the total cost that security controls, realizing security control total cost by internal layer and outer field dual-layer optimization minimizes; Formula (40) is corresponding to inequality constrain formula (6)-(10) in Preventive Control Model (I), and by them, with multidimensional function f (x)≤z unified representation, (such as formula (6) can be expressed as with formula (41) is corresponding to equality constraint formula (4)-(5), and formula (42) is the Transient Stability Constraints represented with PDSR; Z, h are the constant column vectors being merged into right side in above-mentioned simplification process; Ω _d(l _i) be accident l _ibasic Practical Dynamic Security Region, Ω _d(l _j, e _j) for considering emergency control measure e _jexpansion Practical Dynamic Security Region, they all represent by lineoid form used in formula (10), node injecting power vector need the common factor space being arranged in above-mentioned two security domains.

This model is one the non-linear dual-layer optimization problem of the mixing of dynamic constrained.Outer optimization attempts contingency set to carry out optimal dividing, obtains optimum prevention and control accident set Γ _pwith emergency control accident set Γ _e, thus provide fixing constraint set for internal layer optimization; The Γ that internal layer optimization provides for skin optimization _pand Γ _e, select prevention and control and the emergency control scheme of overhead control cost minimization.

The present invention adopts " active constraint relaxation method " [9] security Comprehensive Control scheme to each period to be optimized.First the method allows prevention and control be responsible for the security of whole forecast accident, and at this moment prevention and control cost is maximum, and emergency control cost is zero.Then the accident in prevention and control subproblem corresponding to part (or all) active constraint (i.e. " active constraint ") from Γ _pin " relaxing " out, be incorporated to Γ _ein, re-start the internal layer optimization under new constraint, and observe the change of total cost.If total cost becomes large, then the result that last optimization obtains is exactly optimum solution; If total cost diminishes, then still need the accident corresponding to part (or whole) active constraint in this step to be also incorporated to Γ _e, proceed new internal layer optimization.So repeatedly carry out, until total cost no longer reduces.Detailed calculation process as shown in Figure 2.

5 sample calculation analysis

Adopt the Power System Planning solution security Valuation Method that New England's system (Fig. 3) proposes as Example Verification the present invention.This system has 39 nodes, 46 branch roads.At planning standard year, system has 10 generators, and the type of each unit, place node and generating capacity are as shown in table 1.

Table 1 New England system standard year power configuration situation

* note: bracket inner digital represents unit quantity.

The peak load of system standard year is 5683.1MW, and estimating, in planning level year, will increase 400MW load, and for details, see the appendix for the load value of each node.

In order to meet the workload demand of system in planning level year, planning and generating capacity is increased 800MW.Now two kinds of programmes are proposed:

Scheme one---without wind-powered electricity generation scheme

In programme one, extend two the 600MW coal unit of original system at 35,37 nodes for 1000MW extra-supercritical unit, power configuration situation is as shown in table 2.Because each bar transmission line of electricity all meets transmission capacity limits requirement when normally running, therefore without the need to newly-built circuit.

Table 2 scheme one power configuration situation

Scheme two---wind-powered electricity generation scheme

In programme two, newly-built two Large Scale Wind Farm Integrations of system, be connected to node 2 and node 11 respectively, its power configuration situation is as shown in table 3.Due in the newly-built 600MW wind energy turbine set of node 2, sending demand to meet wind-powered electricity generation, being double back by circuit 25-26 by single time enlarging, being shown in Fig. 3.For details, see the appendix for the investment data of wind energy turbine set and transmission line of electricity.

Table 3 scheme two power configuration situation

For ease of comparing, the present invention all adopts same leading contingency set to two programmes, wherein altogether containing 25 faults (N-1 fault 18, N-2 fault 6, N-3 fault 1).The method that each accident that forecast accident is concentrated all adopts document [10] to propose calculates PDSR lineoid coefficient.From chain accident forecast, may cause the chain stoppage in transit of All other routes after circuit 1-39,10-13 break down, the circuit that next stage is stopped transport is respectively circuit 2-3 and 6-11.

Calculating for simplifying, making the following assumptions:

(1) namely the contingency set of annual all period employings is all identical does not consider the time dependent situation of forecast accident.

(2) the wind power prediction method adopting document [11] to propose, obtains the sequential output of wind electric field data of annual 8760 hours, and supposes all output of wind electric field all according to the curvilinear motion of same sequential output of wind electric field.

(3) when supposing emergency control generator cut after unused time be 10 hours, after shutdown, vacancy generated energy is supplemented by equilibrator; On each node, the shutdown of generator and start-up cost are set to different constants according to generator capacity.

(4) when hypothetical accident causes system unstability, total system loses a halfload, and power off time is taken as 1 hour, and the average value B=4250 unit/MWh of unit loss electricity, the social influence factor is taken as 4.56.

(5) node load predicated error is taken as 5%, and wind power prediction error is taken as 20%.

(6) five kinds of pollutants such as carbon dioxide, carbon monoxide, sulphuric dioxide, oxides of nitrogen, total suspended particles are considered in the calculating of discharge costs.

Adopt Matlab programming language to realize previously described security value appraisal procedure, table 4 gives the result of calculation of two programmes being carried out to security value assessment, and the cost in table is the cost after the year value such as being converted to.

Can see from table 4, the power construction cost difference of two programmes is not very large, about differs 100,000,000 yuan, if the manufacturing cost of wind power equipment continues to decline, the Wind Power Generation investment of programme two will have the space declined further.For safety cost, the prevention and control cost of scheme one is more than scheme two, and this is mainly fired power generating unit due to the power supply in scheme one, and the fuel that its prevention and control increase and Environmental costs all comparatively scheme two are large.It should be noted that emergency control cost, 0.02 hundred million yuan is all only had for two programmes, also less than 0.5% in total safety cost.This mainly ensures safety due to forecast accident most of in assessment by prevention and control, and the accident that emergency control is responsible for is relatively less, is even entirely ensured by prevention and control at the forecast accident that a lot of period is all; Meanwhile, the probability that accident occurs will also be counted in the calculating of emergency control cost.Therefore, from the angle of engineering reality, if system leaves enough margins of safety, when needing the forecast accident of consideration fewer in scheme evaluation, then can be similar to and ignore emergency control cost.For the social loss cost that control measure are invalid, scheme two whole year is scheme more than 1 2.8 hundred million yuan comparatively.Obviously, because wind power output has larger uncertainty, result in scheme two planning level year horal dangerous probability all large than scheme one, thus result in its larger social loss cost.If can improve wind speed and wind power prediction precision, the social loss cost of scheme two will reduce.Scheme two total safety cost 2.49 hundred million yuan than scheme more than, and total cost has more 1.348 hundred million yuan than scheme one.

Table 4 security value assessment result

Although from the angle of power system security, the total cost of scheme one is little compared with scheme two, seems to have more advantage, and security is only the aspect weighing programme, therefore can not show that building thermoelectricity is just better than the conclusion of building wind-powered electricity generation from table 4.If contrasted from the angle of power supply, the value assessment of the power network planning scheme whole society, table 5 gives another kind of different result.

Table 5 whole society value assessment result

Unit: hundred million yuan	Programme one	Programme two
			Wind Power Generation cost	0	6.41
The investment of other power supplys and operation expense	7.42	0
			The investment of power transmission and transforming equipment and operation expense	0	0.158
Web-based exercise	1.75	1.46
			Fuel cost	79.0	76.1
Environmental costs	10.3	10.1
			Safety cost	6.58	9.07
Total social cost	105.05	103.298

In table 5, fuel cost refers to that all thermoelectricitys, Gas Generator Set are at fuel (coal, the rock gas) cost that the consumes whole year in planning level year, and Environmental costs refer to the pollutants such as CO2, NOX that unit discharges be converted to economic loss after cost.Because wind-power electricity generation does not need cost can not produce various pollutant yet, therefore the fuel cost of scheme two and Environmental costs all little than scheme one, and the ratio of this part in total cost is very large, and therefore total social cost of scheme two is less than scheme one.

List of references

[7] Yu Yixin. the technique study commentary of power system security territory. University Of Tianjin's journal, 2008,41 (6): 635-646.

[8] Liu Hui, Yu Yixin. based on the power system security Comprehensive Control of Practical Dynamic Security Region. Proceedings of the CSEE, 2005,25 (20): 31-36

[9] Yu Yixin, Wang Dongtao. the risk assessment of transmission system dynamic security and optimization. Chinese science (E collects: technological sciences), 2009,39 (2): 286-292.

[10] Zeng Yuan, Yu Yixin. the pragmatic solution of Dynamic Security Regions of Power Systems. Proceedings of the CSEE, 2003,23 (5): 24-28

[11]ZhangNing，KangChongqingCQ，DuanChanggangCG，etal.Simulationmethodologyofmultiplewindfarmsoperationconsideringwindspeedcorrelation.TheThirdIASTEDAsianConferenceonPowerandEnergySystems，Beijing，2009.

Claims (2)

1., containing a security value appraisal procedure for the Power System Planning scheme of large-scale wind power, comprise the following steps:

The first step, sets up power supply, the electrical network unified planning model of taking into account static system and dynamic security value, and the cost wherein containing the Power System Planning scheme of large-scale wind power mainly comprises the content of following several aspect: the 1. cost of development C of wind-powered electricity generation _w; 2. the investment of other power supplys and operation expense C except wind-powered electricity generation _g; 3. the investment of power transmission and transforming equipment and operation expense C _n; 4. prevention and control cost C ^pre; 5. emergency control cost C ^emer; 6. invalid social loss C is controlled ^out; 7. Web-based exercise C ^loss; First can try to achieve 1. according to the given data of power generating facilities and power grids-3.;

Second step, start to carry out detailed working train family to candidate's programme, thus the security value of evaluation scheme: first according to prediction every day day part node load and wind power output carry out Unit Combination a few days ago, with determine load and wind power output fluctuation condition under, whether current power source planning scheme can meet the requirement of system frequency modulation peak regulation and day part margin capacity;

3rd step, after Unit Combination, by the period safety evaluation is carried out to system: in a certain period, first forecast accident scanning and chain accident forecast is carried out, filter out the contingency set of the single failure that this period needs are considered and the contingency set of the chain stoppage in transit of circuit may be caused, the common contingency set forming this period;

4th step, to this contingency set, the optimization of security Comprehensive Control is carried out by " active constraint relaxation method ", obtain optimum prevention and control contingency set and the division of emergency control contingency set, and in the prevention and control optimization and emergency control Optimized model of each period use safety territory method taking into account of achieving that dynamic security retrains, thus optimum control scheme and corresponding prevention and control cost, the emergency control cost of this period can be obtained and control invalid social loss, namely obtain the total safety cost of system in this period;

5th step, by the assessment to annual 8760 hours of planning level year, and in conjunction with the initial outlay of power supply, power network planning scheme, namely the annual total security cost of programme is drawn, in conjunction with the first step, the total cost of programme can be drawn, in this, as the foundation of program evaluation scheme;

The security method of cost accounting is refined as:

If grid is by n+1 node, n _bbar circuit forms, and its interior joint 0 is slack bus, with G:={1, and 2 ..., n _grepresent the set of generator node; Use L:={n _g+ 1 ..., n} represents the set of load bus; The set of whole node is represented, i.e. N=G ∪ L ∪ 0 with N; The set of whole circuit is represented with B; CTS is the contingency set of single failure, and CTS is by contingency set and the prevention and control contingency set Γ that need take Control Measure _pwith emergency control measure need be taked to guarantee contingency set and the emergency control contingency set Γ of stability _ecomposition; CTS _cfor the contingency set of the chain stoppage in transit of circuit may be caused: 1.1 prevention and control cost C ^pre

Prevention and control cost is made up of three parts:

$C^{\mathrm{pre}}=C_f^{\mathrm{pre}}+C_e^{\mathrm{pre}}+C_c^{\mathrm{pre}}---\left(2\right)$

In formula be respectively prevention and control fuel cost, prevention and control Environmental costs and prevention and control load summate cost;

The probability that accident of getting when calculating prevention and control cost occurs is 1;

(1) prevention and control fuel cost

If the result of Unit Combination directly meets the various security constraints of system through verifying, then illustrative system is without the need to carrying out extra generation adjustment, and now prevention and control cost is 0; Otherwise, then need to solve the fuel cost that following two optimization problems try to achieve prevention and control:

First problem is the security constrained economic dispatch problem based on AC power flow, and its model is as follows:

$\min F_C=\frac{1}{2}{P}^T{H}_{P}P+f_P^{T}P---\left(3\right)$

$\begin{array}{ccc}s.t.& P_i=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}]& \&ForAll;i\&Element;N---\left(4\right)\end{array}$

$\begin{array}{cc}{Q}_i=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}]& \&ForAll;i\&Element;N---\left(5\right)\end{array}$

$\begin{array}{ccc}\left(I\right)& P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M& \&ForAll;i\&Element;G---\left(6\right)\end{array}$

$\begin{array}{cc}{Q}_{g,i}^m\&le;Q_{g,i}\&le;Q_{g,i}^M& \&ForAll;i\&Element;G\end{array}---\left(7\right)$

$\begin{array}{cc}-P_l^M\&le;P_l\&le;P_l^M& \&ForAll;l\&Element;B---\left(8\right)\end{array}$

$\begin{array}{cc}{U}_i^m\&le;U_i\&le;U_i^M& \&ForAll;i\&Element;N\end{array}---\left(9\right)$

$\begin{array}{cc}\underset{\&ForAll;i\&Element;G\&cup;L}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(e,i)}{P}_i<1& \&ForAll;l\&Element;{\mathrm{\&Gamma;}}_p\&Subset;\mathrm{CTS}---\left(10\right)\end{array}$

Wherein, F _cfor considering Active Generation cost during security constraint, H _p, f _pbe respectively cost of electricity-generating quadratic term matrix of coefficients and Monomial coefficient vector, P is the meritorious injection vector of node, P _i, Q _ibe respectively the meritorious injection of node i and idle injection, G _ij, B _ijbe respectively real part and the imaginary part of bus admittance matrix i-th j element, θ _ijfor branch road ij two ends node voltage phase angle difference; P _g,i, Q _g,iwhat be respectively generator i meritoriously exerts oneself and idlely to exert oneself, and its upper and lower limit is respectively p _lfor the effective power flow of branch road l, its maximum permission through-put power is u _ifor the voltage magnitude of node i, its upper and lower limit is respectively α _{(l, i)}for i-th component of the Practical Dynamic Security Region lineoid coefficient for forecast accident l, subscript T represents transposition, wherein Practical Dynamic Security Region and PDSR, PracticalDynamicSecurityRegion;

Second Problem is the Economic Dispatch Problem without security constraint, and its model is as follows:

$\min F_U=\frac{1}{2}{P}^T{H}_{P}P+f_P^{T}P---\left(11\right)$

$\begin{array}{ccc}\left(\mathrm{II}\right)& s.t.& \underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_i=0---\left(12\right)\end{array}$

$\begin{array}{cc}{P}_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M& \&ForAll;i\&Element;G\end{array}---\left(13\right)$

Wherein, F _ufor disregarding Active Generation cost during security constraint;

The prevention and control fuel cost of t period of system is security constrained economic dispatch and the difference without security constrained economic dispatch objective function

$C_{f,t}^{\mathrm{pre}}=F_{C,t}-F_{U,t}---\left(14\right)$

In formula be the prevention and control fuel cost of t period, F _c,t, F _u,tbe respectively t time safety constrained optimum trend and the target function value without security constraint optimal load flow;

(2) prevention and control Environmental costs

Second component of prevention and control is prevention and control Environmental costs, is made up of two parts: prevention and control environmental loss cost and prevention and control environmental countermeasure cost, is located at the meritorious vector that injects of node corresponding to the optimum solution of segment model I and model II during t and is respectively with the then prevention and control Environmental costs of t period for

$C_{e,t}^{\mathrm{pre}}=[C_{e,t}^{\mathrm{loss}}\left(P_{C,t}^{*}\right)-C_{e,t}^{\mathrm{loss}}\left(P_{U,t}^{*}\right)]+[C_{e,t}^{\mathrm{meas}}\left(P_{C,t}^{*}\right)-C_{e,t}^{\mathrm{meas}}\left(P_{U,t}^{*}\right)]---\left(15\right)$

Part I in formula for prevention and control environmental loss cost, be count security constraint and disregard security constraint environmental loss cost difference; Part II for prevention and control environmental countermeasure cost, be count security constraint and the difference disregarded with generated output relevant portion in security constraint environmental countermeasure cost, refer to second component of environmental countermeasure cost, namely relevant with generated output size environmental countermeasure cost component; Should be noted that, environmental countermeasure cost is former should be made up of two components, one-component is the cost of investment of environment protection equipment, for fixed value, second component is the operation and maintenance cost of equipment, relevant to generated output, due to environment protection equipment once purchase, its cost of investment is in operation irrelevant with Prevention and control strategy;

In formula (15), the environmental loss cost of t period for

$C_{e,t}^{\mathrm{loss}}\left(P_t^{*}\right)=\underset{h=1}{\overset{H}{\mathrm{\&Sigma;}}}{c}_h{Q}_{h,t}\left(P_t^{*}\right)---\left(16\right)$

Wherein, c _hbe the specific emissions cost of h kind pollutant, be the discharge capacity of t period h kind pollutant, or the optimum solution of the corresponding segment model I and model II when t respectively, H is the pollution species number considered;

The environmental countermeasure cost that t period is relevant with generated output for

$C_{e,t}^{\mathrm{meas}}\left(P_t^{*}\right)=\underset{h=1}{\overset{H}{\mathrm{\&Sigma;}}}{c}_{h,t}^o\left(P_t^{*}\right)---\left(17\right)$

Wherein, for the t period is for the operation and maintenance cost of the environment protection equipment of h kind pollutant;

(3) prevention and control load summate cost

As follows with the minimum system compensation Controlling model for target of load summate total amount:

$\min \underset{j\&Element;L}{\mathrm{\&Sigma;}}{P}_j^{\mathrm{cut}}---\left(18\right)$

$\begin{array}{ccc}s.t.& P_i^{\&prime;}=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}]& \&ForAll;i\&Element;N---\left(19\right)\end{array}$

$\begin{array}{cc}{Q}_i^{\&prime;}=U_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j[G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}]& \&ForAll;i\&Element;N\end{array}---\left(20\right)$

$\begin{array}{ccc}& P_{g,i}^m\&le;P_{g,i}\&le;P_{g,i}^M& \&ForAll;i\&Element;G---\left(21\right)\end{array}$

$\begin{array}{ccc}\left(\mathrm{III}\right)& Q_{g,i}^m\&le;Q_{g,i}\&le;Q_{g,i}^M& \&ForAll;i\&Element;G---\left(26\right)\end{array}$

$\begin{array}{cc}{U}_i^m\&le;U_i\&le;U_i^M& \&ForAll;i\&Element;N\end{array}---\left(23\right)$

$\begin{array}{cc}-P_l^M\&le;P_l\&le;P_l^M& \&ForAll;l\&Element;B---\left(24\right)\end{array}$

$\begin{array}{cc}0\&le;P_j^{\mathrm{cut}}\&le;P_{d,j}& \&ForAll;j\&Element;L---\left(25\right)\end{array}$

$\begin{array}{cc}0\&le;Q_j^{\mathrm{cut}}\&le;Q_{d,j}& \&ForAll;j\&Element;L---\left(26\right)\end{array}$

$\begin{array}{cc}\underset{\&ForAll;i\&Element;G}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}{P}_{g,i}+\underset{\&ForAll;i\&Element;L}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{(l,i)}(P_{d,i}-P_i^{\mathrm{cut}})<1& \&ForAll;l\&Element;{\mathrm{\&Gamma;}}_p\&Subset;\mathrm{CTS}---\left(27\right)\end{array}$

In formula be respectively meritorious, the load or burden without work reduction on a jth load bus, P _d,j, Q _d,jbe respectively a jth node to gain merit and load or burden without work; α _{(l, i)}for the Practical Dynamic Security Region for forecast accident l, i.e. i-th component of PDSR, PracticalDynamicSecurityRegion lineoid coefficient; for the power factor of node i reduction plans, and be constant for node i, the burden with power reduction of t period node i can be obtained by above load summate model then the load summate cost of this period is

$C_{c,t}^{\mathrm{pre}}=\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_{i,t}^{\mathrm{cut}}\&times;{\mathrm{IEAR}}_i---\left(28\right)$

IEAR in formula _ifor the load loss Assessment Rate of node i, namely cut down the loss of outage cost of specific load;

Like this, whole system is made up of prevention and control fuel cost, prevention and control Environmental costs and prevention and control load summate cost three part at the prevention and control cost of period t

$C_t^{\mathrm{pre}}=C_{f,t}^{\mathrm{pre}}+C_{e,t}^{\mathrm{pre}}+C_{c,t}^{\mathrm{pre}}---\left(29\right)$

Corresponding with formula (2), the prevention and control cost of system whole year is

$C^{\mathrm{pre}}=C_f^{\mathrm{pre}}+C_e^{\mathrm{pre}}+C_c^{\mathrm{pre}}=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}(C_{f,t}^{\mathrm{pre}}+C_{e,t}^{\mathrm{pre}}+C_{c,t}^{\mathrm{pre}})---\left(30\right)$ 1.2 emergency control cost C ^emer

Namely the cost of emergency control measure constitutes the Part II of security cost, is expressed as follows:

C ^emer＝C ^repl+C ^repair+C ^load(31)

C in formula ^repl, C ^repair, C ^loadbe respectively alternative cost, shutdown and start-up cost and load rejection cost, to a certain accident l that forecast accident is concentrated, controlling cost of emergency control measure e is taked to comprise shutdown and start-up cost, alternative cost and load rejection cost three part, to t the period in planning level year, emergency control cost can be expressed as

$\begin{array}{cc}{C}_t^{\mathrm{emer}}(l,e)=C_t^{\mathrm{repair}}(l,e)+C_t^{\mathrm{repl}}(l,e)+C_t^{\mathrm{load}}(l,e)& \&ForAll;l\&Element;{\mathrm{\&Gamma;}}_e---\left(32\right)\end{array}$

Section 1 in formula for shutdown and start-up cost, be set to constant value herein; Section 2 for alternative cost, refer to that genset cut in emergency control its generated energy during stopping transport should be born by emergency power supply, the difference of the cost of electricity-generating caused therefrom, its computing formula is as follows:

$C_t^{\mathrm{repl}}(l,e)=(c_{\mathrm{emerg}}-c_{\mathrm{orig}})\&times;P_G\&times;h---\left(33\right)$

C in formula _emergfor the unit cost of electricity-generating of emergency power supply, c _origfor unit cost of electricity-generating when cut unit normally runs, P _gfor the generating capacity of loss, h is the idle time of cut unit,

To a certain forecast accident, if take to cut machine measure still can not ensure its transient stability, then must adopt cutting load control measure, the load rejection loss caused thus is embodied in the Section 3 of formula (32) in; Load rejection cost computing formula as follows:

$C_t^{\mathrm{load}}(l,e)=\underset{i\&Element;N}{\mathrm{\&Sigma;}}{P}_{i,t}^{\mathrm{load}}\&times;h_i^{\&prime;}\&times;{\mathrm{IEAR}}_i---\left(34\right)$

In formula for the load of node i excision, h ' is the power off time of cut load;

The emergency control cost of system whole year is

$C^{\mathrm{emer}}=C^{\mathrm{repair}}+C^{\mathrm{repl}}+C^{\mathrm{load}}=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}(C_t^{\mathrm{repair}}+C_t^{\mathrm{repl}}+C_t^{\mathrm{load}})---\left(35\right)$

1.3 control invalid social loss

At period t, the invalid social loss cost caused is controlled to a certain forecast accident l for

$C_t^{\mathrm{out}}\left(l\right)=B\&times;P\left(l\right)\&times;h_l\&times;(1+C\left(l\right))---\left(36\right)$

In formula, B is that the average loss of unit loss electricity is worth, and can be obtained by electric system to the statistical data of accident in the past, also can by gross domestic product (GDP) and year power consumption be divided by and obtain; The power failure load general power that P (l) causes for control measure are invalid; h _lfor power off time; C (l) is the society that middle reflection system security controls measure is invalid to be caused, the political fallout factor;

According to above analysis, after taking into account load and the probabilistic impact of wind power output, system is calculated as follows at the safety cost of t period of planning level year

$C_t^{\sec }=C_t^{\mathrm{pre}}+\underset{{l\&Element;\mathrm{\&Gamma;}}_e}{\mathrm{\&Sigma;}}\mathrm{Pr}\left(l\right)C_t^{\mathrm{emer}}(l,e_l)+\underset{l\&Element;\mathrm{\&Gamma;}}{\mathrm{\&Sigma;}}\mathrm{Pr}\left(l\right)\mathrm{Pr}\left(\mathrm{DIS}\right|l)C_t^{\mathrm{out}}\left(l\right)---\left(37\right)$

The probability that in formula, Pr (l) occurs for forecast accident l, Pr (DIS|l) is the dynamic unsafe probability of system after generation forecast accident l; Γ is the contingency set of present period, Γ _efor emergency control measure need be taked to guarantee its stable contingency set, and ${\mathrm{\&Gamma;}}_e\&Subset;\mathrm{\&Gamma;};$

System at the safety cost that planning level year is total is

$C_S=\underset{t=1}{\overset{8760}{\mathrm{\&Sigma;}}}{C}_t^{\sec }---\left(38\right)$

Wherein, " active constraint relaxation method " is, first allow prevention and control be responsible for the security of whole forecast accident, at this moment prevention and control cost is maximum, and emergency control cost is zero; Then active constraint part or all of in prevention and control subproblem, the accident namely corresponding to " active constraint " is from prevention and control contingency set Γ _pin " relaxing " out, be incorporated to Γ _ein, re-start the internal layer optimization under new constraint, and observe the change of total cost; If total cost becomes large, then the result that last optimization obtains is exactly optimum solution; If total cost diminishes, then still need the accident corresponding to the part or all of active constraint in this step to be also incorporated to Γ _e, proceed new internal layer optimization; So repeatedly carry out, until total cost no longer reduces.

2. the security value appraisal procedure of the Power System Planning scheme containing large-scale wind power as claimed in claim 1, the 4th step is further refined as:

(1) the Power System Planning model of considering security cost: the Power System Planning scheme that after considering security cost C, total cost is minimum, its mathematical description is as follows:

minC＝C _w+C _g+C _n+C ^pre+C ^emer+C ^out+C ^loss(1)

(2) security value estimation flow:

First according to prediction every day day part node load and wind power output carry out Unit Combination a few days ago, with determine load and wind power output fluctuation condition under, whether current power source planning scheme can meet the requirement of system frequency modulation peak regulation and day part margin capacity, the method of Unit Combination is the net load curve first drawing system, namely load deducts wind power output, then combines conventional power unit accordingly; Unit Combination herein does not consider that network constraint is interior;

To described contingency set, optimized by security Comprehensive Control, obtain optimum prevention and control contingency set and the division of emergency control contingency set, thus optimum control scheme and corresponding prevention and control cost, the emergency control cost of this period can be obtained and control invalid social loss, namely obtain the total safety cost of system in this period.