CN104600695A - Trend load flow calculating method based on online status estimation and real-time scheduling plans - Google Patents

Abstract

The invention provides a trend load flow calculating method based on online status estimation and real-time scheduling plans. The method comprises the following steps of determining future trend operating mode data of a power grid; performing rationality identification and automatic adjustment on real-time scheduling plans; establishing a multi-section active power control model to perform automatic fine-tuning generation active power control; performing reactive power voltage local control, and distributing reactive power unbalance of hub nodes; performing load flow calculation to generate load flow calculation data used for power grid future trend dynamic safety assessment. The trend load flow calculating method based on online status estimation and real-time scheduling plans solves the problem that traditional online dynamic safety assessment algorithms cannot effectively analysis future power grid short safety and stability within a short time, and by analyzing further power grid safety change trend in advance, assists scheduling personnel to arrange and adjust current operating manners.

Description

Estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence

Technical field

The invention belongs to powernet simulation analysis field, be specifically related to a kind ofly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence.

Background technology

Along with the progressively formation of China's extra-high voltage alternating current-direct current series-parallel connection bulk power grid general layout, power system safety and stability characteristic and mechanism increasingly sophisticated, operation of power networks controls difficulty and continues to increase, the accuracy that in-circuit emulation is analyzed and advancedly proposed to new requirement.The basis of carrying out powernet safety analysis in advance forms the trend flow data characterizing following power system operating mode.The key factor forming following power system operating mode comprises:

1) plan in real time, the changes of operating modes such as multiple labour, power trade adjustment, temporary scheduling operation and electric network fault are stopped according to load variations situation, equipment, consider that systematic economy runs and security constraint, each generating start-stop of generator set plan and the plan of exerting oneself in the given short time, interregional active power exchange plan;

2) ultra-short term bus load prediction, as the important foundation of electrical network Real-time generation control, on the factor bases such as historical load data and Changes in weather, each bus burden with power variation tendency in the forecast short time;

3) presence data estimator, describes equipment static parameter under the current operation of electrical network, topological structure, generator measures and the operating mode such as controling parameters, load active power factor, load tap changer position.

The convergence of trend trend and precision determine accuracy and the reasonability of following power system operating mode safety analysis result to a great extent.At present, the existing research of the AC power flow computational methods based on operation plan, effective power flow precision improves constantly, and substantially meets the accuracy requirement of the out-of-limit check of effective power flow.But owing to lacking the idle plan of generating set and bus reactive load forecasting, the idle plan trend reasonability in power flow solutions is poor.On the one hand, the reactive power flow that deviation is larger can affect the precision of the out-of-limit check of circuit rated current and the out-of-limit check result of main transformer rated capacity; On the other hand, irrational reactive voltage distribution also can affect the computational accuracy of effective power flow to a certain extent, and has a strong impact on the reasonability of voltage out-of-limit check and stability check result.Simultaneously, the uncertain factors such as ultra-short term error, new forms of energy fluctuating error, electric energy Plan rescheduling and operation of power networks operation, can have a negative impact to the robustness of trend Load flow calculation, computational convergence also can be caused to affect serious problem by the quality of data.

Summary of the invention

In order to overcome above-mentioned the deficiencies in the prior art, the invention provides and a kind ofly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, cannot effectively to the problem that safety and stability in the following electrical network short time is analyzed for solving traditional online dynamic secure estimation algorithm, analyze in advance following power grid security variation tendency, auxiliary dispatching personnel carry out arrangement adjustment to current operational mode.

In order to realize foregoing invention object, the present invention takes following technical scheme:

The invention provides and a kind ofly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, said method comprising the steps of:

Step 1: determine electrical network future trend running mode data;

Step 2: reasonability identification is carried out to Real-Time Scheduling plan, and it is adjusted automatically;

Step 3: set up multibreak real power control model, carries out automatic fine tuning generating active power controller;

Step 4: carry out reactive voltage and control on the spot, and distribute hub node idle amount of unbalance;

Step 5: carry out Load flow calculation, generates the Load flow calculation data being used for electrical network future trend dynamic secure estimation.

In described step 1, electrical network future trend running mode data is included in line states data estimator, Real-Time Scheduling planning data and ultra-short term data.

According to the online grid measurement data obtained, adopt statistical method to estimate dynamic power system internal state, obtain presence data estimator;

On the basis of operation plan a few days ago, in conjunction with ultra-short term information, ad hoc inspection and repair information and transregional electricity transaction application information, to formulate in the following 5min of electrical network or in 60min the corresponding period real-time generation schedule and get in touch with trading program in real time, obtain Real-Time Scheduling planning data;

Utilize existing history daily load data and meteorological data, the load value of corresponding period in the following 5min of electrical network or in 60min is estimated, complete the ultra-short term comprising system loading prediction and bus load prediction, obtain ultra-short term data.

Described step 2 specifically comprises the following steps:

Step 2-1: carry out reasonability identification to Real-Time Scheduling plan, comprises plan Time-Series analysis, plan association analysis and the analysis of achieve an equilibrium ikn planning degree;

Step 2-2: according to the expertise of electrical network plan, set up the constraint knowledge storehouse of Real-Time Scheduling plan, and Real-Time Scheduling plan is adjusted automatically, the expertise constraint in constraint knowledge storehouse comprises generating maintenance mutual exclusive restrict, generation schedule temporal constraint, maintenance scheduling mutual exclusive restrict and generating set units limits.

In plan Time-Series analysis, be divided into following two kinds of situations:

A) judge whether Real-Time Scheduling planning data or ultra-short term data exist the shortage of data of certain time point, if exist, Real-Time Scheduling planning data or the ultra-short term data of this time point are considered as singular point;

B) in real-time generation schedule, generating set plan goes out force value rate of change in sequential and whether exceedes generating set plan and to exert oneself setting threshold, as exceeded, is considered as singular point.

In plan association analysis, determine whether rational Real-Time Scheduling plan according to following four Rule of judgment, arbitraryly not meet if wherein have, be then judged to be irrational Real-Time Scheduling plan;

A) in real-time generation schedule, generating set plan goes out force value not higher than the actual upper limit of exerting oneself of generating set;

B) not grid-connected generating set does not have plan and exerts oneself;

C) in real-time generation schedule generating set plan go out force value and real-time examination and repair in the works putting equipment in service state be consistent;

D) real-time examination and repair in the works putting equipment in service state there is not numerical value contradiction.

During achieve an equilibrium ikn planning degree is analyzed, the analysis of achieve an equilibrium ikn planning degree is carried out to ultra-short term, real-time generation schedule, in real time contact trading program, has:

${\hat{P}}_L=\underset{g\∈M}{\mathrm{\Σ}}{P}_g-P_{\mathrm{exch}}=\underset{d\∈N}{\mathrm{\Σ}}{P}_d+P_{\mathrm{loss}}+\underset{\mathrm{pl}\∈O}{\mathrm{\Σ}}{P}_{\mathrm{pl}}$

Wherein, for system loading predicted value, P _gfor generating set plan goes out force value, P _exchforce value is gone out, P for getting in touch with trading program in real time _dfor bus load predicted value, P _lossfor system losses, P _plfor the station service active power in power plant, M is generating set quantity, and N is bus load quantity, and O is power plant quantity.

In generating maintenance mutual exclusive restrict, there is mutual exclusion when conflicting in real-time generation schedule and real-time examination and repair plan, adjusts, specifically have with real-time examination and repair state for benchmark is exerted oneself to generating set:

$P_{G_{\mathrm{\α}},t}^{\′}=P_{G_{\mathrm{\α}},t}\·S_{G_{\mathrm{\α}},t}$

Wherein, force value is gone out for adjusting generating set plan in rear real-time generation schedule, force value is gone out for adjusting generating set plan in front real-time generation schedule, for real-time examination and repair putting equipment in service state in the works;

In generation schedule temporal constraint, when real-time generation schedule exists singular point in sequential, go out force value adjustment by the generating set plan near its time point, specifically have:

$P_{G_{\mathrm{\α}},t}^{\′}=\frac{P_{\mathrm{\α},t-1}+P_{\mathrm{\α},t+1}}{2}$

Wherein, P _{α, t-1}for the generating set plan near the time t-1 of singular point place goes out force value, P _{α, t+1}for the plan near the time t+1 of singular point place goes out force value;

In maintenance scheduling mutual exclusive restrict, real-time examination and repair when putting equipment in service state exists mutual exclusive restrict in the works, adjusts maintenance scheduling for benchmark with devices in system actual motion state, specifically has:

S′ _β,plan＝S _β,ope

Wherein, S ' _{β, plan}the state that puts into operation of real-time examination and repair equipment β in the works after adjustment, S _{β, ope}for the actual motion state of devices in system β;

In generating set units limits, when generating set plan goes out that in force value and system, generating set exists mutual exclusive restrict in real-time generation schedule, for benchmark, force value adjustment to generating set plan in real-time generation schedule with generating set in system, specifically has:

$P_{G_{\mathrm{\&alpha;}},t}^{\&prime;}=\left\{\begin{array}{cc}{P}_{G_{\mathrm{\&alpha;}},\max }& P_{G_{\mathrm{\&alpha;}},t}>P_{G_{\mathrm{\&alpha;}},\max }\\ P_{G_{\mathrm{\&alpha;}},t}& P_{G_{\mathrm{\&alpha;}},\min }<P_{G_{\mathrm{\&alpha;}},t}<P_{G_{\mathrm{\&alpha;}},\max }\\ P_{G_{\mathrm{\&alpha;}},\min }& P_{G_{\mathrm{\&alpha;}},t}<P_{G_{\mathrm{\&alpha;}},\min }\end{array}\right.$

Wherein, for, with be respectively generating set in system to exert oneself upper and lower bound.

Described step 3 specifically comprises the following steps:

Step 3-1: set up multibreak real power control model, increases section active power deviation equation in power flow algorithm, and section active power deviation the Representation Equation is:

$\mathrm{\&Delta;}{P}_{\mathrm{cut}}\left(\mathrm{\&gamma;}\right)=\stackrel{N_{\mathrm{line}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{line}}-P_{\mathrm{des}}\left(\mathrm{\&gamma;}\right)=0$

Wherein, Δ P _cut(γ) be the active power deviation of section γ, P _linefor the active power of section γ, be the N of this section _linethe active power summation of bar circuit; P _des(γ) be the active power desired value of section γ;

Step 3-2: the active power adjustable strategies adopting the method for successive approximation, the active power deviation according to section is adjusted by multistep, gradually section active power is adjusted to active power desired value;

Step 3-3: set up generating set active power controller equation, specifically have:

P _δ(α)＝f _p(α)+η(δ)ΔP _vail(α)

Wherein, P _δ(α) be δ the active power controlling α platform generating set in a group of planes, f _p(α) be the topological constraints of generating set node, η (δ) is δ the active power controller factor controlling a group of planes, Δ P _vail(α) exert oneself for the generating set of Weight is adjustable.

Described step 4 specifically comprises the following steps:

Step 4-1: carry out reactive voltage based on voltage power-less subregion and control on the spot;

Step 4-2: idle for hub node amount of unbalance is dispensed to the reactive power source in voltage power-less subregion.

In described step 5, adopt and carry out iterative computation based on the Newton method of node power equilibrium equation, and the running status of certainty annuity, determine the Load flow calculation data of electrical network future trend dynamic secure estimation, comprise the power distribution in voltage magnitude on each bus and phase angle, network and power loss.

Compared with prior art, beneficial effect of the present invention is:

The invention solves traditional online dynamic secure estimation algorithm cannot effectively to the problem that safety and stability in the following electrical network short time is analyzed, for the following operational mode safety analysis of electrical network provides trend flow data accurately.Reasonability identification is carried out to Real-Time Scheduling plan, error message in automatic Identification Scheme data, and carry out Plan rescheduling according to actual electric network expertise, for trend Load flow calculation provides good data basis; Set up multibreak real power control model of large-scale interconnected power system, adopt the adjustable strategies of Step wise approximation true value, automatic fine tuning generating real power control, avoid system gain merit deviation larger time trend situation about not restraining; Adopt hierarchical and regional balance principle, carry out reactive voltage based on idle partition method and control on the spot, the convergence of further raising trend power flow algorithm.

Accompanying drawing explanation

Fig. 1 is the trend tidal current computing method flow chart based on presence estimation and Real-Time Scheduling plan in the embodiment of the present invention.

Fig. 2 is the 500kV electric pressure branch road accuracy rate statistics schematic diagram of system in the embodiment of the present invention;

Fig. 3 is the 220kV electric pressure branch road accuracy rate statistics schematic diagram of system in the embodiment of the present invention.

Embodiment

Below in conjunction with accompanying drawing, the present invention is described in further detail.

As Fig. 1, the invention provides and a kind ofly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, said method comprising the steps of:

Step 1: determine electrical network future trend running mode data;

Step 2: reasonability identification is carried out to Real-Time Scheduling plan, and it is adjusted automatically;

Step 3: set up multibreak real power control model, carries out automatic fine tuning generating active power controller;

Step 4: carry out reactive voltage and control on the spot, and distribute hub node idle amount of unbalance;

Step 5: carry out Load flow calculation, generates the Load flow calculation data being used for electrical network future trend dynamic secure estimation.

In described step 1, electrical network future trend running mode data is included in line states data estimator, Real-Time Scheduling planning data and ultra-short term data.

According to the online grid measurement data obtained, adopt statistical method to estimate dynamic power system internal state, obtain presence data estimator;

On the basis of operation plan a few days ago, in conjunction with ultra-short term information, ad hoc inspection and repair information and transregional electricity transaction application information, to formulate in the following 5min of electrical network or in 60min the corresponding period real-time generation schedule and get in touch with trading program in real time, obtain Real-Time Scheduling planning data;

Utilize existing history daily load data and meteorological data, the load value of corresponding period in the following 5min of electrical network or in 60min is estimated, complete the ultra-short term comprising system loading prediction and bus load prediction, obtain ultra-short term data.

Described step 2 specifically comprises the following steps:

Step 2-1: carry out reasonability identification to Real-Time Scheduling plan, comprises plan Time-Series analysis, plan association analysis and the analysis of achieve an equilibrium ikn planning degree;

Step 2-2: according to the expertise of electrical network plan, set up the constraint knowledge storehouse of Real-Time Scheduling plan, and Real-Time Scheduling plan is adjusted automatically, the expertise constraint in constraint knowledge storehouse comprises generating maintenance mutual exclusive restrict, generation schedule temporal constraint, maintenance scheduling mutual exclusive restrict and generating set units limits.

In plan Time-Series analysis, be divided into following two kinds of situations:

A) judge whether Real-Time Scheduling planning data or ultra-short term data exist the shortage of data of certain time point, if exist, Real-Time Scheduling planning data or the ultra-short term data of this time point are considered as singular point;

B) in real-time generation schedule, generating set plan goes out force value rate of change in sequential and whether exceedes generating set plan and to exert oneself setting threshold, as exceeded, is considered as singular point.

In plan association analysis, determine whether rational Real-Time Scheduling plan according to following four Rule of judgment, arbitraryly not meet if wherein have, be then judged to be irrational Real-Time Scheduling plan;

A) in real-time generation schedule, generating set plan goes out force value not higher than the actual upper limit of exerting oneself of generating set;

B) not grid-connected generating set does not have plan and exerts oneself;

C) in real-time generation schedule generating set plan go out force value and real-time examination and repair in the works putting equipment in service state be consistent;

D) real-time examination and repair in the works putting equipment in service state there is not numerical value contradiction (numerical value contradiction refers to two records that there is same equipment, and the state that puts into operation is contrary).

During achieve an equilibrium ikn planning degree is analyzed, the analysis of achieve an equilibrium ikn planning degree is carried out to ultra-short term, real-time generation schedule, in real time contact trading program, has:

${\hat{P}}_L=\underset{g\&Element;M}{\mathrm{\&Sigma;}}{P}_g-P_{\mathrm{exch}}=\underset{d\&Element;N}{\mathrm{\&Sigma;}}{P}_d+P_{\mathrm{loss}}+\underset{\mathrm{pl}\&Element;O}{\mathrm{\&Sigma;}}{P}_{\mathrm{pl}}---\left(1\right)$

Wherein, for system loading predicted value, P _gfor generating set plan goes out force value, P _exchforce value is gone out, P for getting in touch with trading program in real time _dfor bus load predicted value, P _lossfor system losses, P _plfor the station service active power in power plant, M is generating set quantity, and N is bus load quantity, and O is power plant quantity.

In generating maintenance mutual exclusive restrict, there is mutual exclusion when conflicting in real-time generation schedule and real-time examination and repair plan, adjusts, specifically have with real-time examination and repair state for benchmark is exerted oneself to generating set:

$P_{G_{\mathrm{\&alpha;}},t}^{\&prime;}=P_{G_{\mathrm{\&alpha;}},t}\&CenterDot;S_{G_{\mathrm{\&alpha;}},t}---\left(2\right)$

Wherein, force value is gone out for adjusting generating set plan in rear real-time generation schedule, force value is gone out for adjusting generating set plan in front real-time generation schedule, for real-time examination and repair putting equipment in service state in the works;

In generation schedule temporal constraint, when real-time generation schedule exists singular point in sequential, go out force value adjustment by the generating set plan near its time point, specifically have:

$P_{G_{\mathrm{\&alpha;}},t}^{\&prime;}=\frac{P_{\mathrm{\&alpha;},t-1}+P_{\mathrm{\&alpha;},t+1}}{2}---\left(3\right)$

Wherein, P _{α, t-1}for the generating set plan near the time t-1 of singular point place goes out force value, P _{α, t+1}for the plan near the time t+1 of singular point place goes out force value;

In maintenance scheduling mutual exclusive restrict, real-time examination and repair when putting equipment in service state exists mutual exclusive restrict in the works, adjusts maintenance scheduling for benchmark with devices in system actual motion state, specifically has:

S′ _β,plan＝S _β,ope(4)

Wherein, S ' _{β, plan}the state that puts into operation of real-time examination and repair equipment β in the works after adjustment, S _{β, ope}for the actual motion state of devices in system β;

In generating set units limits, when generating set plan goes out that in force value and system, generating set exists mutual exclusive restrict in real-time generation schedule, for benchmark, force value adjustment to generating set plan in real-time generation schedule with generating set in system, specifically has:

$P_{G_{\mathrm{\&alpha;}},t}^{\&prime;}=\left\{\begin{array}{cc}{P}_{G_{\mathrm{\&alpha;}},\max }& P_{G_{\mathrm{\&alpha;}},t}>P_{G_{\mathrm{\&alpha;}},\max }\\ P_{G_{\mathrm{\&alpha;}},t}& P_{G_{\mathrm{\&alpha;}},\min }<P_{G_{\mathrm{\&alpha;}},t}<P_{G_{\mathrm{\&alpha;}},\max }\\ P_{G_{\mathrm{\&alpha;}},\min }& P_{G_{\mathrm{\&alpha;}},t}<P_{G_{\mathrm{\&alpha;}},\min }\end{array}\right.---\left(5\right)$

Wherein, for, with be respectively generating set in system to exert oneself upper and lower bound.

Described step 3 specifically comprises the following steps:

Step 3-1: set up multibreak real power control model, increases section active power deviation equation in power flow algorithm, and section active power deviation the Representation Equation is:

$\mathrm{\&Delta;}{P}_{\mathrm{cut}}\left(\mathrm{\&gamma;}\right)=\stackrel{N_{\mathrm{line}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{line}}-P_{\mathrm{des}}\left(\mathrm{\&gamma;}\right)=0---\left(6\right)$

Wherein, Δ P _cut(γ) be the active power deviation of section γ, P _linefor the active power of section γ, be the N of this section _linethe active power summation of bar circuit; P _des(γ) be the active power desired value of section γ;

Step 3-2: the active power adjustable strategies adopting the method for successive approximation, the active power deviation according to section is adjusted by multistep, gradually section active power is adjusted to active power desired value;

Control measure are divided into one-step control and multistep to control by the active power adjustment strategy of Step wise approximation.Each step setting one substep target during multistep controls, the desired value of this substep target Step wise approximation section.Each decoupled method is equivalent to an one-step control, and its step-length is the difference of substep target and this step initial value.

The calculation procedure of one-step control is as follows:

1) select according to principle given in advance the unit and the number of units that control and balance a group of planes;

2) utilize the process to the constraint of generator limit value, make section power control to become the start and stop of continuous print generator and self-regulating process, adjustment process adjusts non-regulation and control unit by priority and regulation and control unit realizes.

On the basis of one-step control, the calculation procedure that multistep controls adds following content:

1) next step step-length and this step control in the maximum variable quantity of line voltage be inversely proportional to;

2) the trend solution that previous step calculates walks initial value as this, makes iteration starting point close to true solution;

3) do not restrain if calculate, step-length reduces by half automatically, again this step controlling calculation;

4) when step-length is less than default threshold, trend does not still restrain, then control failure, terminates to calculate.

Step 3-3: set up generating set active power controller equation, specifically have:

P _δ(α)＝f _p(α)+η(δ)ΔP _vail(α) (7)

Wherein, P _δ(α) be δ the active power controlling α platform generating set in a group of planes, f _p(α) be the topological constraints of generating set node, η (δ) is δ the active power controller factor controlling a group of planes, Δ P _vail(α) exert oneself for the generating set of Weight is adjustable.

Described step 4 specifically comprises the following steps:

Step 4-1: carry out reactive voltage based on voltage power-less subregion and control on the spot;

Step 4-2: idle for hub node amount of unbalance is dispensed to the reactive power source in voltage power-less subregion.

Voltage power-less subregion is that whole system is divided into several subregions by hierarchical and regional balance principle, for layering and zoning, the in-situ balancing realizing idle control provides basis.Concrete, by the set of electric network composition chosen in advance hub node, according to electrical distance, system node is divided into several reactive balance region Ω be made up of hub node and multiple reactive power source node _i, and calculate Ω _iinterior each reactive power source node is to the idle distribution coefficient λ of hub node i _ki, have:

${\mathrm{\&lambda;}}_{\mathrm{ki}}=\frac{1/x_{\mathrm{ik}}}{\underset{J\&Element;\mathrm{\&Omega;}}{\mathrm{\&Sigma;}}1/x_{\mathrm{ij}}}---\left(8\right)$

Wherein, x _ijfor the branch road reactance between hub node i and reactive power source node j.

Reactive voltage controls on the spot, new expanding node type is increased in power flow equation, and increase var-volt regulation measure in its iterative process, according to idle units limits and voltage constraint, idle for node amount of unbalance is dispensed to each reactive source, realizes automatically adjusting on the spot of reactive voltage.

New extensions node type: type, the meritorious P of its generating _gknown, generate electricity idle Q _git is unknown with voltage magnitude V, $\tilde{Q}=\left\{Q_G\right|Q_{G\min }\&le;Q_G{\&le;Q}_{G\max }\},\tilde{V}=\{V|V_{\min }\&le;V\&le;V_{\max }\};$ type, its active-power P that generates electricity _gthe unknown, voltage-phase θ is known, and generate electricity idle and voltage magnitude the unknown, but can change within the specific limits.Based on above-mentioned expanding node type, in power flow equation iterative process, increase var-volt regulation equation, as shown in the formula:

$Q_{\mathrm{Gi}}^{\left(t\right)}=Q_{\mathrm{Gi}}^{(t-1)}+Q_{\mathrm{fpi}}---\left(9\right)$

Wherein, Q _fpito be assigned in node i when being the t-1 time iteration idle exerts oneself; with be respectively the iterative value of idle the t-1 time of exerting oneself of node i and t time.

In idle iterative process, according to idle distribution coefficient the idle amount of unbalance of node i is dispensed to each reactive source in reactive balance region, and the idle amount of unbalance of computing node i again shown in (10) ~ (11).

$Q_{\mathrm{fpi}}=l_{\mathrm{bcl}}\mathrm{\&Delta;}{Q}_i^{(t-1)}+\underset{j\&Element;{\mathrm{\&Omega;}}_i}{\mathrm{\&Sigma;}}{l}_{\mathrm{bc}2}{\mathrm{\&lambda;}}_{\mathrm{ij}}\mathrm{\&Delta;}{Q}_j^{(t-1)}+\mathrm{\&Delta;}{Q}_{\mathrm{dyyxi}}+\underset{j\&Element;{\mathrm{\&Omega;}}_i}{\mathrm{\&Sigma;}}\mathrm{\&Delta;}{Q}_{\mathrm{dyyxj}}+\mathrm{\&Delta;}{Q}_{\mathrm{clyxi}}+\underset{j\&Element;{\mathrm{\&Omega;}}_i}{\mathrm{\&Sigma;}}\mathrm{\&Delta;}{Q}_{\mathrm{clyxi}}---\left(10\right)$

$\mathrm{\&Delta;}{Q}_i^{\left(t\right)}=Q_{\mathrm{Gi}}^{\left(t\right)}-Q_{\mathrm{Li}}^{\left(t\right)}-U_i^{\left(t\right)}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{U}_j^{\left(t\right)}(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}^{\left(t\right)}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}^{\left(t\right)})---\left(11\right)$

In formula, l _bc1and l _bc2for step-size factor; with be respectively the idle amount of unbalance of t-1 iteration node i, j; Δ Q _dyyxjfor being assigned to the reactive power of node j because of node i voltage out-of-limit; Δ Q _dyyxifor the adjustment amount of himself reactive power during node i voltage out-of-limit; Δ Q _clyxjfor being assigned to the reactive power of node j because node i reactive power is out-of-limit; Δ Q _clyxifor node i reactive power out-of-limit time himself reactive power adjustment amount.

According to Reactive-power control ability and the voltage-regulation nargin thereof of each reactive power source node, by idle for hub node amount of unbalance Δ Q _ibe dispensed to each reactive power source node of surrounding, shown in (12) ~ (13).

Q′ _Gi＝Q _Gi-l _bc1|ΔQ _i|sgn(ΔQ _i) (12)

Q′ _Gk＝Q _Gk-l _bc2λ _ki|ΔQ _i|sgn(ΔQ _i) (13)

In formula, Q _gk, Q ' _gkbe respectively reactive power source node k in subregion adjust before and idle after adjustment exert oneself; Sgn (Δ Q _i) be variable Δ Q _isign function, value-1 or 1.

In described step 5, adopt and carry out iterative computation based on the Newton method of node power equilibrium equation, and the running status of certainty annuity, determine the Load flow calculation data of electrical network future trend dynamic secure estimation, comprise the power distribution in voltage magnitude on each bus and phase angle, network and power loss.

According to the method described above, using certain real system as Knowledge Verification Model, carry out statistical analysis to the trend trend accuracy on April 16th, 2014, wherein the sampling interval is 15min, i.e. 96 time points of 0:15-24:00.Statistic analysis result is as shown in the table, and the Average Accuracy of 550kV trend trend is the Average Accuracy of 94.56%, 220kV trend trend is 94.56%, concrete as table 1:

Table 1

Accuracy statistical information	Content
		500kV number of branches	380～410
220kV number of branches	2640～2660
		500kV trend accuracy rate	94.56％
220kV trend accuracy rate	95.48％

Above-mentioned instance analysis shows: this method overcomes the traditional online dynamic secure estimation algorithm of solution cannot effectively to the problem that safety and stability in the following electrical network short time is analyzed, based on presence data estimator and Real-Time Scheduling plan, simulation actual electric network plan operation characteristic, calculate rational reactive voltage plan and generator output control, the dynamic secure estimation for electrical network future trend provides flow data accurately.

Finally should be noted that: above embodiment is only in order to illustrate that technical scheme of the present invention is not intended to limit; those of ordinary skill in the field still can modify to the specific embodiment of the present invention with reference to above-described embodiment or equivalent replacement; these do not depart from any amendment of spirit and scope of the invention or equivalent replacement, are all applying within the claims of the present invention awaited the reply.

Claims (11)

1. estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: said method comprising the steps of:

Step 1: determine electrical network future trend running mode data;

Step 2: reasonability identification is carried out to Real-Time Scheduling plan, and it is adjusted automatically;

Step 3: set up multibreak real power control model, carries out automatic fine tuning generating active power controller;

Step 4: carry out reactive voltage and control on the spot, and distribute hub node idle amount of unbalance;

Step 5: carry out Load flow calculation, generates the Load flow calculation data being used for electrical network future trend dynamic secure estimation.

2. according to claim 1ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: in described step 1, electrical network future trend running mode data is included in line states data estimator, Real-Time Scheduling planning data and ultra-short term data.

3. according to claim 2ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: according to the online grid measurement data obtained, adopt statistical method to estimate dynamic power system internal state, obtain presence data estimator;

On the basis of operation plan a few days ago, in conjunction with ultra-short term information, ad hoc inspection and repair information and transregional electricity transaction application information, to formulate in the following 5min of electrical network or in 60min the corresponding period real-time generation schedule and get in touch with trading program in real time, obtain Real-Time Scheduling planning data;

Utilize existing history daily load data and meteorological data, the load value of corresponding period in the following 5min of electrical network or in 60min is estimated, complete the ultra-short term comprising system loading prediction and bus load prediction, obtain ultra-short term data.

4. according to claim 1ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: described step 2 specifically comprises the following steps:

Step 2-1: carry out reasonability identification to Real-Time Scheduling plan, comprises plan Time-Series analysis, plan association analysis and the analysis of achieve an equilibrium ikn planning degree;

Step 2-2: according to the expertise of electrical network plan, set up the constraint knowledge storehouse of Real-Time Scheduling plan, and Real-Time Scheduling plan is adjusted automatically, the expertise constraint in constraint knowledge storehouse comprises generating maintenance mutual exclusive restrict, generation schedule temporal constraint, maintenance scheduling mutual exclusive restrict and generating set units limits.

5. according to claim 4ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: plan, in Time-Series analysis, to be divided into following two kinds of situations:

A) judge whether Real-Time Scheduling planning data or ultra-short term data exist the shortage of data of certain time point, if exist, Real-Time Scheduling planning data or the ultra-short term data of this time point are considered as singular point;

B) in real-time generation schedule, generating set plan goes out force value rate of change in sequential and whether exceedes generating set plan and to exert oneself setting threshold, as exceeded, is considered as singular point.

6. according to claim 4ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: in plan association analysis, rational Real-Time Scheduling plan is determined whether according to following four Rule of judgment, arbitraryly not meet if wherein have, be then judged to be irrational Real-Time Scheduling plan;

A) in real-time generation schedule, generating set plan goes out force value not higher than the actual upper limit of exerting oneself of generating set;

B) not grid-connected generating set does not have plan and exerts oneself;

C) in real-time generation schedule generating set plan go out force value and real-time examination and repair in the works putting equipment in service state be consistent;

D) real-time examination and repair in the works putting equipment in service state there is not numerical value contradiction.

7. according to claim 4ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: during achieve an equilibrium ikn planning degree is analyzed, the analysis of achieve an equilibrium ikn planning degree is carried out to ultra-short term, real-time generation schedule, in real time contact trading program, has:

${\hat{P}}_L=\underset{g\&Element;M}{\mathrm{\&Sigma;}}{P}_g-P_{\mathrm{exch}}=\underset{d\&Element;N}{\mathrm{\&Sigma;}}{P}_d+P_{\mathrm{loss}}+\underset{\mathrm{pl}\&Element;O}{\mathrm{\&Sigma;}}{P}_{\mathrm{pl}}$

Wherein, for system loading predicted value, P _gfor generating set plan goes out force value, P _exchforce value is gone out, P for getting in touch with trading program in real time _dfor bus load predicted value, P _lossfor system losses, P _plfor the station service active power in power plant, M is generating set quantity, and N is bus load quantity, and O is power plant quantity.

8. according to claim 4ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: in generating maintenance mutual exclusive restrict, there is mutual exclusion when conflicting in real-time generation schedule and real-time examination and repair plan, adjust for benchmark is exerted oneself to generating set with real-time examination and repair state, specifically have:

$P_{G_{\mathrm{\&alpha;}},t}^{\&prime;}=P_{G_{\mathrm{\&alpha;}},t}\&CenterDot;S_{G_{\mathrm{\&alpha;}},t}$

Wherein, force value is gone out for adjusting generating set plan in rear real-time generation schedule, force value is gone out for adjusting generating set plan in front real-time generation schedule, for real-time examination and repair putting equipment in service state in the works;

In generation schedule temporal constraint, when real-time generation schedule exists singular point in sequential, go out force value adjustment by the generating set plan near its time point, specifically have:

$P_{G_{\mathrm{\&alpha;}},t}^{\&prime;}=\frac{P_{\mathrm{\&alpha;},t-1}+P_{\mathrm{\&alpha;},t+1}}{2}$

Wherein, P _{α, t-1}for the generating set plan near the time t-1 of singular point place goes out force value, P _{α, t+1}for the plan near the time t+1 of singular point place goes out force value;

In maintenance scheduling mutual exclusive restrict, real-time examination and repair when putting equipment in service state exists mutual exclusive restrict in the works, adjusts maintenance scheduling for benchmark with devices in system actual motion state, specifically has:

S′ _β,plan＝S _β,ope

Wherein, S ' _{β, plan}the state that puts into operation of real-time examination and repair equipment β in the works after adjustment, S _{β, ope}for the actual motion state of devices in system β;

In generating set units limits, when generating set plan goes out that in force value and system, generating set exists mutual exclusive restrict in real-time generation schedule, for benchmark, force value adjustment to generating set plan in real-time generation schedule with generating set in system, specifically has:

$P_{G_{\mathrm{\&alpha;},t}}^{\&prime;}=\left\{\begin{array}{cc}{P}_{G_{\mathrm{\&alpha;}},\max }& P_{G_{\mathrm{\&alpha;}},t}>P_{G_{\mathrm{\&alpha;}},\max }\\ P_{G_{\mathrm{\&alpha;}},t}& P_{G_{\mathrm{\&alpha;}},\min }<P_{G_{\mathrm{\&alpha;}},t}<P_{G_{\mathrm{\&alpha;}},\max }\\ P_{G_{\mathrm{\&alpha;}},\min }& P_{G_{\mathrm{\&alpha;}},t}<P_{G_{\mathrm{\&alpha;}},\min }\end{array}\right.$

Wherein, for, with be respectively generating set in system to exert oneself upper and lower bound.

9. according to claim 1ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: described step 3 specifically comprises the following steps:

Step 3-1: set up multibreak real power control model, increases section active power deviation equation in power flow algorithm, and section active power deviation the Representation Equation is:

${\mathrm{\&Delta;P}}_{\mathrm{cut}}\left(\mathrm{\&gamma;}\right)=\stackrel{N_{\mathrm{line}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{line}}-P_{\mathrm{des}}\left(\mathrm{\&gamma;}\right)=0$

Wherein, Δ P _cut(γ) be the active power deviation of section γ, P _linefor the active power of section γ, be the N of this section _linethe active power summation of bar circuit; P _des(γ) be the active power desired value of section γ;

Step 3-2: the active power adjustable strategies adopting the method for successive approximation, the active power deviation according to section is adjusted by multistep, gradually section active power is adjusted to active power desired value;

Step 3-3: set up generating set active power controller equation, specifically have:

P _δ(α)＝f _p(α)+η(δ)ΔP _vail(α)

Wherein, P _δ(α) be δ the active power controlling α platform generating set in a group of planes, f _p(α) be the topological constraints of generating set node, η (δ) is δ the active power controller factor controlling a group of planes, Δ P _vail(α) exert oneself for the generating set of Weight is adjustable.

10. according to claim 1ly to estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: described step 4 specifically comprises the following steps:

Step 4-1: carry out reactive voltage based on voltage power-less subregion and control on the spot;

Step 4-2: idle for hub node amount of unbalance is dispensed to the reactive power source in voltage power-less subregion.

11. according to claim 1ly estimate and the trend tidal current computing method of Real-Time Scheduling plan based on presence, it is characterized in that: in described step 5, adopt and carry out iterative computation based on the Newton method of node power equilibrium equation, and the running status of certainty annuity, determine the Load flow calculation data of electrical network future trend dynamic secure estimation, comprise the power distribution in voltage magnitude on each bus and phase angle, network and power loss.