CN104636831A - Multi-power-grid-oriented hydropower station short period peak load regulation characteristic value searching method - Google Patents

Abstract

The invention discloses a multi-power-grid-oriented hydropower station short period peak load regulation characteristic value searching method, and belongs to the technical field of hydroelectric energy optimized operation and electrical power system electricity generation optimized dispatching. With the maximal and minimal values of each electricity receiving power grid residual load as a characteristic value, the output at the period of time of the characteristic value is adjusted gradually in a characteristic value limit range, and the electricity receiving process of each power grid is corrected gradually in order to meet the hydropower station operation limit requirement. Multi-power-grid short period peak regulation is divided into two scheduling modes of a non-abandoned water mode and an abandoned water mode according to available water quantity of a scheduling period. Under the non-abandoned water mode, based on the consideration of power grid typical load characteristic and peak regulation requirement, under the precondition that a lower course comprehensive water usage demand is met, hydropower station peak regulation capacity gain is exerted as much as possible, more electricity is generated at electricity utilization peak and less electricity is generated at electricity utilization valley; under the abandoned water mode, the hydropower station is in fully loaded running, under the requirement of appointed power grid distribution ratio, with minimal rest load variance as a target, a power grid electricity receiving plan is made by a classical particle swarm optimization algorithm.

Description

A kind of power station short-term peak regulation eigenvalue search method towards many electrical networks

Technical field

The invention belongs to HYDROELECTRIC ENERGY optimizing operation and electric system generation optimization dispatching technique field, more specifically, relate to a kind of power station short-term peak regulation eigenvalue search method towards many electrical networks.

Background technology

The current large hydropower station building or put into operation such as small stream Luo Du, Burner zone, ERTAN Hydroelectric ProJect etc., all simultaneously to multiple provincial power network power transmission, but because the load magnitude of each provincial power network, peak valley occur that period, peak-valley difference are different, these power stations need take into full account the part throttle characteristics of each powered electrical network when formulating and exerting oneself plan, with the peak regulation requirement of each electrical network of dynamic response.

More than the existing many employings of power station short-term peak regulation lotus successively move afterwards method and successively cutting load method to generate electricity under carrying out many grid transmissions situation planning.But, these methods carry out balance of electric power and ener calculating by limiting the mode of power station to each peak load regulation network capacity, although the gross capability after superposition meets power station and to exert oneself upper limit constraint, limit the performance of power station peak, each electrical network all can not get satisfied peak regulation effect.And due to the generating capacity in power station limited and exert oneself and to require and climbing rate constrained by safe and stable operation, if only divide electricity to carry out balance of electric power and ener calculating than respectively to hydropower station amount by electrical network, when several powered electrical network crest segment overlaps, the power curve obtained after superposing easily is caused not meet power station actual motion demand.Therefore, current methods, when processing power station many peak load regulation networks problem, is difficult to give full play to power station peaking capacity benefits, and cannot takes into account part throttle characteristics and the peak regulation demand of multiple powered electrical network.

Summary of the invention

For above defect or the Improvement requirement of prior art, the invention provides a kind of power station short-term peak regulation eigenvalue search method towards many electrical networks, hydroelectric station operation restriction and electrical network can be met by electricity proportion requirement to formulate, while can take into account again the generation schedule of each peaking demand of power grid.By on the basis considering each powered network load characteristic, using the most value of powered electrical network residue load as eigenwert, in eigenwert limited field, progressively adjust the eigenwert place period exerts oneself, and progressively revises each electrical network by electric process, to meet hydroelectric station operation restriction requirement.

The invention provides a kind of power station short-term peak regulation eigenvalue search method towards many electrical networks, comprise the following steps:

Step 1 calculates schedule periods water volume that can be utilized, and estimates the maximum output capacity of power station schedule periods, judges whether power station abandons water according to average letdown flow, has and then performs step 7, otherwise performs step 2;

The each electrical network of step 2 with described maximum output capacity for the upper limit, obtain many power curves by the respective prediction load curve successively cutting load method that is used alone and add up, obtain power station initially to exert oneself process, and judge whether described process of initially exerting oneself meets period minimax units limits, if do not meet, the part not meeting constraint is added up and has been used as the adjustment step delta E that exerts oneself, then perform step 3, otherwise perform step 4;

Step 3 when hydropower station is not enough, then performs sub-step (3-1); When hydropower station is too much, then perform sub-step (3-2):

(3-1) exert oneself by described adjustment step delta E increase of exerting oneself, divide electricity to compare R by each electrical network the described adjustment step delta E that exerts oneself ^gbe divided into G part: Δ E ^g=R ^gΔ E, wherein, Δ E ^grepresent and need the adjustment step-length of exerting oneself being assigned to g electrical network, g ∈ [1, G], and find the maximum period of lotus more than each electrical network, and by Δ E ^gbe dispensed to this period; If affined restriction, the electricity of increase can not be dissolved by the single period completely, then the lotus maximum period more than being dispensed to by dump energy outside the adjusted period, repeat this process until Δ E ^gjoin complete, recalculate power station gross capability, perform step 4;

(3-2) reduce to exert oneself by the described adjustment step delta E that exerts oneself, by the method for described sub-step (3-1), under the prerequisite meeting firm output powcr constraint, travel through each powered electrical network one by one, the period reduction finding remaining lotus minimum is exerted oneself, until Δ E is assigned; Recalculate power station gross capability, perform step 4;

Step 4 process power station exert oneself minimum duration constraint and exert oneself luffing constraint;

Step 5 to be exerted oneself process computation power station day part water level and letdown flow process according to power station, and the period of violating letdown flow and operating water level constraint is processed, judge that the scheduling end of term calculates water level and whether meets the last restriction of water level of control, if the last water level of described calculating is not equal to the last water level of described control, then perform step 6; Otherwise must arrive power station to exert oneself and the powered result of each electrical network, flow process terminates;

If the last water level of the described calculating of step 6 is greater than the last water level of described control, then hydropower station is not enough, and by the described adjustment step delta E that exerts oneself, increase is exerted oneself, and goes to and performs described step 3; If the last water level of described calculating is less than the last water level of described control, then hydropower station is excessive, reduces to exert oneself, go to and perform described step 3 by the described adjustment step delta E that exerts oneself;

The each electrical network of step 7 stochastic generation is by electric process, as the initial position of particle, the constraint of process minimum duration, and by classical particle group algorithm, to remain, load variance is minimum carries out optimizing for target, if particle position do not meet electrical network divide electricity than constraint, then by the method in described step 3 to revising by electric process, when algorithm reaches maximum iteration time, calculate and stop, using population globally optimal solution as the powered planned outcome of each electrical network.

In general, the above technical scheme conceived by the present invention compared with prior art, has following beneficial effect:

1, the present invention can take into account part throttle characteristics and the peak regulation demand of multiple powered electrical network, give full play to power station peaking capacity benefits, peak-valley difference after peak regulation all significantly reduces compared with before peak regulation with residue load variance index (evaluating residue load planarization), and each electrical network obtains satisfied peak regulation effect; Meanwhile, the scheme of exerting oneself that the present invention formulates can meet safe and stable operation of exerting oneself and require and climbing rate constrained, and the power curve obtained meets power station actual motion demand, has certain engineering practicability.

2, the peak regulating method computing velocity of the present invention's proposition is fast, and counting yield is high.In identical running environment, working time is much smaller than conventional dynamic programming algorithm, will be more remarkable carrying out the raising of Hydropower Stations in Large Scale group many peak load regulation networks computational valid time rate.

Accompanying drawing explanation

Fig. 1 is the power station short-term peak regulation eigenvalue search method flow diagram of the present invention towards many electrical networks;

Fig. 2 is the powered procedure chart of embodiment of the present invention Sichuan Electric Power Network;

Fig. 3 is the powered procedure chart of embodiment of the present invention Chongqing electricity grid;

Fig. 4 is embodiment of the present invention ERTAN Hydroelectric ProJect letdown flow and water level process figure.

Embodiment

In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.In addition, if below in described each embodiment of the present invention involved technical characteristic do not form conflict each other and just can mutually combine.

Figure 1 shows that the power station short-term peak regulation eigenvalue search method flow diagram of the present invention towards many electrical networks, specifically comprise the following steps:

Step 1: according to upper water power station letdown flow, local inflow forecast and next day water level to disappear plan, calculate power station schedule periods water volume that can be utilized W by following formula:

$W=\underset{t=1}{\overset{T}{\mathrm{\Σ}}}{B}_t\mathrm{\Δt}+\underset{t=1}{\overset{T}{\mathrm{\Σ}}}\underset{r\∈\mathrm{\Φ}}{\mathrm{\Σ}}{Q}_{r-1,t-{\mathrm{\τ}}_r}\mathrm{\Δt}+V_0-V_T$

Wherein, hop count (if scheduling slot length is 1 hour, then T is 24, and other situations by that analogy) when T represents that in a day, schedule periods is total; T ∈ [1, T] represents scheduling slot number; B _trepresent the local inflow of power station in the t period; Δ t represents that scheduling slot is long; φ represents that upstream and current calculating power station have the power station of direct hydraulic connection to gather; Q _{r-1, t-τ}for r power station, upstream is at the letdown flow of t period; τ _rrepresent the current time lag between r power station, upstream and current calculating power station; V ₀with V _trepresent power station schedule periods just last storage capacity respectively.

Pass through formula schedule periods water volume that can be utilized W is converted into average letdown flow downstream of hydro-power plant water level Z is obtained again by inquiry power station letdown flow-level of tail water relation table _d; And then, by formula calculate power station average water head wherein Z ₀with Z _trepresent power station schedule periods just last water level respectively.Then according to formula calculate the average generating capacity E of power station schedule periods ₀, wherein A represents output of power station coefficient.Last according to average water head power station maximum output capacity P under average water head next day is obtained by inquiry power station water consumption rate table ^max.According to average letdown flow judge whether power station abandons water, when average letdown flow be greater than power station when completely sending out flow, power station is in abandons aqueous mode, and all units are all full to be sent out, and unnecessary water is discharged by outlet structure escape works, then perform step 7 carry out abandoning aqueous mode under the powered plan of electrical network; Otherwise power station is in non-ly abandons aqueous mode, then perform step 2.

Step 2: determine that process and each electrical network are initially exerted oneself by electric process in power station;

Assuming that power station power transmission electrical network number is G, then electricity can be divided to compare R according to each electrical network ^g(g ∈ [1, G]) (R ^gfor the ratio of each power grid electric is sent in power station, be definite value in embodiments of the present invention) by power station schedule periods generating capacity E ₀be divided into G part, shown in following formula (1):

$\left\{\begin{array}{c}{E_0}^g=E_0{R}^g\\ R^1+R^2+\·\·\·+R^g+\·\·\·+R^G=1\end{array}\right.---\left(1\right)$

At each electrical network by electricity when determining, power station maximum output capacity P under the following per day head in each power station ^maxfor the upper limit, successively cutting load method is used to carry out balance of electric power and ener calculating respectively by the prediction load curve (this prediction load curve obtains by electrical network short-term load forecasting) of each electrical network, finally obtain power station the t period (t ∈ [1, T]) give g electrical network go out force value added up, can initially to be exerted oneself process in this power station this process of exerting oneself likely violates period minimax units limits, calculates by following formula (2) the adjustment step delta E that exerts oneself:

$\mathrm{\&Delta;E}=\underset{t,P_t>p^{\max }}{\overset{T}{\mathrm{\&Sigma;}}}(P_t-P^{\max })+\underset{t,P_t<p^{\min }}{\overset{T}{\mathrm{\&Sigma;}}}(P_t-P^{\min })---\left(2\right)$

Wherein, P _tfor the t period exerts oneself; p ^minfor period minimum load.

If do not meet the period exert oneself limit value constraint, i.e. Δ E ≠ 0, then perform step 3, otherwise perform step 4.

Step 3: adjust each electrical network by electric process by eigenvalue search method;

When hydropower station is not enough, then perform sub-step (3-1); When hydropower station is too much, then perform sub-step (3-2):

(3-1) increase by the adjustment step delta E that exerts oneself and exert oneself, considering that power station is to many grid transmissions situation, can be G part by a point electric score by Δ E, shown in following formula (3):

ΔE ^g＝R ^gΔE (3)

Wherein, Δ E ^grepresent and need the adjustment step-length of exerting oneself being assigned to g electrical network.

In embodiments of the present invention, as follows by Δ E ^gbe dispensed to corresponding powered electrical network: find the period that remaining lotus is maximum, and by Δ E ^gbe dispensed to this period; The restriction of condition if be tied, the electricity of increase can not be dissolved by the single period completely, then the lotus maximum period more than being dispensed to by dump energy outside the adjusted period, repeat this process until Δ E ^gbe assigned; Recalculate power station gross capability, perform step 4;

(3-2) reduce to exert oneself with step delta E, by the method for above-mentioned sub-step (3-1), under the prerequisite meeting firm output powcr constraint, travel through each powered electrical network one by one, the period reduction finding remaining lotus minimum is exerted oneself, until Δ E is assigned; Recalculate power station gross capability, perform step 4.

Step 4: process power station exert oneself minimum duration constraint and exert oneself luffing constraint, specifically comprise following sub-step:

(4-1) minimum duration constraint of exerting oneself is processed.Using the mean value of exerting oneself in minimum duration Δ t as going out force value in Δ t, constant with firm output powcr total amount while meeting minimum duration constraint;

(4-2) luffing constraint of exerting oneself is processed.Judge that adjacent two periods exert oneself P _t-1and P _twhether meet the luffing constraint of exerting oneself shown in following formula (4):

|P _t-1-P _t|≤ΔP (4)

Wherein, Δ P allows maximum output luffing the power station period.

When not meeting above-mentioned luffing of exerting oneself and retraining (4):

If P _t-1> P _t, then by P _tincrease to P _t-1-Δ P, now the t period exerts oneself and compares the additional issue P that initially exerts oneself _t-1-Δ P-P _t, by downward for whole power curve translation (P _t-1-Δ P-P _t)/T, the total amount that makes to exert oneself is constant, performs step 5;

If P _t-1< P _t, then by P _tbe reduced to P _t-1+ Δ P, whole curve is translation (P upwards _t-P _t-1-Δ P)/T, the total amount that makes to exert oneself is constant, performs step 5.

When meeting above-mentioned luffing of exerting oneself and retraining, perform step 5.

Step 5: according to output of power station process computation water level process;

Carry out " determining water with electricity " by gross capability process (known task of exerting oneself, to become a mandarin and water level at the beginning of the period, ask the minimum water consumption in power station) simulation calculation, determine power station day part water level and letdown flow process, and the following formula of violation (the 5) ~ letdown flow of (7) and the period of restriction of water level are processed, namely when the letdown flow calculated or water level value be not formula (5) ~ (7) Suo Shi within the scope of boundary value, then directly boundary value is set to.

Operating water level retrains:

$Z_t^{\min }\&le;Z_t\&le;Z_t^{\max }---\left(5\right)$

Letdown flow retrains:

$Q_t^{\min }\&le;Q_t\&le;Q_t^{\max }---\left(6\right)$

Period water level/letdown flow luffing constraint:

$\left\{\begin{array}{c}|Z_t-Z_{t-1}|\&le;\mathrm{\&Delta;Z}\\ |Q_t-Q_{t-1}|\&le;\mathrm{\&Delta;Q}\end{array}\right.---\left(7\right)$

Wherein, represent power station t period water level up-and-down boundary respectively; represent power station t period letdown flow up-and-down boundary respectively; Δ Z, Δ Q represent that the power station period allows maximum stage luffing and letdown flow luffing respectively.

Judge that the scheduling end of term calculates water level Z _cwhether meet the scheduling end of term restriction of water level shown in following formula (8):

Z _c＝Z _g(8)

Wherein, Z _cwith Z _grepresent that power station calculates last water level and scheduling end of term water level control value respectively.

If calculate last water level Z _cbe not equal to and control last water level Z _g(namely not meeting last restriction of water level), then perform step 6, revises the period and exert oneself to reach last water lev el control requirement; Otherwise must arrive power station to exert oneself and the powered result of each electrical network, flow process terminates.

Step 6: last water level correction strategy, determines the adjustment step delta E that exerts oneself, and adjusts each electrical network by electric process by eigenvalue search method;

If calculate last water level Z _cbe greater than last water level control value Z _g, then hydropower station is not enough, can increase and exert oneself, go to step 3 corrections by electric process by the adjustment step delta E that exerts oneself; If calculate last water level Z _cbe less than last water level control value Z _g, then hydropower station is excessive, need reduce to exert oneself by the adjustment step delta E that exerts oneself, go to step 3 corrections by electric process.

Exert oneself shown in the adjustment following formula of step delta E computing method (9):

ΔE＝a(Z _c-Z _g)+b (9)

Wherein, a and b represents step-length regulation coefficient respectively, and a suitably adjusts according to the summation of exerting oneself of power station under current average water head, and b determines jointly according to water level solving precision ε and coefficient a, i.e. b=ε × a.Obtain the suitable adjustment step delta E that exerts oneself thus and can ensure that iterations is less, computing velocity is faster.

Step 7: abandon the powered plan of electrical network under aqueous mode.Stochastic generation 1st ~ G-1 electrical network is subject to electric process, as the initial position of particle;

Abandoning under aqueous mode, power station all periods are completely sent out, and process of exerting oneself is certain.For playing the peaking capacity benefits in power station as far as possible, make each electrical network residue load after the peak clipping of power station smooth, minimum for the powered plan of target making electrical network to remain load variance.Respective objects function representation is:

$\left\{\begin{array}{c}\min E^g=\frac{1}{T}\underset{t}{\overset{T}{\mathrm{\&Sigma;}}}{w}^g{(D_t^g-\stackrel{\&OverBar;}{D^g})}^2\\ \begin{array}{cc}\stackrel{\&OverBar;}{D^g}=\frac{1}{T}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{D}_t^g& D_t^g=C_t^g-P_t^g\end{array}\end{array}\right.g\&Element;[1,G]---\left(10\right)$

Wherein, E ^grepresent g electrical network residue load variance, G is total electrical network quantity; represent the residue load of t period g electrical network; represent g electrical network prediction load; represent that the power station t period supplies exerting oneself of g electrical network; w ^grepresent the peak regulation target weight coefficient of g electrical network.This weight coefficient not only can reflect the peak regulation preference of dispatcher, namely more lays particular emphasis on and carries out peak regulation to certain electrical network, also can eliminate part because the deviation of the optimization solution that power transmission ratio is different or each peak load regulation network magnitude difference is brought.Therefore, weight coefficient value is as follows:

$w^g=\frac{R^g-{\mathrm{\&alpha;}}^g}{{\mathrm{\&lambda;}}^g}---\left(11\right)$

Wherein, α ^gfor artificial slack variable, generally get 1, also by arranging the size of variable, peak regulation preference between artificial adjustment electrical network; R ^gfor a point electricity ratio; λ ^gfor each electrical network peak-valley difference accounts for the ratio of peak-valley difference sum, computing formula is as follows:

${\mathrm{\&lambda;}}^g=\frac{C_{\max }^g-C_{\min }^g}{\underset{i=1}{\overset{G}{\mathrm{\&Sigma;}}}(C_{\max }^i-C_{\min }^i)}---\left(12\right)$

Wherein, represent maximal value and the minimum value of the prediction load of g electrical network respectively, represent maximal value and the minimum value of the prediction load of No. i-th electrical network respectively.According to point electricity than requiring, be subject to electricity according to fixed electrical network, above-mentioned objective function comprises T × G decision variable abandoning to be under aqueous mode simple quadratic function, in embodiments of the present invention use classical particle group's algorithm (PSO) solve.Following constraint condition (13) and (14) need be met during classical particle group Algorithm for Solving:

Many electrical networks period total powered Constraints of Equilibrium:

$\underset{g=1}{\overset{G}{\mathrm{\&Sigma;}}}{P}_t^g=P^{\max }---\left(13\right)$

Single electrical network divides electricity than Constraint:

$\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_t^g=E_0\&times;R^g,g=\mathrm{1,2},...,G---\left(14\right)$

Wherein, particle initial position is determined as follows:

$P_t^g=\left\{\begin{array}{cc}r\&times;p^{\max }& \mathrm{if}(g=1)\\ r\&times;(p^{\max }-\underset{i=1}{\overset{g-1}{\mathrm{\&Sigma;}}}{P}_t^i)& \mathrm{if}(g>1\mathrm{andg}<G)\\ p^{\max }-\underset{i=1}{\overset{G-1}{\mathrm{\&Sigma;}}}{P}_t^i& \mathrm{if}(g=G)\end{array}\right.---\left(15\right)$

Wherein, r is the random number of 0 ~ 1.In searching process by electric process likely violate electrical network divide electricity than constraint, if do not meet point electricity than constraint, then calculate each electrical network by following formula (16) should adjust by electricity, and revise by electric process with the eigenvalue search method described in above-mentioned steps 3:

${\mathrm{\&Delta;E}}^g=R^g\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_t-\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_t^g---\left(16\right)$

After electric process correction, with the constraint of method process minimum duration described in above-mentioned steps (4-1), and classical particle group algorithm is used to carry out optimizing; When reaching algorithm maximum iteration time, calculating and stopping, using population globally optimal solution as the powered planned outcome of each electrical network.

The present invention with ERTAN Hydroelectric ProJect to Sichuan Electric Power Network and Chongqing electricity grid power transmission for embodiment, according to the power station short-term peak regulation eigenvalue search method flow towards many electrical networks shown in Fig. 1, carry out the simulation of many electrical networks short-term peak regulation, to embody the effect that the present invention reaches.

ERTAN Hydroelectric ProJect is positioned at Yalongjiang River downstream, total installation of generating capacity 3,300,000 kilowatts, normal pool level 1200 meters, total reservoir storage 58 billion cubic meter, be responsible for the peak-frequency regulation task of two provincial power networks in Sichuan and Chongqing, its Short-term Optimal Operation is a typical power station short-term peaking problem towards many electrical networks.In embodiments of the present invention, assuming that the first water level of ERTAN Hydroelectric ProJect is 1182.32 meters, end water level is 1181.82 meters, luffing of exerting oneself is constrained to 1,400,000 kilowatts, a point electricity ratio to Sichuan Electric Power Network and Chongqing electricity grid power transmission is 7:3, the all units in power station are without maintenance, and the power station time interval of exerting oneself between twice change is not less than 4 periods (1 hour).Embodiment is that ERTAN Hydroelectric ProJect is non-abandons water phase power generation dispatching in schedule periods simulation with day, obtains Sichuan Electric Power Network and Chongqing electricity grid by electric process, and is analyzed scheduling result.The invention process concrete steps are as follows:

Step 1: according to power station schedule periods water volume that can be utilized, calculates its average letdown flow, schedule periods generating capacity and maximum output capacity.Judge whether power station abandons water, if do not abandon water, go to step 2, otherwise go to step 5;

Step 2: determine each electrical network initially by electric process by successively cutting load method, power station gross capability process is determined after superposition, judge whether it meets restriction of exerting oneself, if do not meet, add up and be used as not meeting the part of restriction of exerting oneself adjustment step-length of exerting oneself, and gone to step 3 use eigenvalue search method corrections and exert oneself;

Step 3: use eigenvalue search method to adjust by electric process each electrical network;

Step 4: process power station exert oneself luffing constraint and exert oneself minimum duration constraint;

Step 5: " determining water with electricity " calculates water level process, judges whether last water level meets and controls last restriction of water level, if do not meet, then go to step 3; If meet, then calculate end, power station must be arrived and exert oneself and the powered result of each electrical network;

Step 6: the difference according to last water level control value and the last water level of calculating determines adjustment step-length of exerting oneself, and uses the correction of eigenvalue search method to exert oneself;

Step 7: use classical particle group algorithm, calculates each electrical network under abandoning aqueous mode and, by electric process, judges whether that meeting electrical network divides electricity than constraint, if do not meet, then use each electrical network of eigenvalue search method correction by electric process; If meet, calculate end, obtain the powered result of each electrical network.

The results are shown in Figure 2,3,4 after the invention process.Figure 2 shows that the powered procedure chart of embodiment of the present invention Sichuan Electric Power Network, Figure 3 shows that the powered procedure chart of embodiment of the present invention Chongqing electricity grid.As the display of Fig. 2 and Fig. 3 result, ERTAN Hydroelectric ProJect operation near this maintenance bond of load paddy segment base is exerted oneself; Then exert oneself at load crest segment and significantly increase, take full advantage of the peak in power station.Result before and after contrast peak regulation, Sichuan Electric Power Network and Chongqing electricity grid peak-valley difference reduce 40% and 24.8% respectively, remaining lotus variance reduces by 56.0% and 41.3% respectively.The above results shows, the present invention can take into account part throttle characteristics and the peak regulation demand of multiple powered electrical network, makes each electrical network all obtain significant peak regulation effect.

Figure 4 shows that embodiment of the present invention ERTAN Hydroelectric ProJect letdown flow and water level process figure, owing to considering downstream comprehensive water-using requirement, let out under all the period of time power station and be all greater than minimum discharging flow, meet power station actual motion requirement; Meanwhile, ERTAN Hydroelectric ProJect scheduling end of term water level reaches control overflow water level, effectively ensure that the periodicity of scheduling.In addition, in the whole schedule periods of ERTAN Hydroelectric ProJect, Sichuan and Chongqing electricity grid are 6.99:3.01 by electricity allocation proportion, more close than 7:3 ten points with set point electricity, meet the demand of many grid transmissions, therefore, the power generation dispatching process that the present invention obtains meets power station actual motion requirement, has certain engineering practicability.

Those skilled in the art will readily understand; the foregoing is only preferred embodiment of the present invention; not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.

Claims (9)

1., towards a power station short-term peak regulation eigenvalue search method for many electrical networks, it is characterized in that, comprising:

Step 1 calculates schedule periods water volume that can be utilized, and estimates the maximum output capacity of power station schedule periods, judges whether power station abandons water according to average letdown flow, has and then performs step 7, otherwise performs step 2;

The each electrical network of step 2 with described maximum output capacity for the upper limit, obtain many power curves by the respective prediction load curve successively cutting load method that is used alone and add up, obtain power station initially to exert oneself process, and judge whether described process of initially exerting oneself meets period minimax units limits, if do not meet, the part not meeting constraint is added up and has been used as the adjustment step delta E that exerts oneself, then perform step 3, otherwise perform step 4;

Step 3 when hydropower station is not enough, then performs sub-step (3-1); When hydropower station is too much, then perform sub-step (3-2):

(3-1) exert oneself by described adjustment step delta E increase of exerting oneself, divide electricity to compare R by each electrical network the described adjustment step delta E that exerts oneself ^gbe divided into G part: Δ E ^g=R ^gΔ E, wherein, Δ E ^grepresent and need the adjustment step-length of exerting oneself being assigned to g electrical network, g ∈ [1, G], and find the maximum period of lotus more than each electrical network, and by Δ E ^gbe dispensed to this period; If affined restriction, the electricity of increase can not be dissolved by the single period completely, then the lotus maximum period more than being dispensed to by dump energy outside the adjusted period, repeat this process until Δ E ^gjoin complete, recalculate power station gross capability, perform step 4;

(3-2) reduce to exert oneself by the described adjustment step delta E that exerts oneself, by the method for described sub-step (3-1), under the prerequisite meeting firm output powcr constraint, travel through each powered electrical network one by one, the period reduction finding remaining lotus minimum is exerted oneself, until Δ E is assigned; Recalculate power station gross capability, perform step 4;

Step 4 process power station exert oneself minimum duration constraint and exert oneself luffing constraint;

Step 5 to be exerted oneself process computation power station day part water level and letdown flow process according to power station, and the period of violating letdown flow and operating water level constraint is processed, judge that the scheduling end of term calculates water level and whether meets the last restriction of water level of control, if the last water level of described calculating is not equal to the last water level of described control, then perform step 6; Otherwise must arrive power station to exert oneself and the powered result of each electrical network, flow process terminates;

If the last water level of the described calculating of step 6 is greater than the last water level of described control, then hydropower station is not enough, and by the described adjustment step delta E that exerts oneself, increase is exerted oneself, and goes to and performs described step 3; If the last water level of described calculating is less than the last water level of described control, then hydropower station is excessive, reduces to exert oneself, go to and perform described step 3 by the described adjustment step delta E that exerts oneself;

The each electrical network of step 7 stochastic generation is by electric process, as the initial position of particle, the constraint of process minimum duration, and by classical particle group algorithm, to remain, load variance is minimum carries out optimizing for target, if particle position do not meet electrical network divide electricity than constraint, then by the method in described step 3 to revising by electric process, when algorithm reaches maximum iteration time, calculate and stop, using population globally optimal solution as the powered planned outcome of each electrical network.

2. the method for claim 1, is characterized in that, in described step 1 according to upper water power station letdown flow, local inflow forecast and next day water level to disappear plan, calculate schedule periods water volume that can be utilized described in power station by following formula:

$W=\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{B}_t\mathrm{\&Delta;t}+\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}\underset{r\&Element;\mathrm{\&Phi;}}{\mathrm{\&Sigma;}}{Q}_{r-1,t-{\mathrm{\&tau;}}_r}\mathrm{\&Delta;t}+V_0-V_T$

Wherein, hop count when T represents that in a day, schedule periods is total; T ∈ [1, T] represents scheduling slot number; B _trepresent the local inflow of power station in the t period; Δ t represents that scheduling slot is long; φ represents that upstream and current calculating power station have the power station of direct hydraulic connection to gather; Q _{r-1, t-τ}for r power station, upstream is at the letdown flow of t period; τ _rrepresent the current time lag between r power station, upstream and current calculating power station; V ₀with V _trepresent power station schedule periods just last storage capacity respectively.

3. the method for claim 1, is characterized in that, exert oneself described in being calculated as follows in described step 2 adjustment step delta E:

$\mathrm{\&Delta;E}=\underset{t,P_t>p^{\max }}{\overset{T}{\mathrm{\&Sigma;}}}(P_t-P^{\max })+\underset{t,P_t<p^{\min }}{\overset{T}{\mathrm{\&Sigma;}}}(P_t-P^{\min })$

Wherein, hop count when T represents that in a day, schedule periods is total; T ∈ [1, T] represents scheduling slot number; P _trepresent that the t period exerts oneself; P ^maxrepresent period maximum output; p ^minrepresent period minimum load.

4. the method according to any one of claim 1-3, it is characterized in that, exert oneself described in process in described step 4 minimum duration constraint time using the mean value of exerting oneself in minimum duration Δ t as going out force value in described minimum duration Δ t, constant with firm output powcr total amount while meeting minimum duration constraint.

5. the method according to any one of claim 1-3, is characterized in that, exert oneself described in process in described step 4 luffing constraint time judge that adjacent two periods exert oneself P _t-1and P _twhether meet the luffing constraint of exerting oneself be shown below:

|P _t-1-P _t|≤ΔP

Wherein, Δ P allows maximum output luffing the power station period;

When luffing of exerting oneself described in not meeting retrains:

If P _t-1> P _t, then by P _tincrease to P _t-1-Δ P, performs described step 5;

If P _t-1< P _t, then by P _tbe reduced to P _t-1+ Δ P, performs described step 5;

When luffing of exerting oneself described in meeting retrains, perform described step 5.

6. the method according to any one of claim 1-3, is characterized in that, the constraint condition in described step 5 comprises:

Operating water level retrains:

$Z_t^{\min }\&le;Z_t\&le;Z_t^{\max }$

Letdown flow retrains:

$Q_t^{\min }\&le;Q_t\&le;Q_t^{\max }$

Period water level/letdown flow luffing constraint:

$\left\{\begin{array}{c}|Z_t-Z_{t-1}|\&le;\mathrm{\&Delta;Z}\\ |Q_t-Q_{t-1}|\&le;\mathrm{\&Delta;Q}\end{array}\right.$

Wherein, represent power station t period water level up-and-down boundary respectively; represent power station t period letdown flow up-and-down boundary respectively; Δ Z, Δ Q represent that the power station period allows maximum stage luffing and letdown flow luffing respectively.

7. the method according to any one of claim 1-3, is characterized in that, adjustment step delta E computing method of exerting oneself described in described step 6 are as follows:

ΔE＝a(Z _c-Z _g)+b

Wherein, Z _crepresent the last water level of described calculating; Z _grepresent the last water level of described control; A and b represents step-length regulation coefficient respectively.

8. the method according to any one of claim 1-3, is characterized in that, in described step 7, the initial position of particle is determined by following formula:

${P_t}^g=\left\{\begin{array}{cc}r\&times;p^{\max }& \mathrm{if}(g=1)\\ r\&times;(p^{\max }-\underset{i=1}{\overset{g-1}{\mathrm{\&Sigma;}}}{P_t}^i)& \mathrm{if}(g>1\mathrm{andg}<G)\\ p^{\max }-\underset{i=1}{\overset{G-1}{\mathrm{\&Sigma;}}}{P_t}^i& \mathrm{if}(g=G)\end{array}\right.$

Adjusting each electrical network by the adjustment step size computation formula of exerting oneself of electricity is:

$\mathrm{\&Delta;}{E}^g=R^g\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_t-\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P_t}^g$

Wherein, r is the random number of 0 ~ 1; P _trepresent that the t period exerts oneself; P ^maxrepresent period maximum output.

9. the method according to any one of claim 1-3, is characterized in that, be the powered plan of target making electrical network to remain load variance minimum in described step 7, respective objects function representation is:

$\left\{\begin{array}{cc}\min E^g=\frac{1}{T}\underset{t}{\overset{T}{\mathrm{\&Sigma;}}}{w}^g{(D_t^g-\stackrel{\&OverBar;}{D^g})}^2& \\ \stackrel{\&OverBar;}{D^g}=\frac{1}{T}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{D}_t^g& D_t^g=C_t^g-{P_t}^g\end{array}\right.g\&Element;[1,G]$

Wherein, E ^grepresent g electrical network residue load variance, G is total electrical network quantity; represent the residue load of t period g electrical network; represent g electrical network prediction load; P _t ^grepresent that the power station t period supplies exerting oneself of g electrical network; w ^grepresent the peak regulation target weight coefficient of g electrical network, its value is as follows:

$w^g=\frac{R^g\&CenterDot;{\mathrm{\&alpha;}}^g}{{\mathrm{\&lambda;}}^g}$

Wherein, α ^grepresent artificial slack variable; λ ^grepresent that each electrical network peak-valley difference accounts for the ratio of peak-valley difference sum, its computing formula is as follows:

${\mathrm{\&lambda;}}^g=\frac{C_{\max }^g-C_{\min }^g}{\underset{i=1}{\overset{G}{\mathrm{\&Sigma;}}}(C_{\max }^i-C_{\min }^i)}$

Wherein, represent maximal value and the minimum value of the prediction load of g electrical network respectively; represent maximal value and the minimum value of the prediction load of No. i-th electrical network respectively.