CN105528668A - Dynamic environment and economy scheduling method of grid-connected wind power system - Google Patents

Abstract

The invention discloses a dynamic environment and economy scheduling method of a grid-connected wind power system. The grid-connected wind power system which comprises a wind farm, more than one thermal power generating set and a scheduling center is arranged, the scheduling center comprises a scheduling control system and a scheduling optimization center, the scheduling control system collects the required supply quantity of the system, and the supply quantity equals the sum of individual users, the medium load and the heavy load; and the scheduling control system also collects related data of the wind farm and the thermal power generating sets, and transmits the data to a scheduling decision center, the scheduling decision center obtains an optimization result of the related data according to a data model and back feeds the result to the scheduling control system, the scheduling control system controls operation of the wind farm and the thermal generating sets, and the scheduling decision center arranges the model and makes data interaction with the scheduling control system.

Description

A kind of wind-electricity integration Electrical Power System Dynamic environmental economy dispatching method

Technical field

The present invention relates to electric system, particularly relate to a kind of wind-electricity integration electric power system dispatching method.

Background technology

Along with the quick growth of China's economy, present the problems such as energy supply anxiety, coal transport power deficiency, the deterioration of the ecological environment gradually.By strengthening competitively priced and mandatory market share policy, the ratio of new forms of energy in China's overall energy structure progressively improves.Wind energy as a kind of continually, never exhausted green energy resource, oneself is through becoming the important substitute energy of fossil energy.Compared with generating electricity with fired power generating unit, wind-powered electricity generation can bring the income on more, more long-range environment and economy for the management and running of electric system.But different from conventional Power Generation Mode, the number of output of wind electric field depends on the size of wind speed at that time, and this is closely related with objective factors such as meteorology, geographical environments again, causes uncertainty and the intermittence of wind power output thus.Although forecasting wind speed technology is at development, far away not as good as the prediction effect of load.

Wind-electricity integration electric system, it comprises some fired power generating unit and wind energy turbine set, and institute's electric energy that obtains is for daily or industrial requirement, and dispatching center controls the stop and start of fired power generating unit in real time by prediction.Dynamic environment economic load dispatching namely meeting workload demand and run constraint prerequisite under, reasonably to distribute in the whole cycle in electrical network the optimum of fired power generating unit in day part and exert oneself, make cost of electricity-generating and pollutant emission in dispatching cycle little as far as possible.The uncertainty that wind-powered electricity generation self is difficult to eliminate must affect to the safe and stable operation of electric system, wind-electricity integration Electrical Power System Dynamic environmental economy scheduling problem also becomes more complicated, solves the corresponding increase of difficulty, and existing scheduling model and algorithm exist model error, solve the deficiencies such as difficulty.Therefore, analyze the impact that wind-power electricity generation runs for electric power system dispatching, propose modeling that is new or that improve and method for solving, how research carries out electric power system dispatching efficiently, adapts to low-carbon economy " new normality " significant for promotion China electrical production scheduling theory.

When wind energy turbine set access electrical network after, electric power system dispatching personnel institute faced by problem be not only the output power of how to distribute fired power generating unit, also to take into account simultaneously arrangement Wind turbines the plan of exerting oneself.In order to ensure that system can safe and reliable operation, enough margin capacities must be reserved to make up contingent generated energy surplus or deficiency in scheduling process.The additional spinning reserve capacity caused by wind-powered electricity generation uncertainty can be considered in Optimized model, and using the reference value of the certain percentage of wind-powered electricity generation total amount as selected additional standby requirement.This modeling method dealing with wind-powered electricity generation random variation characteristic by increasing margin capacity, the effect of safeguards system safe and reliable operation can be played to a certain extent, but owing to not giving statistical study to the uncertain characteristic of wind-powered electricity generation, only have the spinning reserve of the wind-powered electricity generation capacity level of reserved 100% can guarantee that system is perfectly safe.The optimum results that this deterministic analytical approach obtains will be too conservative, and again because wind power can not remain near power-handling capability always, this will cause the waste of margin capacity, and then indirectly adds the cost of electricity-generating of economic load dispatching.

In order to overcome above deficiency, the emulation mode based on random probability distribution is arisen at the historic moment.The Wind speed model based on the Weibull regularity of distribution is adopted for the economic load dispatching under chronic mode, by wind speed--powertrace is converted to corresponding wind power probability distribution function, in put forward scheduling model, introduces the risk that stand-by cost and punishment cost over-evaluate wind power respectively and underestimate simultaneously take in the objective function of optimization problem.But Weibull distribution is generally used for the probability distribution situation describing annual mean wind speed, so this type of distributes and is not suitable for arranging short-term electricity generation scheduling problem.Different from the scheduling problem under chronic mode, then need under short term scheduling environment to utilize wind speed/wind power prediction technology to estimate the reference value of following concrete moment wind power output.Due to randomness and the intermittence of wind energy itself, comparatively load prediction is more difficult to make wind power prediction, is difficult to obtain result comparatively accurately as load prediction.Therefore, be necessary to analyze wind speed/wind power prediction error.For the uncertainty of outburst prediction, common way sets up probability Distribution Model to predicated error.According to this thinking, the basis of predicting wind speed adopts Normal Distribution density function to describe forecasting wind speed error, thus make wind-powered electricity generation probability model more close to actual schedule situation.

Summary of the invention

In view of the above-mentioned problem that need solve, the present invention take System Thinking as thought guidance, on the basis of deep anatomy environmental economy scheduling connotative definition, break through from method the limitation that traditional environment economic load dispatching model can not portray enchancement factor, decision-making guard and radical this to contradiction between seek the path optimizing of Wind Power Development.Specifically, the randomness of the present invention's outstanding wind-powered electricity generation in wind-electricity integration electric system, take into account the uncertainty of load prediction and fired power generating unit stoppage in transit, discuss under certain reliability, validity level control, and under conventional constraint, relation between wind power output plan, the thermoelectricity plan of exerting oneself, positive and negative spinning reserve three, probes into the equilibrium point of their coordinated developments.By Evolution Paths and the interpretation of result of policy different under multiple sight, the policy that realizes is preferred, provides scientific approach and theory support for decision maker formulates power scheduling policy, and provides simulation tool for the validity of inspection or the policy of evaluation.

In order to reach above-mentioned goal of the invention, first the present invention constructs a class wind-electricity integration Electrical Power System Dynamic environmental economy scheduling model, be embodied in following form, as Fig. 1, wind-electricity integration electric system is set, comprises wind energy turbine set, more than one fired power generating unit, and dispatching center, dispatching center comprises Dispatching Control System and optimizing scheduling center, supply needed for Dispatching Control System acquisition system, and described supply comprises personal user, medium load and heavy load sum; Described Dispatching Control System also gathers wind energy turbine set and fired power generating unit related data, be transferred to scheduling decision center, scheduling decision center obtains after above-mentioned related data draws optimum results according to data model and feeds back to Dispatching Control System, Dispatching Control System controls the operation of wind energy turbine set and fired power generating unit, and data model comprises:

Objective function:

1) cost of electricity-generating

Electric system cost of electricity-generating minimizes and can be expressed as:

${\mathrm{minF}}_1=min\underset{t=1}{\overset{T}{\Σ}}\underset{i=1}{\overset{M}{\Σ}}{C}_i\left(T_{i,t}\right)---\left(1\right)$

In formula: M is fired power generating unit number; T is time interval number; C _i() is the cost function of i-th fired power generating unit; T _i,tthat the plan of i-th fired power generating unit in period t is exerted oneself.

Specifically, consider the fired power generating unit cost of electricity-generating of valve point effect, the form of quadratic function and sine function sum can be shown as:

$C_i\left(T_{i,t}\right)=a_i{T}_{i,t}^2+b_i{T}_{i,t}+d_i+\left|e_{i}sin\left(f_i\right(T_i^{\min }-T_{i,t})\right)|,1\≤i\≤M,1\≤t\≤T---\left(2\right)$

In formula: a _i, b _i, d _i, e _iand f _iit is the cost of electricity-generating coefficient of i-th fired power generating unit; T _i ^minit is the meritorious lower limit of exerting oneself of i-th fired power generating unit.

2) pollutant emission

In power generation process, fired power generating unit more or less can give off the pollutants such as a large amount of oxysulfide, oxides of nitrogen, and be similar to cost of electricity-generating functional expression (2), pollutant emission minimizes and can be expressed as:

${\mathrm{minF}}_2=min\underset{t=1}{\overset{T}{\Σ}}\underset{i=1}{\overset{M}{\Σ}}{E}_i\left(T_{i,t}\right)---\left(3\right)$

In formula: E _i() is the pollutant emission function of i-th fired power generating unit.

Specifically, pollutant discharge amount and fired power generating unit gain merit and exert oneself between funtcional relationship, the form of quadratic function and exponential function sum can be shown as:

$E_i\left(T_{i,t}\right)={\mathrm{\α}}_i{T}_{i,t}^2+{\mathrm{\γT}}_{i,t}+{\mathrm{\λ}}_i+{\mathrm{\δ}}_i\exp \left({\mathrm{\τ}}_i{T}_{i,t}\right),1\≤i\≤M,1\≤t\≤T---\left(4\right)$

In formula: α _i, γ _i, λ _i, δ _iand τ _iit is the pollutant discharge coefficient of i-th fired power generating unit.

Constraint condition:

1) the power-balance constraint of power outages is considered

$\mathrm{Pr}\{\underset{i=1}{\overset{M}{\Σ}}{T}_{i,t}+W_t\≤P_{D,t}+P_{L,t}\}\≤{\mathrm{\η}}_1,1\≤t\≤T---\left(5\right)$

In formula: W _tthat the meritorious of wind energy turbine set in period t is exerted oneself; P _d,tl _t,sit is system load demand in period t; P _l,tit is the power outages in period t; η ₁it is the confidence level meeting workload demand.

Consider that circuit capacity retrains, DC power flow method will be used for calculating the poower flow of each bar circuit.Based on the loss equation of Kron, the loss in formula (5) can be expressed as within each time period:

$P_{L,t}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{M}{\Σ}}{T}_{i,t}{B}_{ij}{T}_{j,t}+\underset{i=1}{\overset{M}{\Σ}}{B}_{i0}{T}_{i,t}+B_{00},1\≤t\≤T---\left(6\right)$

In formula: B _ij, B _i0and B ₀₀it is the loss factor in electric power networks transmission power loss matrix B.

2) unit operation constraint

T _i ^min≤T _i,t≤T _i ^max,1≤t≤T(7)

In formula: T _i ^maxit is the meritorious upper limit of exerting oneself of i-th fired power generating unit.

3) unit climbing rate constraint

T _i,t-T _i,t-1≤UR _i·T ₆₀,1≤t≤T(8)

T _i,t-1-T _i,t≤DR _i·T ₆₀,1≤t≤T(9)

In formula: UR _i, DR _irepresent the upper and lower climbing rate restriction of i-th fired power generating unit respectively; Δ t is the time interval of each period.

4) spinning reserve constraint

In typical power system, margin capacity can be used for disposal system load prediction error and fired power generating unit and to stop transport the impact brought.But for wind-electricity integration electric system, because wind power output predicated error is far longer than system loading predicated error, the uncertainty of system obviously increases, be necessary to introduce positive and negative margin capacity to reduce adverse effect.When wind-electricity integration scale is given, wind power output predicted value is larger, once lose or reduce this part electricity, other units must improve output power with the disappearance of quick response system power, and namely system need provide more positive rotation margin capacity to deal with the impact underestimated wind power output and cause.On the other hand, in order to maintain the active balance of system, the increasable amount of wind power output is larger, and the power that fired power generating unit need be forced down is more, and the demand that system bears spinning reserve is also higher.When wind energy turbine set reaches maximum output, the rising space becomes 0, and namely electric system is without the need to providing negative spinning reserve capacity.After large-scale wind power is grid-connected, bringing the interference of electric power system dispatching to reduce wind power output randomness, should consider that positive and negative spinning reserve capacity retrains simultaneously.Wherein, positive rotation Reserve Constraint can be expressed as:

$\mathrm{Pr}\{{\mathrm{SSR}}_t^u\≤P_{D,t}\·\mathrm{\α}+W_t\·{\mathrm{\β}}_u\}\≤{\mathrm{\η}}_2,1\≤t\≤T---\left(10\right)$

In formula: α is the demand that system loading predicated error aligns spinning reserve; β _ufor wind power output predicated error aligns the demand of spinning reserve; η ₂it is the confidence level meeting positive rotation standby requirement.Wherein, positive rotation is for subsequent use specifically can be expressed as:

${\mathrm{SSR}}_t^u=\underset{i=1}{\overset{M}{\Σ}}{\mathrm{SR}}_{i,t}^u=\underset{i=1}{\overset{M}{\Σ}}\min \{T_i^{\max }-T_{i,t},\frac{{\mathrm{UR}}_i}{6}\},1\≤t\≤T---\left(11\right)$

The present invention suppose positive rotation that fired power generating unit provides for subsequent use with 10 minutes for measurement unit, and its size retrains relevant with the climbing rate of fired power generating unit.From period t to period t+1, the ratio of slope upper limit of climbing of i-th unit is UR _i.Correspondingly, in 10 minutes, the corresponding ratio of slope upper limit of climbing just becomes

Similarly, negative rotation turns Reserve Constraint and can be expressed as:

$\mathrm{Pr}\{{\mathrm{SSR}}_t^d\≤(W_{max}-W_t)\·{\mathrm{\β}}_d\}\≤{\mathrm{\η}}_3,1\≤t\≤T---\left(12\right)$

In formula: W _maxfor the rated power of wind energy turbine set; β _dfor wind power output predicated error is to the demand of negative spinning reserve; η ₃meet the confidence level that negative rotation turns standby requirement.Wherein, negative spinning reserve specifically can be expressed as:

${\mathrm{SSR}}_t^d=\underset{i=1}{\overset{M}{\Σ}}{\mathrm{SR}}_{i,t}^d=\underset{i=1}{\overset{M}{\Σ}}min\{T_{i,t}-T_i^{\min },\frac{{\mathrm{DR}}_i}{6}\},1\≤t\≤T---\left(13\right)$

The wind-electricity integration electric power system dispatching optimization problem that the present invention studies is containing stochastic variable in constraint condition, and does not comprise stochastic variable in objective function.For this characteristic, on the basis about wind power output statistical nature, Chance-Constrained Programming Model of the present invention can be converted into deterministic optimization model.Specifically, constraint condition Chinese style (5) can be converted into:

$\begin{array}{cc}\begin{array}{c}\mathrm{Pr}\{\underset{i=1}{\overset{M}{\Σ}}{T}_{i,t}+W_t\≤P_{D,t}+P_{L,t}\}\\ =\mathrm{Pr}\{W_t\≤P_{D,t}+P_{L,t}-\underset{i=1}{\overset{M}{\Σ}}{T}_{i,t}\}\\ =F(P_{D,t}+P_{L,t}-\underset{i=1}{\overset{M}{\Σ}}{T}_{i,t})\≤{\mathrm{\η}}_1\end{array}& 1\≤t\≤T\end{array}---\left(14\right)$

In order to represent the distribution function F () of wind power output, its probability density function function f () is summarized as:

$f_W\left(\mathrm{\&rho;}\right)=\mathrm{\&theta;}\frac{{\mathrm{klv}}_i}{{\mathrm{cw}}_{rated}}{\left(\frac{(1+\mathrm{\&rho;}l)v_i}{c}\right)}^{k-1}\exp (-{\left(\frac{(1+\mathrm{\&rho;}l)v_i}{c}\right)}^k)+(1+\mathrm{\&theta;})\frac{{\mathrm{lv}}_i}{w_{rated}{\mathrm{\&beta;}}^{\mathrm{\&xi;}}\mathrm{\&Gamma;}\left(\mathrm{\&zeta;}\right)}{\left(\right(1+\mathrm{\&rho;}l\left)v_i\right)}^{\mathrm{\&xi;}-1}\exp (-\frac{(1+\mathrm{\&rho;}l)v_i}{\mathrm{\&beta;}}),0<\mathrm{\&rho;}<1---\left(15\right)$

Wherein:

$\mathrm{\&rho;}=\frac{w}{w_{rated}},l=\frac{{\mathrm{\&nu;}}_r-v_i}{v_i}---\left(16\right)$

In formula: v _iit is incision wind speed; v _oit is cut-out wind speed; v _rit is wind rating; w _ratedthat rated power exports; K is the form parameter of Weibull distribution; C is the scale parameter of Weibull distribution; ζ is the form parameter of Gamma distribution; β is the scale parameter of Gamma distribution; Γ () is Gamma function; θ is the weight parameter of Hybrid IC and Weibull distribution gained GW distribution.About the mixing probability density function curve of wind power output is as Fig. 2.

The distribution function F () of wind power output can be expressed as:

$F\left(w\right)=\left\{\begin{array}{cc}0& w<0\\ \mathrm{\&theta;}\left(\exp \right(-{(-\frac{b}{ac})}^k)-\exp (-{(-\frac{w-b}{ac})}^k))& \\ +(1-\mathrm{\&theta;})\left(\frac{\mathrm{\&gamma;}(\mathrm{\&xi;},-(b/a\mathrm{\&beta;}))-\mathrm{\&gamma;}(\mathrm{\&xi;},(w-b)/a\mathrm{\&beta;})}{\mathrm{\&Gamma;}\left(\mathrm{\&xi;}\right)}\right)& 0\&le;w<w_{rated}\\ 1& w\&GreaterEqual;w_{rated}\end{array}\right.---\left(17\right)$

In addition, wind power output in the probable value of discrete portions is:

Pr{W＝0}＝F _v(v _i)+(1-F _v(v _o))(18)

Pr{W＝w _rated}＝F _v(v _o)-F _v(v _r)(19)

Similarly, formula (10), (12) also can transform following deterministic form respectively:

$\begin{array}{cc}\begin{array}{c}\mathrm{Pr}\{{\mathrm{SSR}}_t^u\&le;P_{D,t}\&CenterDot;\mathrm{\&alpha;}+W_t\&CenterDot;{\mathrm{\&beta;}}_u\}\\ =\mathrm{Pr}\{W_t\&GreaterEqual;\frac{1}{{\mathrm{\&beta;}}_u}({\mathrm{SSR}}_t^u-P_{D,t}\&CenterDot;\mathrm{\&alpha;})\}\\ =1-F\left(\frac{1}{{\mathrm{\&beta;}}_u}\right({\mathrm{SSR}}_t^u-P_{D,t}\&CenterDot;\mathrm{\&alpha;}))\&le;{\mathrm{\&eta;}}_2\end{array}& 1\&le;t\&le;T\end{array}---\left(20\right)$

$\begin{array}{cc}\begin{array}{c}\mathrm{Pr}\{{\mathrm{SSR}}_t^d\&le;(W_{\max }-W_t)\&CenterDot;{\mathrm{\&beta;}}_d\}\\ =\mathrm{Pr}(W_t\&le;W_{\max }-\frac{{\mathrm{SSR}}_t^d}{{\mathrm{\&beta;}}_d})\\ =F(W_{\max }-\frac{{\mathrm{SSR}}_t^d}{{\mathrm{\&beta;}}_d})\&le;{\mathrm{\&eta;}}_3\end{array}& 1\&le;t\&le;T\end{array}---\left(21\right)$

Based on above change, the Chance-Constrained Programming Model about the scheduling of wind-electricity integration Electrical Power System Dynamic environmental economy finally can be converted into the form of following deterministic optimization model:

$\left\{\begin{array}{c}\min \&lsqb;F_1\left(T\right),F_2\left(T\right)\&rsqb;\\ \begin{array}{ccc}s.t.& g_j\left(T\right)\&le;0& j=1,2,\mathrm{...},6.\end{array}\end{array}\right.---\left(22\right)$

In formula: T be fired power generating unit plan exert oneself form decision vector; G (), j=1,2 ..., 6} is the inequality constrain function about decision vector.

The dynamic environment Economic Dispatch Problem that the present invention constructs has stronger binding character, simultaneously containing non-linear component.This makes feasible zone not only narrow and small but also topological structure is complicated, and conventional optimized algorithm efficiency is very low.The crucial part of such problem of Efficient Solution is effectively to process strong constraint.For this reason, the present invention adopts heuristic strategies to revise the solution that NSGA-II algorithm obtains, and makes every effort to the diversity improving colony and the degree of accuracy solved, expands search volume, guarantee to obtain more excellent scheduling solution.

Consider the meritorious T that exerts oneself of each unit in day part _i,t(i=1,2 ..., M; T=1,2 ..., T) be decision variable, adopt real coding mode, by all T _i,tbody is one by one connected into, corresponding expression scheduling scheme according to period and number order.

Because decision variable dimension is n=MT, individual X is corresponding can be expressed as:

X＝[T _1,1,...,T _1,T,...,T _M,1,...,T _M,T](23)

In initialization procedure, T _i,t(i=1,2 ..., M; T=1,2 ..., T) can be obtained by following formula:

$T_{i,t}=T_i^{\min }+\mathrm{\&xi;}\&CenterDot;(T_i^{\max }-T_i^{\min })---\left(24\right)$

In formula: ξ is for obeying (0,1) equally distributed random number.

Under different constraint condition, respectively dynamic constrained process is carried out to the individual X in optimizing process:

1) power-balance constraint dynamic process

In Electrical Power System Dynamic optimizing scheduling process, the plan of thermoelectricity genset in time period t is exerted oneself T _i,tlikely do not meet the power-balance constraint given by formula (5).In order to meet above-mentioned constraint condition, can by the existing scheduling scheme of heuristic strategies correction.Concrete operation step is as follows:

Step 1: time period numbering t=1;

Step 2: calculation interval t internal power Constraints of Equilibrium violation value Δ _t:

${\mathrm{\&Delta;}}_t=F(P_{D,t}+P_{L,t}-\underset{i=1}{\overset{M}{\&Sigma;}}{T}_{i,t})-{\mathrm{\&eta;}}_1---\left(25\right)$

Step 3: if Δ _t≤ 0, then jump to step 5;

Step 4: each fired power generating unit plan is exerted oneself T _i,t(i=1,2 ..., M) adjust:

T _i,t=min{T _i ^max, T _i,t+ Δ _tω _i,t, wherein weight ${\mathrm{\&omega;}}_{i,t}=\frac{T_i^{\max }-T_{i,t}}{\underset{i=1}{\overset{M}{\&Sigma;}}(T_i^{\max }-T_{i,t})}---\left(26\right)$

To exert oneself T to check the plan after adjustment _i,twhether meet constraint condition, jump to step 2;

Step 5: if t<T, so t=t+1 jump to step 2;

Step 6: power-balance constraint dynamic process terminates.

2) unit operation constraint, climbing rate constraint dynamic process

In order to meet unit operation constraint given by formula (7) and formula (8), (9) given unit climbing rate retrain, also can by the existing scheduling scheme of heuristic strategies correction.Specific operation process is as follows:

Step 1: initialization thermoelectricity genset numbering i=1, time period numbering t=1;

Step 2: as t=1, upper is set _i,t=T _i ^max, lower _i,t=T _i ^min; As t>1,

Upper is set _i,t=min{T _{i, t-1}+ UR _i, T _i ^max, lower _i,t=max{T _{i, t-1}-DR _i, T _i ^min(27)

Step 3: the T if fired power generating unit plan is exerted oneself _i,tdo not meet unit operation constraint, the constraint of climbing rate, so revise as follows respectively:

Step 4: if t<T, so t=t+1 jump to step 2;

Step 5: if i<M, so i=i+1 jump to step 2;

Step 6: unit operation constraint, climbing rate constraint dynamic process terminate.

3) spinning reserve constraint dynamic process

In order to meet positive rotation Reserve Constraint condition, equally can by the existing scheduling scheme of heuristic strategies correction.Specific operation process is as follows:

Step 1: time period numbering t=1;

Step 2: positive rotation Reserve Constraint violation value Δ in calculation interval t _t:

${\mathrm{\&Delta;}}_t=F\left(\frac{1}{{\mathrm{\&beta;}}_u}\right({\mathrm{SSR}}_t^u-P_{D,t}\&CenterDot;\mathrm{\&alpha;}))+{\mathrm{\&eta;}}_2-1---\left(29\right)$

Step 3: if Δ _t>=0, then jump to step 5;

Step 4: each fired power generating unit plan is exerted oneself T _i,t(i=1,2 ..., M) adjust:

T _i,t=max{T _i ^min, T _i,t-Δ _tω _i,t, wherein weight ${\mathrm{\&omega;}}_{i,t}=\frac{T_{i,t}-T_i^{\min }}{\underset{i=1}{\overset{M}{\&Sigma;}}(T_{i,t}-T_i^{\min })}---\left(30\right)$

To exert oneself T to check the plan after adjustment _i,twhether meet constraint condition, jump to step 2;

Step 5: if t<T, so t=t+1 jump to step 2;

Step 6: positive rotation Reserve Constraint dynamic process terminates.

Similarly, the amendment scheme that negative rotation turns Reserve Constraint condition can also be obtained:

Step 1: time period numbering t=1;

Step 2: in calculation interval t, negative rotation turns Reserve Constraint violation value Δ _t:

${\mathrm{\&Delta;}}_t=F(W_{max}-\frac{{\mathrm{SSR}}_t^d}{{\mathrm{\&beta;}}_d})-{\mathrm{\&eta;}}_3---\left(31\right)$

Step 3: if Δ _t≤ 0, then jump to step 5;

Step 4: each fired power generating unit plan is exerted oneself T _i,t(i=1,2 ..., M) adjust:

T _i,t=min{T _i ^max, T _i,t+ Δ _tω _i,t, wherein weight ${\mathrm{\&omega;}}_{i,t}=\frac{T_i^{\max }-T_{i,t}}{\underset{i=1}{\overset{M}{\&Sigma;}}(T_i^{\max }-T_{i,t})}---\left(32\right)$

To exert oneself T to check the plan after adjustment _i,twhether meet constraint condition, jump to step 2;

Step 5: if t<T, so t=t+1 jump to step 2;

Step 6: negative rotation turns Reserve Constraint dynamic process and terminates.

Through effective process to strong constraint, with heuristic strategies NSGA-II algorithm flow as shown in Figure 3.Dynamic environment economic load dispatching model belongs to complicated multi-objective constrained optimization problem, is difficult to obtain feasible Pareto efficient solution, must designs suitable decision-making technique according to target signature by existing penalty function method.First the present invention generates certain proportion feasible solution as initial solution, then uses Feasible degree to merge standardized objective function and constraint degree of running counter to generation affinity of antibody, finally adopts repair strategy to obtain the antibody that certain feasible solution generates restructuring.

Based on fuzzy set theory as the scheme extracting optimizing decision.Suppose that obtained Pareto efficient solution set is X={x ₁, x ₂, L, x _k, concrete steps in decision-making is:

Step 1: calculate each Pareto efficient solution x _i∈ X, i=1,2 ..., the linear membership function of each sub-goal of K:

${\mathrm{\&mu;}}_m\left(x_i\right)=\{\begin{array}{cc}1& F_m\left(x_i\right)\&le;F_m^{\min }\left(x_i\right)\\ \frac{F_m^{\max }\left(x_i\right)-F_m\left(x_i\right)}{F_m^{\max }\left(x_i\right)-F_m^{\min }\left(x_i\right)}& F_m^{\min }\left(x_i\right)<F_m\left(x_i\right)<F_m^{\max }\left(x_i\right)\\ 0& F_m\left(x_i\right)\&GreaterEqual;F_m^{\max }\left(x_i\right)\end{array},m=1,2.---\left(33\right)$

Step 2: calculate each Pareto efficient solution x _ithe standardization satisfaction of ∈ X:

$\mathrm{\&eta;}\left(x_i\right)=\frac{\underset{m=1}{\overset{2}{\&Sigma;}}{\mathrm{\&mu;}}_m\left(x_i\right)}{\underset{i=1}{\overset{K}{\&Sigma;}}\underset{m=1}{\overset{2}{\&Sigma;}}{\mathrm{\&mu;}}_m\left(x_i\right)}---\left(34\right)$

Step 3: the Pareto efficient solution of selection standard Maximum Satisfaction is optimum solution.

Accompanying drawing explanation

Fig. 1 wind-electricity integration electric system schematic diagram

Fig. 2 wind power output mixing probability density function

Fig. 3 is with the NSGA-II algorithm flow chart of heuristic strategies

Optimum forward position (the η of Fig. 4 Pareto ₁=η ₂=η ₃=0.1)

Optimum forward position (the η of Fig. 5 Pareto ₁=η ₂=η ₃=0.15)

The cluster result that Fig. 6 Pareto separates

Embodiment

The electric system that this example is formed for 10 fired power generating unit and 1 Large-scale Wind Power place is to simulate dynamic environment Economic Dispatch Problem.In annex, A.2 A.1 form have recorded operational factor and the electric load demand of classical 10 machine test macros respectively with form.In addition, loss also will be considered, in special time interval, suppose that loss factor remains unchanged, and concrete loss matrix is shown in annex.Wind energy turbine set parameter is shown in that in annex, form A.3.By MATLAB7.0, the NSGA-II with heuristic strategies is used to minimize financial cost and pollutant discharge amount.Group size is set to 20, and maximum iteration time is set to 100, and crossing-over rate, aberration rate are set to 0.9,0.2 respectively.

In order to probe into, wind-powered electricity generation uncertain factor is exerted oneself for optimal plan, the impact of financial cost, pollutant discharge amount, the Pareto forward position that Fig. 3, Fig. 4 obtain under compared for different confidence level.Pareto forward position in Fig. 4,5 is evenly distributed and has a very wide distribution, and fully shows the applicability of the NSGA-II with heuristic strategies, thus provides abundant information for meeting each other half way between economic benefit and environmental protection.Along with the raising meeting each constraint condition confidence level, namely the risk level that causes of wind-powered electricity generation uncertainty rises, the corresponding increase of feasible zone of dynamic economic dispatch optimization problem, and the financial cost of wind-electricity integration electric system entirety, pollutant emission is corresponding changes.In order to reflect more intuitively Fig. 4,5 difference, table 1 lists end points place, Pareto forward position minimum economic cost under different confidence level, minimum contamination discharge single object optimization extreme value.

Extreme value under the different confidence level of table 1

According to table 1 analysis result, along with the rising meeting confidence level in constraint condition, fired power generating unit gross capability monotone decreasing in electric system, the corresponding increase of output of wind electric field, minimum financial cost is down to 2326600 $ by 2560400 $, meanwhile, minimum dusty gas discharge capacity reduces to 244860lb by 303280lb.Illustrate thus: although the increase of output of wind electric field proportion is conducive to economy and the feature of environmental protection in electric system, also bring larger risk to system cloud gray model.If dispatcher only considers economic benefit, when the scheduling scheme selecting financial cost minimum, pollutant discharge amount can increase, and is unfavorable for the feature of environmental protection; If only minimum for target with pollutant discharge amount, then to increase cost of electricity-generating for cost, be unfavorable for economy.Therefore need the factor considering each side, fully excavate the information that Pareto optimum solution contains, and then provide aid decision making for electric power system dispatching personnel.

In the aid decision making stage, take confidence level as the Pareto optimum collection of 0.1, first based on clustering methodology, cluster is carried out to Pareto optimum collection.Fig. 6 describes the result 40 Pareto optimum solutions being divided into 3 classes, has stronger meaning directly perceived.Wherein, category-A contains 13 decision vectors, represents the preference of dispatcher to environmental protection; Category-B contains 15 decision vectors, represents dispatcher and takes into account economic benefit and environmental protection, without special preference; C class contains 12 decision vectors, represents the preference of dispatcher to economic benefit.

Application fuzzy set theory obtains recommending compromise to separate, as shown in Figure 3 after sorting to the decision vector that each class Pareto optimum is concentrated.In order to verify the accuracy of solution, table 2,3,4 gives corresponding fired power generating unit plan and exerts oneself.Under specific optimization aim and constraint condition, the optimal plan of different unit to exert oneself within each period fluctuation obviously.In addition, when the environmental protection of dispatcher's preference, the cost of electricity-generating total in 24 hours of whole electric system is 2573700 $, and pollutant emission is 303460lb.When dispatcher takes into account economic benefit and environmental protection, during without special preference, the cost of electricity-generating total in 24 hours of whole electric system is 2567400 $, and pollutant emission is 304040lb.When dispatcher's preference economic benefit, the cost of electricity-generating total in 24 hours of whole electric system is 2561500 $, and pollutant emission is 304990lb.

Table 2 compromise is separated optimal plan in (category-A) corresponding day part and is exerted oneself (MW)

Table 3 compromise is separated optimal plan in (category-B) corresponding day part and is exerted oneself (MW)

Table 4 compromise is separated optimal plan in (C class) corresponding day part and is exerted oneself (MW)

Annex

The operational factor of the classical 10 machine test macros of Table A .1

Table A .2 electric load demand (10 machine system)

Table A .3 wind energy turbine set parameter (10 machine system)

In 10 machine system normal course of operation, loss matrix B is:

$B=\left(\begin{array}{cccccccccc}0.000049& 0.000014& 0.000015& 0.000015& 0.000016& 0.000017& 0.000017& 0.000018& 0.000019& 0.000020\\ 0.000014& 0.000045& 0.000016& 0.000016& 0.000017& 0.000015& 0.000015& 0.000016& 0.000018& 0.000018\\ 0.000015& 0.000016& 0.000039& 0.000010& 0.000012& 0.000012& 0.000014& 0.000014& 0.000016& 0.000016\\ 0.000015& 0.000016& 0.000010& 0.000040& 0.000014& 0.000010& 0.000011& 0.000012& 0.000014& 0.000015\\ 0.000016& 0.000017& 0.000012& 0.000014& 0.000035& 0.000011& 0.000013& 0.000013& 0.000015& 0.000016\\ 0.000017& 0.000015& 0.000012& 0.000010& 0.000011& 0.000036& 0.000012& 0.000012& 0.000014& 0.000015\\ 0.000017& 0.000015& 0.000014& 0.000011& 0.000013& 0.000012& 0.000038& 0.000016& 0.000016& 0.000018\\ 0.000018& 0.000016& 0.000014& 0.000012& 0.000013& 0.000012& 0.000016& 0.000040& 0.000015& 0.000016\\ 0.000019& 0.000018& 0.000016& 0.000014& 0.000015& 0.000014& 0.000016& 0.000015& 0.000042& 0.000019\\ 0.000020& 0.000018& 0.000016& 0.000015& 0.000016& 0.000015& 0.000018& 0.000016& 0.000019& 0.000044\end{array}\right)$

Described content be only the present invention conceive under basic explanation, and according to any equivalent transformation that technical scheme of the present invention is done, all should protection scope of the present invention be belonged to.

Claims (3)

1. a wind-electricity integration Electrical Power System Dynamic environmental economy dispatching method, wind-electricity integration electric system is set, comprise wind energy turbine set, more than one fired power generating unit, and dispatching center, dispatching center comprises Dispatching Control System and optimizing scheduling center, supply needed for Dispatching Control System acquisition system, described supply comprises personal user, medium load and heavy load sum; Described Dispatching Control System also gathers wind energy turbine set and fired power generating unit related data, be transferred to scheduling decision center, scheduling decision center obtains after above-mentioned related data draws optimum results according to data model and feeds back to Dispatching Control System, Dispatching Control System controls the operation of wind energy turbine set and fired power generating unit, it is characterized in that: scheduling decision center arrange as drag and with Dispatching Control System interaction data:

Two objective functions are as follows:

1) cost of electricity-generating

Electric system cost of electricity-generating minimizes and is expressed as:

${\mathrm{minF}}_1=min\underset{t=1}{\overset{T}{\&Sigma;}}\underset{i=1}{\overset{M}{\&Sigma;}}{C}_i\left(T_{i,t}\right)---\left(1\right)$

In formula: M is fired power generating unit number; T is time interval number; C _i() is the cost function of i-th fired power generating unit; T _i,tthat the plan of i-th fired power generating unit in period t is exerted oneself.

2) pollutant emission

Pollutant emission minimizes and is expressed as:

${\mathrm{minF}}_2=min\underset{t=1}{\overset{T}{\&Sigma;}}\underset{i=1}{\overset{M}{\&Sigma;}}{E}_i\left(T_{i,t}\right)---\left(3\right)$

In formula: E _i() is the pollutant emission function of i-th fired power generating unit.

Four constraint conditions are as follows:

1) the power-balance constraint of power outages is considered

$\mathrm{Pr}\{\underset{i=1}{\overset{M}{\&Sigma;}}{T}_{i,t}+W_t\&le;P_{D,t}+P_{L,t}\}\&le;{\mathrm{\&eta;}}_1,1\&le;t\&le;T---\left(5\right)$

In formula: W _tthat the meritorious of wind energy turbine set in period t is exerted oneself; P _d,tl _t,sit is system load demand in period t; P _l,tit is the power outages in period t; η ₁it is the confidence level meeting workload demand.

2) unit operation constraint

$T_i^{\min }\&le;T_{i,t}\&le;T_i^{\max },1\&le;t\&le;T---\left(7\right)$

In formula: it is the meritorious upper limit of exerting oneself of i-th fired power generating unit.

3) unit climbing rate constraint

T _i,t-T _i,t-1≤UR _i·T ₆₀,1≤t≤T(8)

T _i,t-1-T _i,t≤DR _i·T ₆₀,1≤t≤T(9)

In formula: UR _i, DR _irepresent the upper and lower climbing rate restriction of i-th fired power generating unit respectively; Δ t is the time interval of each period.

4) spinning reserve constraint

Positive rotation Reserve Constraint is expressed as:

$\mathrm{Pr}\{{\mathrm{SSR}}_t^u\&le;P_{D,t}\&CenterDot;\mathrm{\&alpha;}+W_t\&CenterDot;{\mathrm{\&beta;}}_u\}\&le;{\mathrm{\&eta;}}_2,1\&le;t\&le;T---\left(10\right)$

In formula: α is the demand that system loading predicated error aligns spinning reserve; β _ufor wind power output predicated error aligns the demand of spinning reserve; η ₂it is the confidence level meeting positive rotation standby requirement.Wherein, positive rotation is for subsequent use specifically can be expressed as:

${\mathrm{SSR}}_t^u=\underset{i=1}{\overset{M}{\&Sigma;}}{\mathrm{SR}}_{i,t}^u=\underset{i=1}{\overset{M}{\&Sigma;}}min\{T_i^{\max }-T_{i,t},\frac{{\mathrm{UR}}_i}{6}\},1\&le;t\&le;T---\left(11\right)$

Negative rotation turns Reserve Constraint and can be expressed as:

$\mathrm{Pr}\{{\mathrm{SSR}}_t^d\&le;(W_{max}-W_t)\&CenterDot;{\mathrm{\&beta;}}_d\}\&le;{\mathrm{\&eta;}}_3,1\&le;t\&le;T---\left(12\right)$

In formula: W _maxfor the rated power of wind energy turbine set; β _dfor wind power output predicated error is to the demand of negative spinning reserve; η ₃meet the confidence level that negative rotation turns standby requirement.Wherein, negative spinning reserve specifically can be expressed as:

${\mathrm{SSR}}_t^d=\underset{i=1}{\overset{M}{\&Sigma;}}{\mathrm{SR}}_{i,t}^d=\underset{i=1}{\overset{M}{\&Sigma;}}\min \{T_{i,t}-T_i^{\min },\frac{{\mathrm{DR}}_i}{6}\},1\&le;t\&le;T---\left(13\right)$

The meritorious T that exerts oneself of each unit in day part is set _i,t(i=1,2 ..., M; T=1,2 ..., T) be decision variable, adopt real coding mode, by all T _i,tbody is one by one connected into, corresponding expression scheduling scheme according to period and number order.Because decision variable dimension is n=MT, individual X is corresponding can be expressed as:

X=[T _1,1,...,T _1,T,...,T _M,1,...,T _M,T](23)

In initialization procedure, T _i,t(i=1,2 ..., M; T=1,2 ..., T) can be obtained by following formula:

$T_{i,t}=T_i^{\min }+\mathrm{\&xi;}\&CenterDot;(T_i^{\max }-T_i^{\min })---\left(24\right)$

In formula: ξ is for obeying (0,1) equally distributed random number

The power-balance constraint of power outages adopts following steps:

Step 1: time period numbering t=1;

Step 2: calculation interval t internal power Constraints of Equilibrium violation value Δ _t:

${\mathrm{\&Delta;}}_t=F(P_{D,t}+P_{L,t}-\underset{i=1}{\overset{M}{\&Sigma;}}{T}_{i,t})-{\mathrm{\&eta;}}_1---\left(25\right)$

Step 3: if Δ _t≤ 0, then jump to step 5;

Step 4: each fired power generating unit plan is exerted oneself T _i,t(i=1,2 ..., M) adjust:

$T_{i,t}=\min \{T_i^{\max },T_{i,t}+{\mathrm{\&Delta;}}_t\&CenterDot;{\mathrm{\&omega;}}_{i,t}\},$ Wherein weight ${\mathrm{\&omega;}}_{i,t}=\frac{T_i^{\max }-T_{i,t}}{\underset{i=1}{\overset{M}{\&Sigma;}}(T_i^{\max }-T_{i,t})}---\left(26\right)$

To exert oneself T to check the plan after adjustment _i,twhether meet constraint condition, jump to step 2;

Step 5: if t<T, so t=t+1 jump to step 2;

Step 6: power-balance constraint dynamic process terminates.

Unit operation constraint, climbing rate constraint dynamic process are adopted with the following method:

Step 1: initialization thermoelectricity genset numbering i=1, time period numbering t=1;

Step 2: as t=1, is arranged ${\mathrm{upper}}_{i,t}=T_i^{\max },{\mathrm{lower}}_{i,t}=T_i^{\min };$ As t>1,

Arrange ${\mathrm{upper}}_{i,t}=\min \{T_{i,t-1}+{\mathrm{UR}}_i,T_i^{\max }\},{\mathrm{lower}}_{i,t}=\max \{T_{i,t-1}-{\mathrm{DR}}_i,T_i^{\min }\}---\left(27\right)$

Step 3: the T if fired power generating unit plan is exerted oneself _i,tdo not meet unit operation constraint, the constraint of climbing rate, so revise as follows respectively:

Step 4: if t<T, so t=t+1 jump to step 2;

Step 5: if i<M, so i=i+1 jump to step 2;

Step 6: unit operation constraint, climbing rate constraint dynamic process terminate.

The process of positive rotation Reserve Constraint condition is adopted with the following method:

Step 1: time period numbering t=1;

Step 2: positive rotation Reserve Constraint violation value Δ in calculation interval t _t:

${\mathrm{\&Delta;}}_t=F\left(\frac{1}{{\mathrm{\&beta;}}_u}\right({\mathrm{SSR}}_t^u-P_{D,t}\&CenterDot;\mathrm{\&alpha;}))+{\mathrm{\&eta;}}_2-1---\left(29\right)$

Step 3: if Δ _t>=0, then jump to step 5;

Step 4: each fired power generating unit plan is exerted oneself T _i,t(i=1,2 ..., M) adjust:

$T_{i,t}=\max \{T_i^{\min },T_{i,t}-{\mathrm{\&Delta;}}_t\&CenterDot;{\mathrm{\&omega;}}_{i,t}\},$ Wherein weight ${\mathrm{\&omega;}}_{i,t}=\frac{T_{i,t}-T_i^{\min }}{\underset{i=1}{\overset{M}{\&Sigma;}}(T_{i,t}-T_i^{\min })}---\left(30\right)$

To exert oneself T to check the plan after adjustment _i,twhether meet constraint condition, jump to step 2;

Step 5: if t<T, so t=t+1 jump to step 2;

Step 6: positive rotation Reserve Constraint dynamic process terminates.

Negative rotation turns the process of Reserve Constraint condition and adopts with the following method:

Step 1: time period numbering t=1;

Step 2: in calculation interval t, negative rotation turns Reserve Constraint violation value Δ _t:

${\mathrm{\&Delta;}}_t=F(W_{max}-\frac{{\mathrm{SSR}}_t^d}{{\mathrm{\&beta;}}_d})-{\mathrm{\&eta;}}_3---\left(31\right)$

Step 3: if Δ _t≤ 0, then jump to step 5;

Step 4: each fired power generating unit plan is exerted oneself T _i,t(i=1,2 ..., M) adjust:

$T_{i,t}=\min \{T_i^{\max },T_{i,t}+{\mathrm{\&Delta;}}_t\&CenterDot;{\mathrm{\&omega;}}_{i,t}\},$ Wherein weight ${\mathrm{\&omega;}}_{i,t}=\frac{T_i^{\max }-T_{i,t}}{\underset{i=1}{\overset{M}{\&Sigma;}}(T_i^{\max }-T_{i,t})}---\left(32\right)$

To exert oneself T to check the plan after adjustment _i,twhether meet constraint condition, jump to step 2;

Step 5: if t<T, so t=t+1 jump to step 2;

Step 6: negative rotation turns Reserve Constraint dynamic process and terminates;

Electric power supply after optimization is transferred to user by described Dispatching Control System.

2. a wind-electricity integration Electrical Power System Dynamic environmental economy dispatching method as claimed in claim 1, is characterized in that:

According to the fired power generating unit cost of electricity-generating of valve point effect, in formula (1):

$C_i\left(T_{i,t}\right)=a_i{T}_{i,t}^2+b_i{T}_{i,t}+d_i+\left|e_{i}sin\left(f_i\right(T_i^{\min }-T_{i,t})\right)|,1\&le;i\&le;M,1\&le;t\&le;T---\left(2\right)$

In formula: a _i, b _i, d _i, e _iand f _iit is the cost of electricity-generating coefficient of i-th fired power generating unit; it is the meritorious lower limit of exerting oneself of i-th fired power generating unit.

3. a wind-electricity integration Electrical Power System Dynamic environmental economy dispatching method as claimed in claim 1, is characterized in that:

To gain merit the relation between exerting oneself according to pollutant discharge amount and fired power generating unit, in formula (2):

$E_i\left(T_{i,t}\right)={\mathrm{\&alpha;}}_i{T}_{i,t}^2+{\mathrm{\&gamma;T}}_{i,t}+{\mathrm{\&lambda;}}_i+{\mathrm{\&delta;}}_i\exp \left({\mathrm{\&tau;}}_i{T}_{i,t}\right),1\&le;i\&le;M,1\&le;t\&le;T---\left(4\right)$

In formula: α _i, γ _i, λ _i, δ _iand τ _iit is the pollutant discharge coefficient of i-th fired power generating unit.