CN103324848B - Method for optimizing electric-quantity-constrained monthly unit commitment and based on induction target function - Google Patents

Abstract

The invention discloses a method for optimizing electric-quantity-constrained monthly unit commitment and based on an induction target function and belongs to the technical field of power dispatching automation. The method includes building a mathematical model of a monthly unit commitment plan according to unit commitment basic data, building a mathematical model of a loose monthly unit commitment plan and calculating the model to obtain the loose monthly unit commitment plan, counting the number of start-stop variables with the value as 0 or 1 in the loose monthly unit commitment plan, building the induction target function of the mathematical model of the loose monthly unit commitment plan, conducting iteration optimization till the number of the start-top variables with the value as 0 or 1 stops increasing, adopting the induction target function to obtain the monthly unit commitment plan, conducting further optimization if a deviation factor of the monthly unit commitment plan is larger than a deviation threshold and finishing optimization otherwise. The method can be used for optimizing the electric-quantity-constrained monthly unit commitment plan, calculation efficiency is remarkably improved, and the method has important practical significance and good application prospects.

Description

Based on induction objective function containing the monthly Unit Combination optimization method of Constraint

Technical field

The invention belongs to technical field of power dispatching automation, provide especially based on induction objective function containing the monthly Unit Combination optimization method of Constraint.

Background technology

Power scheduling is that electrical network adopts various Optimized-control Techniques, coordinates various generating resources, maintains the equilibrium of supply and demand, guarantees the important means of power grid security economical operation.The essence of power scheduling work be a class in certain hour yardstick, sending out, transmitting electricity resource-constrained in the situation that, from the control problem of time and Spatial Dimension overall arrangement electrical production.According to the difference of time scale, be divided into year, monthly, a few days ago, in a few days, the scheduling problem such as real-time, the scheduling problem of different time yardstick has feature separately.

Monthly Unit Combination optimization is one of important step of power scheduling.Monthly Unit Combination optimization refers to that, electrical network energy resource consumption minimum taking operation of power networks cost is minimum etc. as target, taking start-stop of generator set state and unit output as control variable, meet the constraint conditions such as power balance constraint, Constraint, unit operation constraint, Network Security Constraints, optimize start-stop of generator set and the plan of exerting oneself of every day in following month (or each period, as every 2 hours as a period).Under " three public affairs " (open, fair and impartial) scheduling method of carrying out in China, optimize monthly Unit Combination plan and need to meet unit Constraint, be that the plan electric weight of unit within the moon should equate substantially with its contract electric weight, allow has deviation in certain scope.Optimizing effectively and rationally monthly Unit Combination plan, by reasonable disposition grid generation resource, is the requisite measure that ensures safe operation of electric network, promotes energy-saving and emission-reduction, has obtained the extensive common recognition of academia and industry member.

Existing lower containing the monthly Unit Combination optimization method of Constraint counting yield, optimize overlong time, be difficult to adapt to the needs of dispatching of power netwoks operation real work.Existing research (Fu Yong, Shahidehpour S M and Li Zuyi, Long-term security-constrained unit commitment:hybrid Dantzig-Wolfe decomposition and subgradient approach (long-term safety constraint Unit Combination: the hybrid solution approach of a kind of DW decomposition method and subgradient method) .IEEE Transactions on Power Systems, 2005.20 (4): 2093-2106.) the hybrid solution approach optimization of a kind of Dantzig-Wolfe (DW) decomposition method and subgradient method has been proposed containing the monthly Unit Combination plan of Constraint, but for the electrical network of IEEE-118 node, the method is consuming time exceedes 6 hours, and the nodes of actual electric network is far longer than 118, therefore the counting yield of the method is difficult to meet practical application request.Existing Chinese invention patent (Li Lili, fourth is proper, Geng Jian, Wang Gang, Yang Zhenglin, Xie Lirong. mid-long-term unit commitment optimizing method: Jiangsu, CN102097866A[P] .2011-06-15) a kind of mid-long-term unit commitment optimizing method proposed, the method comprises the following steps: set up the medium-term and long-term security constraint Unit Combination model taking unit generation amount and Expected energy deviation minimum as target, non-linear factor linearization in model is expressed, adopt mixed integer programming method calculate unit in the control cycle start and stop state of each day, rate of load condensate and the meritorious of peak load period are exerted oneself, then the Unit Commitment state obtaining according to optimization and peak load period unit output, consider that overall network monitors element, employing peak load point every day carries out Security Checking, if there is the not element by Security Checking, the security constraints of this element is added in medium-term and long-term security constraint Unit Combination model, again optimize Unit Combination plan, until all elements pass through Security Checking.But the method, taking unit generation amount and Expected energy deviation minimum as target, only can ensure that unit generation amount and Expected energy are comparatively approaching on the whole, cannot ensure that optimum results strictly meets each unit Constraint.Containing the optimization of the monthly Unit Combination plan of Constraint, be an extensive mixed integer programming problem that constraint is strict, optimizing is difficult in essence, be difficult to efficiently try to achieve optimum solution; The counting yield of existing research or invention and precision there is no method and meet grid company by the demand of monthly Unit Combination plan realization generating most optimum distribution of resources.Therefore, grid company is calculated the optimization method containing the monthly Unit Combination plan of Constraint efficient, result is optimum in the urgent need to a kind of, for controlling and scheduling grid generation resource, meet the actual demand that power grid security and unit complete contract electric weight simultaneously, reach the target of most optimum distribution of resources and energy-saving and emission-reduction.

Summary of the invention

The object of the invention is the weak point for overcoming prior art, provide based on induction objective function containing the monthly Unit Combination optimization method of Constraint, the present invention can fully excavate the lax calculated effective information of monthly Unit Combination, build the induction objective function containing the monthly Unit Combination Optimized model of Constraint, have and calculate efficient, the optimum feature of result, for the Automatic Optimal of monthly Unit Combination plan is really moved towards the practical solid foundation of having established.

The invention provides based on induction objective function containing the monthly Unit Combination optimization method of Constraint, monthly Unit Combination optimization refers to that, electrical network energy resource consumption minimum taking operation of power networks cost is minimum etc. as target, taking start-stop of generator set state and unit output as control variable, meet the constraint conditions such as power balance constraint, Constraint, unit operation constraint, Network Security Constraints, optimize start-stop of generator set and the plan of exerting oneself of every day in following month (or each period, as every 2 hours as a period).The present invention includes following steps:

(1) obtain Unit Combination basic data;

Described Unit Combination basic data refers to that the operation characteristic data of genset, monthly load prediction data, each unit contract electric weight, grid topology data and the Optimal Parameters data etc. of following month build the data that monthly Unit Combination plan mathematics model needs;

The operation characteristic data of described genset comprise that fuel cost, start expense, idleness expense, minimum start/stop time, the variation upper limit/lower limit of exerting oneself, the min/max technology of genset go out force data;

Described monthly load prediction data are following month electric load conditions of demand that obtain according to monthly load prediction software, comprise the total load data of following month each day day part electrical network, the node load data of the each node of day part;

To be each unit exceed the quata or vacancy generated energy and the monthly electricity contract of having signed by the following moon cumulative in this month the described unit contract electric weight of following month; Accumulate mode is, if unit this month real generated energy exceeded this month contract electric weight, from the monthly electricity contract of following month, deduct; If unit this month, real generated energy was less than this month contract electric weight, in the monthly electricity contract of following month, supply;

Described grid topology data comprises the meritorious trend limit and the circuit ID comprising, each genset and the node load meritorious transfer distribution factor data to every transmission line of electricity thereof of the node of electric power networks and the annexation of transmission line of electricity, each transmission cross-section;

Described Optimal Parameters data comprise lower deviation ratio and upper deviation ratio, electrical network spinning reserve rate and the deviation threshold data of unit contract electric weight;

(2) build the mathematical model containing the monthly Unit Combination plan of Constraint according to Unit Combination basic data;

The described mathematical model containing the monthly Unit Combination plan of Constraint is made up of objective function and constraint condition;

(2-1) build the objective function containing the monthly Unit Combination plan of Constraint mathematical model, expression formula is as follows:

$\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\underset{t=1}{\overset{T}{\mathrm{\Σ}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]$

Define this objective function for " former objective function "; represent the corresponding cost of electricity-generating of monthly Unit Combination plan; Wherein, represent the fuel cost of unit i at period t, represent the start expense of unit i at period t, represent the idleness expense of unit i at period t; N represents unit number, and T represents period number; Wherein, it is control variable;

(2-2) build the constraint condition containing the monthly Unit Combination plan of Constraint mathematical model, expression formula is as follows respectively:

(2-2-1) fuel cost constraint condition

$\begin{array}{cc}{C}_{i,t}^P\≥A_i^P{P}_{i,t}\mathrm{\Δ}{T}^D& \∀t,\∀i\end{array}$

(2-2-2) start expense restriction condition

$\begin{array}{cc}{C}_{i,t}^U\≥A_i^U[I_{i,t}-I_{i,(t-1)}]& \∀t,\∀i\end{array}$

$\begin{array}{cc}{C}_{i,t}^U\≥0& \∀t,\∀i\end{array}$

(2-2-3) idleness expense constraint condition

$\begin{array}{cc}{C}_{i,t}^D\≥A_i^D[I_{i,(t-1)}-I_{i,t}]& \∀t,\∀i\end{array}$

$\begin{array}{cc}{C}_{i,t}^D\≥0& \∀t,\∀i\end{array}$

Wherein, P _i,trepresent that unit i exerts oneself at the meritorious of period t, I _i,trepresent the start and stop state of unit i at period t, I _i,tvalue can only be 1 or 0, I _i,tbe 1 expression start, I _i,tbe that 0 expression is shut down; P _i,twith I _i,talso be control variable; Δ T ^dintersegmental interval time while representing every two; represent that unit i often sends the fuel cost that unit quantity of electricity needs; with represent respectively the expense that unit i starts shooting once and shut down expense once; Fuel cost constraint condition represents the fuel cost that unit i pays for sending power at period t; Start expense restriction condition represent when unit i at period t-1 the start expense during to period t transition, if unit i starts shooting in this transition period, start expense is if not start, start expense is 0; Idleness expense constraint condition represent when unit i at period t-1 the idleness expense during to period t transition, if unit i shuts down in this transition period, idleness expense is if do not shut down, idleness expense is 0;

(2-2-4) electric network active equilibrium constraint

$\begin{array}{cc}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{i,t}=D_t& \∀t\end{array}$

Wherein, D _trepresent the electrical network total load of period t, in this constraint representation electrical network, the generated output of all units equals electrical network total load;

(2-2-5) electrical network spinning reserve constraint condition

$\begin{array}{cc}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{I}_{i,t}{P}_i^{\max }\≥D_t(1+R^P)& \∀t\end{array}$

Wherein, the maximum technology that represents unit i is exerted oneself, R ^prepresent electrical network spinning reserve rate, in this constraint representation electrical network, the capacity sum of all start units should be greater than the certain proportion of electrical network total load;

(2-2-6) unit minimum start/stop time of constraint condition

$\begin{array}{cc}(T_{i,t}^{\mathrm{on}}-T_{i,\min }^{\mathrm{on}})[I_{i,(t-1)}-I_{i,t}]\≥0& \∀t,\∀i\end{array}$

$T_{i,t}^{\mathrm{on}}=\underset{{k=t-T}_{i,\min }^{\mathrm{on}}}{\overset{t-1}{\mathrm{\Σ}}}{I}_{i,k}$

$\begin{array}{cc}(T_{i,t}^{\mathrm{off}}-T_{i,\min }^{\mathrm{off}})[I_{i,t}-I_{i,(t-1)}]\≥0& \∀t,\∀i\end{array}$

$T_{i,t}^{\mathrm{off}}=\underset{{k=t-T}_{i,\min }^{\mathrm{off}}}{\overset{t-1}{\mathrm{\Σ}}}(1-I_{i,k})$

Wherein, with representing respectively on time and stop time that unit i had experienced continuously before period t, is control variable; with represent respectively minimum on time and the minimum stop time of unit i; Must experience after this constraint representation unit i start time could shut down again, must experience after shutdown time could start shooting again;

(2-2-7) unit output changes bound constraint condition

$P_{i,t}-P_{i,(t-1)}\≤{\mathrm{DP}}_i^{\max }{I}_{i,(t-1)}+[I_{i,t}-I_{i,(t-1)}]P_i^{\min }$

$\begin{array}{cc}+P_i^{\max }(1-I_{i,t})& \∀t,\∀i\end{array}$

$P_{i,(t-1)}-P_{i,t}\≤{\mathrm{DP}}_i^{\min }{I}_{i,t}-[I_{i,t}-I_{i,(t-1)}]P_i^{\min }$

$\begin{array}{cc}+P_i^{\max }[1-I_{i,(t-1)}]& \∀t,\∀i\end{array}$

Wherein, with represent respectively the upper and lower bound that unit output changes, with the maximum technology that represents respectively unit i is exerted oneself and minimum technology is exerted oneself, and determines by generator inherent characteristic; At period t-1, the meritorious variable quantity of exerting oneself during to period t transition should limit within the specific limits this constraint representation unit i; If the transition period of unit i from period t-1 to period t is start process, i.e. I _{i, (t-1)}=0 and I _i,t=1, unit output changes bound constraint condition and is and $P_{i,(t-1)}-P_{i,t}\≤{\mathrm{DP}}_i^{\min }-P_i^{\min }+P_i^{\max },$ Due to I _{i, (t-1)}=0, therefore P _{i, (t-1)}=0, and with constraint condition (2-2-8) acting in conjunction, by P _i,tbe defined as first period after the start of expression unit exerts oneself and can only exert oneself for minimum technology now, for redundancy, actual inoperative; If the transition period of unit i from period t-1 to period t is stopping process, i.e. I _{i, (t-1)}=1 and I _i,t=0, unit output changes bound constraint condition and is $P_{i,t}-P_{i,(t-1)}\≤{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ And $P_{i,(t-1)}-P_{i,t}\≤P_i^{\min },$ Now, $P_{i,t}-P_{i,(t-1)}\≤{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ For redundancy, actual inoperative, due to I _i,t=0, therefore P _i,t=0, and with constraint condition (2-2-8) acting in conjunction, by P _{i, (t-1)}be defined as last period before expression compressor emergency shutdown exerts oneself and can only exert oneself for minimum technology if unit i is from period t-1 to period t always in running status, i.e. I _{i, (t-1)}=1 and I _i,t=1, unit output changes bound constraint condition and is and represent that the bound that unit output changes is respectively with if unit i is from period t-1 to period t always in stopped status, i.e. I _{i, (t-1)}=0 and I _i,t=0, unit output changes bound constraint condition and is and all can be met;

(2-2-8) unit output bound constraint condition

$\begin{array}{cc}{I}_{i,t}{P}_i^{\min }\≤P_{i,t}\≤I_{i,t}{P}_i^{\max }& \∀t,\∀i\end{array}$

When this constraint representation unit start, it is exerted oneself and in certain scope, change; When compressor emergency shutdown, it is exerted oneself is 0;

(2-2-9) unit Constraint condition

$\begin{array}{cc}{W}_i^{\min }\≤\mathrm{\Δ}{T}^D\·\underset{t=1}{\overset{T}{\mathrm{\Σ}}}{P}_{i,t}\≤W_i^{\max }& \∀i\end{array}$

Wherein, with represent respectively electric weight lower limit and the upper limit of unit i, with be multiplied by lower deviation ratio by the contract electric weight of unit i respectively and upper deviation ratio obtains; The plan electric weight of this constraint representation unit within the moon equate substantially with its contract electric weight, and allow has deviation in certain scope;

(2-2-10) Network Security Constraints condition

$\begin{array}{cc}{-P}_s^{\max }\≤\underset{l\∈s}{\mathrm{\Σ}}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{G}_{l-i}{P}_{i,t}-\underset{l\∈s}{\mathrm{\Σ}}\underset{j=1}{\overset{J}{\mathrm{\Σ}}}{G}_{l-j}{D}_{j,t}\≤P_s^{\max }& \∀t,\∀s\end{array}$

Wherein, represent the meritorious trend limit of transmission cross-section s, G _l-irepresent the meritorious transfer distribution factor of unit i to circuit l, G _l-jrepresent the meritorious transfer distribution factor of node load j to circuit l, D _j,trepresent the node load of node load j at period t, l ∈ s represents that circuit l belongs to transmission cross-section s, and J represents the number of node load; In this constraint representation electrical network, the meritorious trend of some transmission cross-section can not exceed its trend limit;

(3) by the I containing in the monthly Unit Combination plan of Constraint mathematical model _i,trelax, make I _i,tany value in desirable [0,1], the mathematical model of the lax monthly Unit Combination plan of structure;

(4) adopt linear programming for solution device, solve the mathematical model of lax monthly Unit Combination plan, obtain lax monthly Unit Combination plan;

(5) in the works, statistics value is 0 or 1 start and stop variable I to the lax monthly Unit Combination obtaining in step (4) _i,tnumber, be designated as H (k), the sequence number that k is current iteration, and define H (0)=0;

(6) judge lax monthly Unit Combination that current iteration obtains in the works value be 0 or 1 start and stop variable I _i,tnumber H (k) whether increase compared with H (k-1); If increase, proceed step (7); If do not increase, jump to step (8);

(7) according to start and stop variable I _i,tvalue, build the induction objective function of lax monthly Unit Combination plan mathematical model;

The former objective function of the mathematical model of lax monthly Unit Combination plan is revised as to the induction objective function of following form:

$\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\underset{t=1}{\overset{T}{\mathrm{\Σ}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]+\underset{i}{\mathrm{\Σ}}\underset{t}{\mathrm{\Σ}}{\mathrm{\ξ}}_{i,t}^{(k+1)}{I}_{i,t}$

Wherein, represent lax monthly Unit Combination that the k time iteration obtain in the works unit i at the start and stop variate-value of period t, B _irepresent fuel cost when unit i operates in minimum technology and exerts oneself; Especially, when time, order wherein ε=0.01, M is B in all units _imaximal value;

Jump to step (4), iterations k adds 1;

(8) adopt induction objective function, build and solve the mathematical model containing the monthly Unit Combination plan of Constraint, obtain monthly Unit Combination plan;

If the value of current iteration number of times k is adopt the induction objective function form when inferior iteration, and make I _i,tonly can value 0 or 1, build the mathematical model containing the monthly Unit Combination plan of Constraint, and adopt mixed integer programming solver to solve this model, obtain monthly Unit Combination plan;

(9) whether the deviation factors λ that judges monthly Unit Combination plan is less than deviation threshold λ ₀;

Calculate the deviation factors λ of monthly Unit Combination plan according to following expression:

$\mathrm{\λ}=\frac{O_{P_0^{*}}-O_{P_S}}{O_{P_0^{*}}}$

Wherein, that step (8) solves the corresponding cost of electricity-generating of monthly Unit Combination plan obtaining; the lax corresponding cost of electricity-generating of monthly Unit Combination plan while being k=1; In fact permanent establishment, therefore 0≤λ < 1 is permanent sets up; The gap of the monthly Unit Combination plan of the monthly Unit Combination plan that λ sign step (8) obtains and optimum;

If λ > is λ ₀, continue step (10); If λ≤λ ₀, jump to step (11);

(10) adopt former objective function, the monthly Unit Combination plan obtaining taking step (8), as initial solution, builds and solves the mathematical model containing the monthly Unit Combination plan of Constraint, obtains monthly Unit Combination plan;

Refer to containing the mathematical model of the monthly Unit Combination plan of Constraint the mathematical model that step (2) builds, adopt mixed integer programming solver to solve this model, and the monthly Unit Combination that step (8) is obtained be intended to be initial solution implant solution procedure, with speed-up computation;

(11) optimize and finish, acquired results is the monthly Unit Combination plan containing Constraint, and grid company is controlled accordingly the start and stop of genset and exerted oneself.

Technical characterstic of the present invention and beneficial effect:

The present invention can fully excavate the lax calculated effective information of monthly Unit Combination, has built the induction objective function containing the monthly Unit Combination Optimized model of Constraint.While adopting induction objective function solving-optimizing model, the number of times of branch-and-bound will significantly reduce, thereby increase substantially the counting yield of monthly Unit Combination plan.Simultaneously, the present invention can judge whether the monthly Unit Combination plan of trying to achieve meets the accuracy requirement that deviation threshold sets flexibly, if do not met, the present invention can ensure to try to achieve exact solution by step (10), does not substantially affect the counting yield of monthly Unit Combination plan simultaneously.Test analysis based on Chinese Provincial electrical network real data shows, counting yield of the present invention can reach 30～50 times of current business software for calculation.To sum up, the present invention is a kind of optimization method containing the monthly Unit Combination plan of Constraint, has and calculates efficient, the optimum feature of result, for the Automatic Optimal of monthly Unit Combination plan is really moved towards the practical solid foundation of having established.Grid company can according to the present invention, optimization obtains monthly Unit Combination plan, rationally control and economic load dispatching grid generation resource, meet power grid security and unit simultaneously and complete the actual demand of contract electric weight, reach the target of most optimum distribution of resources and energy-saving and emission-reduction.

Brief description of the drawings

Fig. 1 is the process flow diagram containing the monthly Unit Combination optimization method of Constraint based on induction objective function;

Fig. 2 (a) is in the embodiment of the present invention, while adopting CPLEX to be optimized, and the figure that the target function value of best feasible solution changed with computing time;

Fig. 2 (b) is in the embodiment of the present invention, while adopting the present invention to be optimized, and the figure that the target function value of best feasible solution changed with computing time.

Embodiment

Below in conjunction with drawings and the embodiments, the present invention is further detailed explanation.Should be appreciated that embodiment described herein can be in order to explain the present invention, but do not limit the present invention.

The invention provides based on induction objective function containing the monthly Unit Combination optimization method of Constraint, as shown in Figure 1, embodiment is as follows:

(1) obtain Unit Combination basic data;

Described Unit Combination basic data refers to that the operation characteristic data of genset, monthly load prediction data, each unit contract electric weight, grid topology data and the Optimal Parameters data etc. of following month build the data that monthly Unit Combination plan mathematics model needs;

The operation characteristic data of described genset comprise that fuel cost, start expense, idleness expense, minimum start/stop time, the variation upper limit/lower limit of exerting oneself, the min/max technology of genset go out force data;

Described monthly load prediction data are following month electric load conditions of demand that obtain according to monthly load prediction software, comprise the total load data of following month each day day part electrical network, the node load data of the each node of day part;

To be each unit exceed the quata or vacancy generated energy and the monthly electricity contract of having signed by the following moon cumulative in this month the described unit contract electric weight of following month; Accumulate mode is, if unit this month real generated energy exceeded this month contract electric weight, from the monthly electricity contract of following month, deduct; If unit this month, real generated energy was less than this month contract electric weight, in the monthly electricity contract of following month, supply;

Described grid topology data comprises the meritorious trend limit and the circuit ID comprising, each genset and the node load meritorious transfer distribution factor data to every transmission line of electricity thereof of the node of electric power networks and the annexation of transmission line of electricity, each transmission cross-section;

Described Optimal Parameters data comprise lower deviation ratio and upper deviation ratio, electrical network spinning reserve rate and the deviation threshold data of unit contract electric weight;

(2) build the mathematical model containing the monthly Unit Combination plan of Constraint according to Unit Combination basic data;

The described mathematical model containing the monthly Unit Combination plan of Constraint is made up of objective function and constraint condition;

(2-1) build the objective function containing the monthly Unit Combination plan of Constraint mathematical model, expression formula is as follows:

$\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]$

Define this objective function for " former objective function "; represent the corresponding cost of electricity-generating of monthly Unit Combination plan; Wherein, represent the fuel cost of unit i at period t, represent the start expense of unit i at period t, represent the idleness expense of unit i at period t; N represents unit number, and T represents period number; Wherein, it is control variable;

(2-2) build the constraint condition containing the monthly Unit Combination plan of Constraint mathematical model, expression formula is as follows respectively:

(2-2-1) fuel cost constraint condition

$\begin{array}{cc}{C}_{i,t}^P\&GreaterEqual;A_i^P{P}_{i,t}\mathrm{\&Delta;}{T}^D& \&ForAll;t,\&ForAll;i\end{array}$

(2-2-2) start expense restriction condition

$\begin{array}{cc}{C}_{i,t}^U\&GreaterEqual;A_i^U[I_{i,t}-I_{i,(t-1)}]& \&ForAll;t,\&ForAll;i\end{array}$

$\begin{array}{cc}{C}_{i,t}^U\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

(2-2-3) idleness expense constraint condition

$\begin{array}{cc}{C}_{i,t}^D\&GreaterEqual;A_i^D[I_{i,(t-1)}-I_{i,t}]& \&ForAll;t,\&ForAll;i\end{array}$

$\begin{array}{cc}{C}_{i,t}^D\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

Wherein, P _i,trepresent that unit i exerts oneself at the meritorious of period t, I _i,trepresent the start and stop state of unit i at period t, I _i,tvalue can only be 1 or 0, I _i,tbe 1 expression start, I _i,tbe that 0 expression is shut down; P _i,twith _{ii, t}also be control variable; Δ T ^dintersegmental interval time while representing every two; represent that unit i often sends the fuel cost that unit quantity of electricity needs; with represent respectively the expense that unit i starts shooting once and shut down expense once; Fuel cost constraint condition represents the fuel cost that unit i pays for sending power at period t; Start expense restriction condition represent when unit i at period t-1 the start expense during to period t transition, if unit i starts shooting in this transition period, start expense is if not start, start expense is 0; Idleness expense constraint condition represent when unit i at period t-1 the idleness expense during to period t transition, if unit i shuts down in this transition period, idleness expense is if do not shut down, idleness expense is 0;

(2-2-4) electric network active equilibrium constraint

$\begin{array}{cc}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{P}_{i,t}=D_t& \&ForAll;t\end{array}$

Wherein, D _trepresent the electrical network total load of period t, in this constraint representation electrical network, the generated output of all units equals electrical network total load;

(2-2-5) electrical network spinning reserve constraint condition

$\begin{array}{cc}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{I}_{i,t}{P}_i^{\max }\&GreaterEqual;D_t(1+R^P)& \&ForAll;t\end{array}$

Wherein, the maximum technology that represents unit i is exerted oneself, R ^prepresent electrical network spinning reserve rate, in this constraint representation electrical network, the capacity sum of all start units should be greater than the certain proportion of electrical network total load;

(2-2-6) unit minimum start/stop time of constraint condition

$\begin{array}{cc}(T_{i,t}^{\mathrm{on}}-T_{i,\min }^{\mathrm{on}})[I_{i,(t-1)}-I_{i,t}]\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

$T_{i,t}^{\mathrm{on}}=\underset{{k=t-T}_{i,\min }^{\mathrm{on}}}{\overset{t-1}{\mathrm{\&Sigma;}}}{I}_{i,k}$

$\begin{array}{cc}(T_{i,t}^{\mathrm{off}}-T_{i,\min }^{\mathrm{off}})[I_{i,t}-I_{i,(t-1)}]\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

$T_{i,t}^{\mathrm{off}}=\underset{{k=t-T}_{i,\min }^{\mathrm{off}}}{\overset{t-1}{\mathrm{\&Sigma;}}}(1-I_{i,k})$

Wherein, with representing respectively on time and stop time that unit i had experienced continuously before period t, is control variable; with represent respectively minimum on time and the minimum stop time of unit i; Must experience after this constraint representation unit i start time could shut down again, must experience after shutdown time could start shooting again;

(2-2-7) unit output changes bound constraint condition

$P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }{I}_{i,(t-1)}+[I_{i,t}-I_{i,(t-1)}]P_i^{\min }$

$\begin{array}{cc}+P_i^{\max }(1-I_{i,t})& \&ForAll;t,\&ForAll;i\end{array}$

$P_{i,(t-1)}-P_{i,t}\&le;{\mathrm{DP}}_i^{\min }{I}_{i,t}-[I_{i,t}-I_{i,(t-1)}]P_i^{\min }$

$\begin{array}{cc}+P_i^{\max }[1-I_{i,(t-1)}]& \&ForAll;t,\&ForAll;i\end{array}$

Wherein, with represent respectively the upper and lower bound that unit output changes, with the maximum technology that represents respectively unit i is exerted oneself and minimum technology is exerted oneself, and determines by generator inherent characteristic; At period t-1, the meritorious variable quantity of exerting oneself during to period t transition should limit within the specific limits this constraint representation unit i; If the transition period of unit i from period t-1 to period t is start process, i.e. I _{i, (t-1)}=0 and I _i,t=1, unit output changes bound constraint condition and is and $P_{i,(t-1)}-P_{i,t}\&le;{\mathrm{DP}}_i^{\min }-P_i^{\min }+P_i^{\max },$ Due to I _{i, (t-1)}=0, therefore P _{i, (t-1)}=0, and with constraint condition (2-2-8) acting in conjunction, by P _i,tbe defined as first period after the start of expression unit exerts oneself and can only exert oneself for minimum technology now, for redundancy, actual inoperative; If the transition period of unit i from period t-1 to period t is stopping process, i.e. I _{i, (t-1)}=1 and I _i,t=0, unit output changes bound constraint condition and is $P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ And $P_{i,(t-1)}-P_{i,t}\&le;P_i^{\min },$ Now, $P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ For redundancy, actual inoperative, due to I _i,t=0, therefore P _i,t=0, and with constraint condition (2-2-8) acting in conjunction, by P _{i, (t-1)}be defined as last period before expression compressor emergency shutdown exerts oneself and can only exert oneself for minimum technology if unit i is from period t-1 to period t always in running status, i.e. I _{i, (t-1)}=1 and I _i,t=1, unit output changes bound constraint condition and is and represent that the bound that unit output changes is respectively with if unit i is from period t-1 to period t always in stopped status, i.e. I _{i, (t-1)}=0 and I _i,t=0, unit output changes bound constraint condition and is and all can be met;

(2-2-8) unit output bound constraint condition

$\begin{array}{cc}{I}_{i,t}{P}_i^{\min }\&le;P_{i,t}\&le;I_{i,t}{P}_i^{\max }& \&ForAll;t,\&ForAll;i\end{array}$

When this constraint representation unit start, it is exerted oneself and in certain scope, change; When compressor emergency shutdown, it is exerted oneself is 0;

(2-2-9) unit Constraint condition

$\begin{array}{cc}{W}_i^{\min }\&le;\mathrm{\&Delta;}{T}^D\&CenterDot;\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_{i,t}\&le;W_i^{\max }& \&ForAll;i\end{array}$

Wherein, with represent respectively electric weight lower limit and the upper limit of unit i, with be multiplied by lower deviation ratio by the contract electric weight of unit i respectively and upper deviation ratio obtains; The plan electric weight of this constraint representation unit within the moon equate substantially with its contract electric weight, and allow has deviation in certain scope;

(2-2-10) Network Security Constraints condition

$\begin{array}{cc}{-P}_s^{\max }\&le;\underset{l\&Element;s}{\mathrm{\&Sigma;}}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{G}_{l-i}{P}_{i,t}-\underset{l\&Element;s}{\mathrm{\&Sigma;}}\underset{j=1}{\overset{J}{\mathrm{\&Sigma;}}}{G}_{l-j}{D}_{j,t}\&le;P_s^{\max }& \&ForAll;t,\&ForAll;s\end{array}$

Wherein, represent the meritorious trend limit of transmission cross-section s, G _l-irepresent the meritorious transfer distribution factor of unit i to circuit l, G _l-jrepresent the meritorious transfer distribution factor of node load j to circuit l, D _j,trepresent the node load of node load j at period t, l ∈ s represents that circuit l belongs to transmission cross-section s, and J represents the number of node load; In this constraint representation electrical network, the meritorious trend of some transmission cross-section can not exceed its trend limit;

(3) by the I containing in the monthly Unit Combination plan of Constraint mathematical model _i,trelax, make I _i,tany value in desirable [0,1], the mathematical model of the lax monthly Unit Combination plan of structure;

(4) adopt linear programming for solution device, solve the mathematical model of lax monthly Unit Combination plan, obtain lax monthly Unit Combination plan;

(5) in the works, statistics value is 0 or 1 start and stop variable I to the lax monthly Unit Combination obtaining in step (4) _i,tnumber, be designated as H (k), the sequence number that k is current iteration, and define H (0)=0;

(6) judge lax monthly Unit Combination that current iteration obtains in the works value be 0 or 1 start and stop variable I _i,tnumber H (k) whether increase compared with H (k-1); If increase, proceed step (7); If do not increase, jump to step (8);

(7) according to start and stop variable I _i,tvalue, build the induction objective function of lax monthly Unit Combination plan mathematical model; The former objective function of the mathematical model of lax monthly Unit Combination plan is revised as to the induction objective function of following form:

$\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]+\underset{i}{\mathrm{\&Sigma;}}\underset{t}{\mathrm{\&Sigma;}}{\mathrm{\&xi;}}_{i,t}^{(k+1)}{I}_{i,t}$

Wherein, represent lax monthly Unit Combination that the k time iteration obtain in the works unit i at the start and stop variate-value of period t, B _irepresent fuel cost when unit i operates in minimum technology and exerts oneself; Especially, when time, order wherein ε=0.01, M is B in all units _imaximal value;

Jump to step (4), iterations k adds 1;

(8) adopt induction objective function, build and solve the mathematical model containing the monthly Unit Combination plan of Constraint, obtain monthly Unit Combination plan;

If the value of current iteration number of times k is adopt the induction objective function form when inferior iteration, and make I _i,tonly can value 0 or 1, build the mathematical model containing the monthly Unit Combination plan of Constraint, and adopt mixed integer programming solver to solve this model, obtain monthly Unit Combination plan;

(9) whether the deviation factors λ that judges monthly Unit Combination plan is less than deviation threshold λ ₀;

Calculate the deviation factors λ of monthly Unit Combination plan according to following expression:

$\mathrm{\&lambda;}=\frac{O_{P_0^{*}}-O_{P_S}}{O_{P_0^{*}}}$

Wherein, that step (8) solves the corresponding cost of electricity-generating of monthly Unit Combination plan obtaining; the lax corresponding cost of electricity-generating of monthly Unit Combination plan while being k=1; In fact permanent establishment, therefore 0≤λ < 1 is permanent sets up; The gap of the monthly Unit Combination plan of the monthly Unit Combination plan that λ sign step (8) obtains and optimum;

If λ > is λ ₀, continue step (10); If λ≤λ ₀, jump to step (11);

(10) adopt former objective function, the monthly Unit Combination plan obtaining taking step (8), as initial solution, builds and solves the mathematical model containing the monthly Unit Combination plan of Constraint, obtains monthly Unit Combination plan;

Refer to containing the mathematical model of the monthly Unit Combination plan of Constraint the mathematical model that step (2) builds, adopt mixed integer programming solver to solve this model, and the monthly Unit Combination that step (8) is obtained be intended to be initial solution implant solution procedure, with speed-up computation;

(11) optimize and finish, acquired results is the monthly Unit Combination plan containing Constraint, and grid company is controlled accordingly the start and stop of genset and exerted oneself.

So far, institute of the present invention extracting method is implemented complete.

Embodiment:

Set forth the monthly Unit Combination optimization method containing Constraint proposed by the invention as an example of certain provincial power network example, and verify the effect that the present invention realizes:

(1) prepare Unit Combination basic data;

Described Unit Combination basic data refers to that the operation characteristic data of genset, monthly load prediction data, each unit contract electric weight of following month, grid topology data, Optimal Parameters data etc. build the data of monthly Unit Combination plan mathematical model;

This provincial power network has 480 nodes, 111 genset.First obtain the operation characteristic data of genset, monthly load prediction data, each unit contract electric weight, grid topology data and the Optimal Parameters data of following month; Concrete data repeat no more;

(2) build the mathematical model containing the monthly Unit Combination plan of Constraint according to Unit Combination basic data;

(2-1) data that adopt step (1) to obtain, build the former objective function containing the monthly Unit Combination planning model of Constraint, and expression formula is as follows:

$\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]$

Wherein, represent the fuel cost of unit i at period t, represent the start expense of unit i at period t, represent the idleness expense of unit i at period t; N represents unit number, is 111 in an embodiment; T represents period number, is 31 in an embodiment; Wherein it is control variable;

(2-2) data that adopt step (1) to obtain, build the constraint condition containing the monthly Unit Combination planning model of Constraint, and expression formula is as follows respectively:

(2-2-1) fuel cost constraint condition

$\begin{array}{cc}{C}_{i,t}^P\&GreaterEqual;A_i^P{P}_{i,t}\mathrm{\&Delta;}{T}^D& \&ForAll;t,\&ForAll;i\end{array}$

(2-2-2) start expense restriction condition

$\begin{array}{cc}{C}_{i,t}^U\&GreaterEqual;A_i^U[I_{i,t}-I_{i,(t-1)}]& \&ForAll;t,\&ForAll;i\end{array}$

$\begin{array}{cc}{C}_{i,t}^U\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

(2-2-3) idleness expense constraint condition

$\begin{array}{cc}{C}_{i,t}^D\&GreaterEqual;A_i^D[I_{i,(t-1)}-I_{i,t}]& \&ForAll;t,\&ForAll;i\end{array}$

$\begin{array}{cc}{C}_{i,t}^D\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

Wherein, P _i,trepresent that unit i exerts oneself at the meritorious of period t, I _i,trepresent the start and stop state of unit i at period t, I _i,tvalue can only be 1 or 0, I _i,tbe 1 expression start, I _i,tbe that 0 expression is shut down; P _i,twith I _i,talso be control variable; Δ T ^dintersegmental interval time while representing every two; represent that unit i often sends the fuel cost that unit quantity of electricity needs; with represent respectively the expense that unit i starts shooting once and shut down expense once; Fuel cost constraint condition represents the fuel cost that unit i pays for sending power at period t; Start expense restriction condition represent when unit i at period t-1 the start expense during to period t transition, if unit i starts shooting in this transition period, start expense is if not start, start expense is 0; Idleness expense constraint condition represent when unit i at period t-1 the idleness expense during to period t transition, if unit i shuts down in this transition period, idleness expense is if do not shut down, idleness expense is 0;

(2-2-4) electric network active equilibrium constraint

$\begin{array}{cc}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{P}_{i,t}=D_t& \&ForAll;t\end{array}$

Wherein, D _trepresent the electrical network total load of period t, in this constraint representation electrical network, the generated output of all units equals electrical network total load;

(2-2-5) electrical network spinning reserve constraint condition

$\begin{array}{cc}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{I}_{i,t}{P}_i^{\max }\&GreaterEqual;D_t(1+R^P)& \&ForAll;t\end{array}$

Wherein, the maximum technology that represents unit i is exerted oneself, R ^prepresent electrical network spinning reserve rate, in this constraint representation electrical network, the capacity sum of all start units should be greater than the certain proportion of electrical network total load;

(2-2-6) unit minimum start/stop time of constraint condition

$\begin{array}{cc}(T_{i,t}^{\mathrm{on}}-T_{i,\min }^{\mathrm{on}})[I_{i,(t-1)}-I_{i,t}]\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

$T_{i,t}^{\mathrm{on}}=\underset{{k=t-T}_{i,\min }^{\mathrm{on}}}{\overset{t-1}{\mathrm{\&Sigma;}}}{I}_{i,k}$

$\begin{array}{cc}(T_{i,t}^{\mathrm{off}}-T_{i,\min }^{\mathrm{off}})[I_{i,t}-I_{i,(t-1)}]\&GreaterEqual;0& \&ForAll;t,\&ForAll;i\end{array}$

$T_{i,t}^{\mathrm{off}}=\underset{{k=t-T}_{i,\min }^{\mathrm{off}}}{\overset{t-1}{\mathrm{\&Sigma;}}}(1-I_{i,k})$

Wherein, with representing respectively on time and stop time that unit i had experienced continuously before period t, is control variable; with represent respectively minimum on time and the minimum stop time of unit i; Must experience after this constraint representation unit i start (shutdown) time could be shut down (start) again;

(2-2-7) unit output changes bound constraint condition

$P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }{I}_{i,(t-1)}+[I_{i,t}-I_{i,(t-1)}]P_i^{\min }$

$\begin{array}{cc}+P_i^{\max }(1-I_{i,t})& \&ForAll;t,\&ForAll;i\end{array}$

$P_{i,(t-1)}-P_{i,t}\&le;{\mathrm{DP}}_i^{\min }{I}_{i,t}-[I_{i,t}-I_{i,(t-1)}]P_i^{\min }$

$\begin{array}{cc}+P_i^{\max }[1-I_{i,(t-1)}]& \&ForAll;t,\&ForAll;i\end{array}$

Wherein, with represent respectively the upper and lower bound that unit output changes, with the maximum technology that represents respectively unit i is exerted oneself and minimum technology is exerted oneself, and determines by generator inherent characteristic; At period t-1, the meritorious variable quantity of exerting oneself during to period t transition should limit within the specific limits this constraint representation unit i; If the transition period of unit i from period t-1 to period t is start process, i.e. I _{i, (t-1)}=0 and I _i,t=1, unit output changes bound constraint condition and is and $P_{i,(t-1)}-P_{i,t}\&le;{\mathrm{DP}}_i^{\min }-P_i^{\min }+P_i^{\max },$ Due to I _{i, (t-1)}=0, therefore P _{i, (t-1)}=0, and with constraint condition (2-2-8) acting in conjunction, by P _i,tbe defined as first period after the start of expression unit exerts oneself and can only exert oneself for minimum technology now, for redundancy, actual inoperative; If the transition period of unit i from period t-1 to period t is stopping process, i.e. I _{i, (t-1)}=1 and I _i,t=0, unit output changes bound constraint condition and is $P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ And $P_{i,(t-1)}-P_{i,t}\&le;P_i^{\min },$ Now, $P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ For redundancy, actual inoperative, due to I _i,t=0, therefore P _i,t=0, and with constraint condition (2-2-8) acting in conjunction, by P _{i, (t-1)}be defined as last period before expression compressor emergency shutdown exerts oneself and can only exert oneself for minimum technology if unit i is from period t-1 to period t always in running status, i.e. I _{i, (t-1)}=1 and I _i,t=1, unit output changes bound constraint condition and is and represent that the bound that unit output changes is respectively with if unit i is from period t-1 to period t always in stopped status, i.e. I _{i, (t-1)}=0 and I _i,t=0, unit output changes bound constraint condition and is and all can be met;

(2-2-8) unit output bound constraint condition

$\begin{array}{cc}{I}_{i,t}{P}_i^{\min }\&le;P_{i,t}\&le;I_{i,t}{P}_i^{\max }& \&ForAll;t,\&ForAll;i\end{array}$

When this constraint representation unit start, it is exerted oneself and in certain scope, change; When compressor emergency shutdown, it is exerted oneself is 0;

(2-2-9) unit Constraint condition

$\begin{array}{cc}{W}_i^{\min }\&le;\mathrm{\&Delta;}{T}^D\&CenterDot;\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_{i,t}\&le;W_i^{\max }& \&ForAll;i\end{array}$

Wherein, with represent respectively electric weight lower limit and the upper limit of unit i, with be multiplied by lower deviation ratio by the contract electric weight of unit i respectively and upper deviation ratio obtains; The plan electric weight of this constraint representation unit within the moon equate substantially with its contract electric weight, and allow has deviation in certain scope;

(2-2-10) Network Security Constraints condition

$\begin{array}{cc}{-P}_s^{\max }\&le;\underset{l\&Element;s}{\mathrm{\&Sigma;}}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{G}_{l-i}{P}_{i,t}-\underset{l\&Element;s}{\mathrm{\&Sigma;}}\underset{j=1}{\overset{J}{\mathrm{\&Sigma;}}}{G}_{l-j}{D}_{j,t}\&le;P_s^{\max }& \&ForAll;t,\&ForAll;s\end{array}$

Wherein, represent the meritorious trend limit of transmission cross-section s, G _l-irepresent the meritorious transfer distribution factor of unit i to circuit l, G _l-jrepresent the meritorious transfer distribution factor of node load j to circuit l, D _j,trepresent the node load of node load j at period t, l ∈ s represents that circuit l belongs to transmission cross-section s, and J represents the number of node load; In this constraint representation electrical network, the meritorious trend of some transmission cross-section can not exceed its trend limit;

(3) by the start and stop variable I in step (2) model _i,tlax, make I _i,tvalue in desirable any [0,1], has built the mathematical model of lax monthly Unit Combination plan;

(4) use business mathematics planning to solve software CPLEX, adopt dual simplex algorithm (Dual Simplex Algorithm), solve the mathematical model of lax monthly Unit Combination plan, obtained lax monthly Unit Combination plan;

(5) the lax monthly Unit Combination of statistics in the works value be 0 or 1 start and stop variable I _i,tnumber H (1) be 1561;

(6) in the works, value is 0 or 1 start and stop variable I to the lax monthly Unit Combination that current iteration obtains _i,tnumber H (1)=1561, have increase compared with H (0)=0, therefore jump to step (7);

(7) according to start and stop variable I _i,tvalue and formula revise the objective function of lax monthly Unit Combination plan mathematical model, and jumped to step (4); Through 6 such iterative process, lax monthly Unit Combination in the works value is 0 or 1 start and stop variable I _i,tnumber H (k) no longer increase, be finally H (6)=3333;

(8) the induction objective function while adopting k=6, has built the mathematical model containing the monthly Unit Combination plan of Constraint; Use business mathematics planning to solve branch's cutting plane algorithm (Branch-and-cut Algorithm) of software CPLEX, solved this mathematical model, obtain monthly Unit Combination plan;

(9) whether the deviation factors that judges monthly Unit Combination plan is less than deviation threshold;

In the present embodiment, deviation threshold λ ₀be set as 1%; Through calculating, deviation factors λ=1.09%>1% that step (8) obtains, therefore, jumps to (10) step;

(10) adopt former objective function, build the mathematical model containing the monthly Unit Combination plan of Constraint, and use business mathematics planning to solve branch's cutting plane algorithm (Branch-and-cut Algorithm) of software CPLEX, the monthly Unit Combination plan obtaining taking step (8) is as initial solution, solve this mathematical model, obtain final monthly Unit Combination plan;

(11) optimize and finish, step (10) acquired results can be used as containing the monthly Unit Combination plan of Constraint and issues, and grid company can be controlled accordingly the start and stop of genset and exert oneself.

Promote for embodying counting yield of the present invention, the computing time and the cost of electricity-generating result that adopt the present invention to be optimized and directly to adopt business Optimization Software CPLEX to be optimized have been shown in table 1 contrast, and the form that is fuel consumption by cost of electricity-generating conversion in table 1 embodies.Directly adopt business Optimization Software CPLEX to be optimized and refer to the mathematical model containing the monthly Unit Combination of Constraint that adopts CPLEX software direct solution step (2) to build, obtain monthly Unit Combination plan, and without step (3)～(10).

The result that table 1 adopts the present invention to be optimized and is optimized with CPLEX contrasts

Fig. 2 (a) and Fig. 2 (b) have further contrasted employing the present invention and have been optimized and adopt CPLEX to be optimized, the situation that the target function value of best feasible solution changed with computing time.Best feasible solution refers in optimizing process, the solution of target function value minimum in current acquired all feasible solutions.Fig. 2 (a) and Fig. 2 (b) show, through the same optimization time, adopt the best feasible solution of the present invention's acquisition to be better than the best feasible solution that adopts CPLEX to obtain; While finally obtaining same optimum results, computing time of the present invention is much smaller than the computing time of CPLEX.

From above specific embodiment, the monthly Unit Combination optimization method containing Constraint that the present invention proposes, not losing under the prerequisite of optimizing precision, has improved approximately 35 times by counting yield.In the example of other provincial power networks, most effectively promote 50 times.According to method provided by the present invention, grid company can be optimized the monthly Unit Combination plan containing Constraint efficiently, optimum results meets power grid security and unit completes the actual demand of contract electric weight, therefore the Unit Combination plan that grid company can obtain according to optimization of the present invention, control the start and stop of genset and exert oneself, reaching the target of most optimum distribution of resources and energy-saving and emission-reduction.Embodiment illustrates that the present invention can meet the actual needs of grid company, has important practical significance and good application prospect.

It is worth mentioning that, the objective function in implementation step proposed by the invention can be selected flexibly as required and customize, and constraint condition can be added according to the actual requirements and delete, extensibility is strong.Therefore, the only unrestricted technical scheme of the present invention in order to explanation of above implementation step.Do not depart from any modification or partial replacement of spirit and scope of the invention, all should be encompassed in the middle of claim scope of the present invention.

Claims (1)

1. Based on induction objective function containing the monthly Unit Combination optimization method of Constraint, it is characterized in that, comprise the following steps:

   (1) obtain Unit Combination basic data;

   Described Unit Combination basic data refers to the data that the operation characteristic data of genset, monthly load prediction data, each unit contract electric weight, grid topology data and the monthly Unit Combination plan of the Optimal Parameters data construct mathematics model of following month need;

   The operation characteristic data of described genset comprise that fuel cost, start expense, idleness expense, minimum start/stop time, the variation upper limit/lower limit of exerting oneself, the min/max technology of genset go out force data;

   Described monthly load prediction data are following month electric load conditions of demand that obtain according to monthly load prediction software, comprise the total load data of following month each day day part electrical network, the node load data of the each node of day part;

   To be each unit exceed the quata or vacancy generated energy and the monthly electricity contract of having signed by the following moon cumulative in this month the described unit contract electric weight of following month; Accumulate mode is, if unit this month real generated energy exceeded this month contract electric weight, from the monthly electricity contract of following month, deduct; If unit this month, real generated energy was less than this month contract electric weight, in the monthly electricity contract of following month, supply;

   Described grid topology data comprises the meritorious trend limit and the circuit ID comprising, each genset and the node load meritorious transfer distribution factor data to every transmission line of electricity thereof of the node of electric power networks and the annexation of transmission line of electricity, each transmission cross-section;

   Described Optimal Parameters data comprise lower deviation ratio and upper deviation ratio, electrical network spinning reserve rate and the deviation threshold data of unit contract electric weight;

   (2) build the mathematical model containing the monthly Unit Combination plan of Constraint according to Unit Combination basic data;

   The described mathematical model containing the monthly Unit Combination plan of Constraint is made up of objective function and constraint condition;

   (2-1) build the objective function containing the monthly Unit Combination plan of Constraint mathematical model, expression formula is as follows:

   $\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]$

   Define this objective function for " former objective function "; represent the corresponding cost of electricity-generating of monthly Unit Combination plan; Wherein, represent the fuel cost of unit i at period t, represent the start expense of unit i at period t, represent the idleness expense of unit i at period t; N represents unit number, and T represents period number; Wherein, it is control variable;

   (2-2) build the constraint condition containing the monthly Unit Combination plan of Constraint mathematical model, expression formula is as follows respectively:

   (2-2-1) fuel cost constraint condition

   $C_{i,t}^P\&GreaterEqual;A_i^P{P}_{i,t}\mathrm{\&Delta;}{T}^D\&ForAll;t,\&ForAll;i$

   (2-2-2) start expense restriction condition

   $C_{i,t}^U\&GreaterEqual;A_i^U[I_{i,t}-I_{i,(t-1)}]\&ForAll;t,\&ForAll;i$

   $C_{i,t}^U\&GreaterEqual;0\&ForAll;t,\&ForAll;i$

   (2-2-3) idleness expense constraint condition

   $C_{i,t}^D\&GreaterEqual;A_i^D[I_{i,(t-1)}-I_{i,t}]\&ForAll;t,\&ForAll;i$

   $C_{i,t}^D\&GreaterEqual;0\&ForAll;t,\&ForAll;i$

   Wherein, P _i,trepresent that unit i exerts oneself at the meritorious of period t, I _i,trepresent the start and stop state of unit i at period t, I _i,tvalue can only be 1 or 0, I _i,tbe 1 expression start, I _i,tbe that 0 expression is shut down; P _i,twith I _i,talso be control variable; Δ T ^dintersegmental interval time while representing every two; represent that unit i often sends the fuel cost that unit quantity of electricity needs; with represent respectively the expense that unit i starts shooting once and shut down expense once; Fuel cost constraint condition represents the fuel cost that unit i pays for sending power at period t; Start expense restriction condition represent when unit i at period t-1 the start expense during to period t transition, if unit i starts shooting in this transition period, start expense is if not start, start expense is 0; Idleness expense constraint condition represent when unit i at period t-1 the idleness expense during to period t transition, if unit i shuts down in this transition period, idleness expense is if do not shut down, idleness expense is 0;

   (2-2-4) electric network active equilibrium constraint

   $\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{P}_{i,t}=D_t\&ForAll;t$

   Wherein, D _trepresent the electrical network total load of period t, in this constraint representation electrical network, the generated output of all units equals electrical network total load;

   (2-2-5) electrical network spinning reserve constraint condition

   $\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{I}_{i,t}{P}_i^{\max }\&GreaterEqual;D_t(1+R^P)\&ForAll;t$

   Wherein, the maximum technology that represents unit i is exerted oneself, R ^prepresent electrical network spinning reserve rate, in this constraint representation electrical network, the capacity sum of all start units should be greater than the certain proportion of electrical network total load;

   (2-2-6) unit minimum start/stop time of constraint condition

   $(T_{i,t}^{\mathrm{on}}-T_{i,\min }^{\mathrm{on}})[I_{i,(t-1)}-I_{i,t}]\&GreaterEqual;0\&ForAll;t,\&ForAll;i$

   $T_{i,t}^{\mathrm{on}}=\underset{k=t-T_{i,\min }^{\mathrm{on}}}{\overset{t-1}{\mathrm{\&Sigma;}}}{I}_{i,k}$

   $(T_{i,t}^{\mathrm{off}}-T_{i,\min }^{\mathrm{off}})\left[{I_{i,t}-I}_{i,(t-1)}\right]\&GreaterEqual;0\&ForAll;t,\&ForAll;i$

   $T_{i,t}^{\mathrm{off}}=\underset{k=t-T_{i,\min }^{\mathrm{off}}}{\overset{t-1}{\mathrm{\&Sigma;}}}(1-I_{i,k})$

   Wherein, with representing respectively on time and stop time that unit i had experienced continuously before period t, is control variable; with represent respectively minimum on time and the minimum stop time of unit i; Must experience after this constraint representation unit i start time could shut down again, must experience after shutdown time could start shooting again;

   (2-2-7) unit output changes bound constraint condition

   $\begin{array}{c}{P}_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }{I}_{i,(t-1)}+[I_{i,t}-I_{i,(t-1)}]P_i^{\min }\\ +P_i^{\max }(1-I_{i,t})\&ForAll;t,\&ForAll;i\end{array}$

   $\begin{array}{c}{P}_{i,(t-1)}-P_{i,t}\&le;{\mathrm{DP}}_i^{\min }{I}_{i,t}-[I_{i,t}-I_{i,(t-1)}]P_i^{\min }\\ +P_i^{\max }(1-I_{i,t(t-1)})\&ForAll;t,\&ForAll;i\end{array}$

   Wherein, with represent respectively the upper and lower bound that unit output changes, with the maximum technology that represents respectively unit i is exerted oneself and minimum technology is exerted oneself, and determines by generator inherent characteristic; At period t-1, the meritorious variable quantity of exerting oneself during to period t transition should limit within the specific limits this constraint representation unit i; If the transition period of unit i from period t-1 to period t is start process, i.e. I _{i, (t-1)}=0 and I _i,t=1, unit output changes bound constraint condition and is and $P_{i,(t-1)}-P_{i,t}\&le;{\mathrm{DP}}_i^{\min }-P_i^{\min }+P_i^{\max },$ Due to I _{i, (t-1)}=0, therefore P _{i, (t-1})=0, and $P_{i,t}-P_{i,(t-1)}\&le;P_i^{\min }$ With constraint condition (2-2-8) acting in conjunction, by P _i,tbe defined as first period after the start of expression unit exerts oneself and can only exert oneself for minimum technology now, for redundancy, actual inoperative; If the transition period of unit i from period t-1 to period t is stopping process, i.e. I _{i, (t-1)}=1 and I _i,t=0, unit output changes bound constraint condition and is $P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ And $P_{i,(t-1)}-P_{i,t}\&le;P_i^{\min },$ Now, $P_{i,t}-P_{i,(t-1)}\&le;{\mathrm{DP}}_i^{\max }-P_i^{\min }+P_i^{\max }$ For redundancy, actual inoperative, due to I _i,t=0, therefore P _i,t=0, and with constraint condition (2-2-8) acting in conjunction, by P _{i, (t-1)}be defined as last period before expression compressor emergency shutdown exerts oneself and can only exert oneself for minimum technology if unit i is from period t-1 to period t always in running status, i.e. I _{i, (t-1)}=1 and I _i,t=1, unit output changes bound constraint condition and is and represent that the bound that unit output changes is respectively with if unit i is from period t-1 to period t always in stopped status, i.e. I _{i, (t-1)}=0 and I _i,t=0, unit output changes bound constraint condition and is and all can be met;

   (2-2-8) unit output bound constraint condition

   $I_{i,t}{P}_i^{\min }\&le;P_{i,t}\&le;I_{i,t}{P}_i^{\max }\&ForAll;t,\&ForAll;i$

   When this constraint representation unit start, it is exerted oneself and in certain scope, change; When compressor emergency shutdown, it is exerted oneself is 0;

   (2-2-9) unit Constraint condition

   $W_i^{\min }\&le;\mathrm{\&Delta;}{T}^D\&CenterDot;\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}{P}_{i,t}\&le;W_i^{\max }\&ForAll;i$

   Wherein, with represent respectively electric weight lower limit and the upper limit of unit i, with be multiplied by lower deviation ratio by the contract electric weight of unit i respectively and upper deviation ratio obtains; The plan electric weight of this constraint representation unit within the moon equate substantially with its contract electric weight, and allow has deviation in certain scope;

   (2-2-10) Network Security Constraints condition

   ${-P}_s^{\max }\&le;\underset{l\&Element;s}{\mathrm{\&Sigma;}}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{G}_{l-i}{P}_{i,t}-\underset{l\&Element;s}{\mathrm{\&Sigma;}}\underset{j=1}{\overset{J}{\mathrm{\&Sigma;}}}{G}_{l-j}{D}_{j,t}\&le;P_s^{\max }\&ForAll;t,\&ForAll;s$

   Wherein, represent the meritorious trend limit of transmission cross-section s, G _l-irepresent the meritorious transfer distribution factor of unit i to circuit l, G _l-jrepresent the meritorious transfer distribution factor of node load j to circuit l, D _j,trepresent the node load of node load j at period t, l ∈ s represents that circuit l belongs to transmission cross-section s, and J represents the number of node load; In this constraint representation electrical network, the meritorious trend of some transmission cross-section can not exceed its trend limit;

   (3) by the I containing in the monthly Unit Combination plan of Constraint mathematical model _i,trelax, make I _i,tany value in desirable [0,1], the mathematical model of the lax monthly Unit Combination plan of structure;

   (4) adopt linear programming for solution device, solve the mathematical model of lax monthly Unit Combination plan, obtain lax monthly Unit Combination plan;

   (5) in the works, statistics value is 0 or 1 start and stop variable I to the lax monthly Unit Combination obtaining in step (4) _i,tnumber, be designated as H (k), the sequence number that k is current iteration, and define H (0)=0;

   (6) judge lax monthly Unit Combination that current iteration obtains in the works value be 0 or 1 start and stop variable I _i,tnumber H (k) whether increase compared with H (k-1); If increase, proceed step (7); If do not increase, jump to step (8);

   (7) according to start and stop variable I _i,tvalue, build the induction objective function of lax monthly Unit Combination plan mathematical model;

   The former objective function of the mathematical model of lax monthly Unit Combination plan is revised as to the induction objective function of following form:

   $\mathrm{Min}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}[C_{i,t}^P+C_{i,t}^U+C_{i,t}^D]+\underset{i}{\mathrm{\&Sigma;}}\underset{t}{\mathrm{\&Sigma;}}{\mathrm{\&xi;}}_{i,t}^{(k+1)}{I}_{i,t}$

   Wherein, represent lax monthly Unit Combination that the k time iteration obtain in the works unit i at the start and stop variate-value of period t, B _irepresent fuel cost when unit i operates in minimum technology and exerts oneself; Especially, when time, order wherein ε=0.01, M is B in all units _imaximal value;

   Jump to step (4), iterations k adds 1;

   (8) adopt induction objective function, build and solve the mathematical model containing the monthly Unit Combination plan of Constraint, obtain monthly Unit Combination plan;

   If the value of current iteration number of times k is adopt the induction objective function form when inferior iteration, and make I _i,tonly can value 0 or 1, build the mathematical model containing the monthly Unit Combination plan of Constraint, and adopt mixed integer programming solver to solve this model, obtain monthly Unit Combination plan;

   (9) whether the deviation factors λ that judges monthly Unit Combination plan is less than deviation threshold λ ₀;

   Calculate the deviation factors λ of monthly Unit Combination plan according to following expression:

   $\mathrm{\&lambda;}=\frac{O_{P_0^{*}}-O_{P_S}}{O_{P_0^{*}}}$

   Wherein, that step (8) solves the corresponding cost of electricity-generating of monthly Unit Combination plan obtaining; the lax corresponding cost of electricity-generating of monthly Unit Combination plan while being k=1;

   If λ > is λ ₀, continue step (10); If λ≤λ ₀, jump to step (11);

   (10) adopt former objective function, the monthly Unit Combination plan obtaining taking step (8), as initial solution, builds and solves the mathematical model containing the monthly Unit Combination plan of Constraint, obtains monthly Unit Combination plan;

   (11) optimize and finish, acquired results is the monthly Unit Combination plan containing Constraint, and grid company is controlled accordingly the start and stop of genset and exerted oneself.