CN104158232B - A kind of generation schedule optimization method considering unit regulation dead band - Google Patents

Abstract

The invention discloses a kind of generation schedule optimization method considering unit regulation dead band, method includes that (1) safeguards and provides the service data needed for generation schedule Optimized model, safeguards the minimum adjustment amount of unit；(2) obtain up-to-date electrical network physical model and real-time mode data, start Calculation of Sensitivity, obtain sensitivity results information；(3) generation schedule Optimized model reads system-based data, minimum adjustment amount and sensitivity results information；(4) all input data read according to step (3), on the basis of generation schedule Optimized model, introduce the minimum adjustment amount constraint of unit, are i.e. modeled minimum adjustment amount constraint, form the Optimized model considering unit regulation dead band.The present invention, by introducing the mathematical model of unit minimum adjustment amount constraint, works out the higher unit output plan of relatively reliable effective degree of being practical, improves unit plan implementation rate, alleviate AGC frequency modulation unit and the regulation pressure of AGC Autonomous Control unit.

Description

A kind of generation schedule optimization method considering unit regulation dead band

Technical field

The present invention relates to a kind of generation schedule optimization method considering unit regulation dead band, belong to electricity Force system dispatching automation technical field.

Background technology

Currently, build prevalence at home with intelligent grid, pre-according to need for electricity next day Survey, it is considered to system balancing constraint, unit operation constraint, operation of power networks constraint, network security are about The various constraintss such as bundle, optimization establishment 96 Unit Combination plans next day and the plan of exerting oneself Generation schedule practical function degree improves the most comprehensively a few days ago.

Plan in practical implementation a few days ago, due to system loading prediction a few days ago and new energy a few days ago Source prediction and physical presence relatively large deviation, cause the plan of a few days ago exerting oneself significantly to adjust, to AGC Unit regulation brings immense pressure, also reduces power grid security index simultaneously.For solving this problem, Real-time generation schedule establishment is brought into schedule, and real-time generation schedule programmed function is according to up-to-date electrical network The change of changes of operating modes, operating states of the units, super short period load requirement forecasting, ultra-short term are new Energy forecast, real-time section limit and planned by electricity, with reference to generation schedule a few days ago, work out future The unit generation plan of multiple periods.Real-Time Scheduling plan with plan identical a few days ago, support energy-conservation Scheduling, electricity market, three 3 kinds of scheduling methods of public scheduling.Plan period interval is 5 points in real time Clock (or 15 minutes), the time range of planning is following 1 to a few hours.Count a few days ago Draw Unit Combination state fixed on the basis of, plan in real time can cycle rolling amendment unit a few days ago The plan of exerting oneself, but Unit Combination state is not adjusted, i.e. only include Unit Economic scheduling merit Can, do not include Unit Combination function, planning considers transmission cross-section security constraint, refers to Surely transmission of electricity element security constraint.

Planning basic data accuracy is high in real time, and ultra-short term prediction data is inclined with real data Difference is the least, and plan implementation rate of exerting oneself increases substantially, and while easing off the pressure for AGC, more can The safety of accurate evaluation 1 to a few hours of future operation of power networks.But, plan and real-time a few days ago After generation schedule and AGC closed loop, plan a few days ago is passed through with the quality of real-time planning result AGC obtains embodying more intuitively.

Due to a few days ago plan and real-time planning do not consider unit regulate dead band, therefore, if Plan a few days ago and adjacent 2 adjustment amounts of exerting oneself of real-time planning unit output are dead less than regulation District, then AGC not assigns regulating command, cause a few days ago with the unit output of real-time planning It is not carried out.Assume that real-time planning result occurs that a large amount of unit output is finely tuned, and unit Regulated quantity is less than regulation dead band, but total regulated quantity is bigger than normal, if following the tracks of a few days ago and in real time in AGC The unit that planned outcome is exerted oneself not performs, and arises that bulk power deviation, for ensureing system Balance, is undertaken by the unit of system frequency modulation unit and AGC Autonomous Control, and this is just to system frequency modulation The unit of unit and AGC Autonomous Control brings the biggest adjustment pressure.

Summary of the invention

The deficiency existed for prior art, it is an object of the present invention to provide a kind of consideration unit regulation The generation schedule optimization method in dead band, by introducing the mathematical model of unit minimum adjustment amount constraint, Work out the higher unit output plan of relatively reliable effective degree of being practical, improve unit meter Draw implementation rate, alleviate AGC frequency modulation unit and the regulation pressure of AGC Autonomous Control unit.

To achieve these goals, the present invention is to realize by the following technical solutions:

A kind of generation schedule optimization method considering unit regulation dead band of the present invention, specifically includes Following step:

(1) safeguard (utilizing existing maintenance tool) and generation schedule Optimized model institute is provided The service data needed, safeguards the minimum adjustment amount data of (being responsible for by maintenance interface) unit simultaneously；

(2) up-to-date electrical network physical model and real-time mode number are obtained by available data interface According to, start Calculation of Sensitivity, obtain full-mesh network and monitor that the sensitivity results of unit is believed by element Breath；

(3) generation schedule Optimized model by optimize data-interface read system-based data, The minimum adjustment amount data of described step (1) and described step (2) calculated sensitivity Object information；

(4) all input data read according to described step (3), in generation schedule optimization On the basis of model, introduce the minimum adjustment amount constraint of unit, i.e. minimum adjustment amount is retrained into Row modeling, forms the Optimized model considering unit regulation dead band, finally exports object information；Institute State minimum adjustment amount constraint to refer to if the variable quantity absolute value of exerting oneself of adjacent two periods is more than Zero, then (adjacent 2 may change, it is also possible to do not become to should be greater than minimum adjustment amount data Change, change, then the absolute value changed should be more than regulation dead band i.e. more than minimum adjustment amount number According to, without change, the most do not consider minimum adjustment amount constraint).

In step (2), described Calculation of Sensitivity specifically uses PQ decoupling method.

In step (3), described system-based data include that system loading prediction, new forms of energy are pre- Survey, repair schedule, section limit, compressor emergency shutdown information, interconnection plan, unit may bring up Power, Unit Economic parameter information and unit reference plan.

Above-mentioned generation schedule Optimized model uses security constraint Unit Combination and security constraint Economic load dispatching Optimized model.

In step (4), the modeling method of described minimum adjustment amount constraint is as follows:

Owing to the change direction of exerting oneself of unit adjacent two periods is indefinite, it may be possible to upper adjustment, also It is probably lower adjustment, therefore, introduces the adjustment amount up and down of non-negative, then adjustment amount and adjacent up and down The exert oneself relationship description of variable quantity of two periods is:

$p_{i,t}-p_{i,t-1}=p_{i,t}^{+}-p_{i,t}^{-}$

Wherein, p_i,tFor exerting oneself of conventional power unit i moment t,Exert oneself for unit i moment t Upper adjustment amount,Exert oneself lower adjustment amount for unit i moment t；

If upper adjustment amount or lower adjustment amount are more than zero, then should be greater than minimum adjustment amount data；

If upper adjustment amount or lower adjustment amount are zero, then without satisfied minimum adjustment amount data, need Whether be the flag bit of zero, i.e. 0/1 variable if introducing upper and lower adjustment amount, now, and minimum adjustment amount Constraint can be expressed as:

$p_{i,t}^{+}\≥h_{i,t}^{+}{m}_{i,t}$

$p_{i,t}^{-}\≥h_{i,t}^{-}{m}_{i,t}$

$p_{i,t}^{+}\≤h_{i,t}^{+}(p_{i,\max }-p_{i,\min })$

$p_{i,t}^{-}\≤h_{i,t}^{-}(p_{i,\max }-p_{i,\min })$

Wherein:Being 0/1 variable, unit i moment t exerts oneself rise integral denotation；For 0/1 variable, unit i moment t exerts oneself downward integral denotation；m_i,tExert oneself for unit i moment t Little adjustment amount；p_i,maxAnd p_i,minBe respectively unit i maximum, minimum technology is exerted oneself；

Above-mentioned minimum adjustment amount constraint expression formula associates upper and lower adjustment amount with adjustment mark up and down Getting up, when upper adjustment amount of exerting oneself is more than zero, raising integral denotation is 1, is otherwise zero；When going out When under power, adjustment amount is more than zero, lowering integral denotation is 1, is otherwise zero；When raising integral denotation it is When 1, then going up adjustment amount and be more than zero, when raising integral denotation and being 0, then going up adjustment amount is zero； Instantly adjust when being masked as 1, then descend adjustment amount more than zero, when adjustment is masked as 0 instantly, then Lower adjustment amount is zero；

For same unit, synchronization has and only one of which adjustment direction, i.e. raises Adjustment whole, lower and not adjusting, therefore, need to meet adjustment direction unique constraints:

$h_{i,t}^{+}+h_{i,t}^{-}\≤1$

So far, i.e. complete the modeling of the minimum adjustment amount constraint of unit, if adjacent two periods Variable quantity of exerting oneself more than zero, change adjustment amount must be more than regulation dead band.

The present invention, on the basis of existing Optimized model is carried out, introduces minimum adjustment amount constraint, improves Optimize the establishment reliability of unit plan and practical level, for AGC frequency modulation unit and AGC from Main control unit eases off the pressure, and the AGC simultaneously improving system regulates standby nargin, it is ensured that electricity Net safe and stable operation；The modeling method of the present invention has calculating intensity spy low, adaptable Point, is suitable for the scheduling institution popularization and application in the various rank of China.

Accompanying drawing explanation

Fig. 1 is a kind of generation schedule optimization method work considering unit regulation dead band of the present invention Make flow chart.

Detailed description of the invention

For the technological means making the present invention realize, creation characteristic, reach purpose and be prone to bright with effect White understanding, below in conjunction with detailed description of the invention, is expanded on further the present invention.

For making to plan and planning in real time and AGC preferably coordinate a few days ago, improve and count a few days ago Drawing the execution level exerted oneself with real-time planning result, plan and real-time planning are excellent a few days ago Change model and introduce the constraint of unit minimum adjustment amount, provide for AGC that degree of being practical is higher exerts oneself Plan, improves plan a few days ago and planning in real time and the closed loop level of AGC.

Present invention physical model statistic property based on electrical network, economic model parameter, network topology data, Load prediction data, wind power prediction data, Unit Combination status data, unit regulation dead band Data, on the basis of optimization routine model, introduce the constraint of unit minimum adjustment amount, develop The unit output plan of relatively reliable effective practical higher level, promotes a few days ago and counts in real time Draw the unit plan implementation rate of establishment, alleviate AGC frequency modulation unit and AGC Autonomous Control unit Regulation pressure, it is ensured that electric power netting safe running.

See Fig. 1, the generation schedule optimization method considering unit regulation dead band of the present invention, tool Body includes following step:

(1) provide data maintenance instrument, safeguard according to the physical attribute that each unit is different and repair Change unit regulation dead band.

(2) obtain up-to-date electrical network physical model and real-time mode data, start Calculation of Sensitivity Obtain monitoring that element is to unit sensitivity information；

(3) read system loading prediction, new forms of energy prediction, repair schedule, section limit, Compressor emergency shutdown information, interconnection plan, unit is adjustable exerts oneself, Unit Economic parameter information, machine It is sensitive that unit regulation dead band that group reference plan, step (1) are safeguarded and step (2) calculate The data such as degree object information.

(4) according to step (3) all inputs data, mould is optimized at conventional SCUC and SCED On the basis of type, introduce the constraint of unit minimum adjustment amount, formed and consider the excellent of unit regulation dead band Change model, finally export object information.

In step (4), the modeling method of described minimum adjustment amount constraint is as follows:

Owing to the change direction of exerting oneself of unit adjacent two periods is indefinite, it may be possible to upper adjustment, also It is probably lower adjustment, therefore, introduces the adjustment amount up and down of non-negative, then adjustment amount and adjacent up and down The exert oneself relationship description of variable quantity of two periods is:

$p_{i,t}-p_{i,t-1}=p_{i,t}^{+}-p_{i,t}^{-}$

Wherein, p_i,tFor exerting oneself of conventional power unit i moment t,Exert oneself for unit i moment t Upper adjustment amount,Exert oneself lower adjustment amount for unit i moment t；

If upper adjustment amount or lower adjustment amount are more than zero, then should be greater than minimum adjustment amount data；

If upper adjustment amount or lower adjustment amount are zero, then without satisfied minimum adjustment amount data, need Whether be the flag bit of zero, i.e. 0/1 variable if introducing upper and lower adjustment amount, now, and minimum adjustment amount Constraint can be expressed as:

$p_{i,t}^{+}\≥h_{i,t}^{+}{m}_{i,t}$

$p_{i,t}^{-}\≥h_{i,t}^{-}{m}_{i,t}$

$p_{i,t}^{+}\≤h_{i,t}^{+}(p_{i,\max }-p_{i,\min })$

$p_{i,t}^{-}\≤h_{i,t}^{-}(p_{i,\max }-p_{i,\min })$

Wherein:Being 0/1 variable, unit i moment t exerts oneself rise integral denotation；For 0/1 variable, unit i moment t exerts oneself downward integral denotation；m_i,tExert oneself for unit i moment t Little adjustment amount；p_i,maxAnd p_i,minBe respectively unit i maximum, minimum technology is exerted oneself；

Above-mentioned minimum adjustment amount constraint expression formula associates upper and lower adjustment amount with adjustment mark up and down Getting up, when upper adjustment amount of exerting oneself is more than zero, raising integral denotation is 1, is otherwise zero；When going out When under power, adjustment amount is more than zero, lowering integral denotation is 1, is otherwise zero；When raising integral denotation it is When 1, then going up adjustment amount and be more than zero, when raising integral denotation and being 0, then going up adjustment amount is zero； Instantly adjust when being masked as 1, then descend adjustment amount more than zero, when adjustment is masked as 0 instantly, then Lower adjustment amount is zero；

For same unit, synchronization has and only one of which adjustment direction, i.e. raises Adjustment whole, lower and not adjusting, therefore, need to meet adjustment direction unique constraints:

$h_{i,t}^{+}+h_{i,t}^{-}\≤1$

So far, i.e. complete the modeling of the minimum adjustment amount constraint of unit, if adjacent two periods Variable quantity of exerting oneself more than zero, change adjustment amount must be more than regulation dead band.

Unit minimum adjustment amount of the present invention retrains, it is achieved consider while establishment unit output plan The operation characteristic in unit regulation dead band, it is to avoid unit adjustment amount is more than zero and regulates dead less than unit District；Improve the reliability of unit output plan, provide more practical unit to go out for AGC Power plan, improves a few days ago and plans the operation with closed ring level with AGC in real time.

The present invention is applied at some regional power grid and provincial power network, and application effect meets expection. Actual application shows, the present invention can be in conjunction with unit actual motion characteristic, and working out relatively reliable has The degree of being practical higher unit output plan of effect, promotes unit plan implementation rate, alleviates AGC frequency modulation unit and the regulation pressure of AGC Autonomous Control unit, improve and plan and real-time a few days ago Plan and the closed loop level of AGC, meet the bulk power grid safe operation demand of lean day by day.

Checking that the present invention carries out under actual electric network data and analysis, it was demonstrated that practical constraint The reasonability of modeling and effectiveness.The present invention, on the basis of optimization routine model, considers each Class operation of power networks retrains, and is simultaneously introduced the constraint of unit minimum adjustment amount, says in a sense, It is the once correction to conventional model unit plan so that revised unit output plan more may be used Lean on, more effectively, more meet the requirement of actual motion, be favorably improved plan a few days ago and send out in real time The intelligent level of electricity scheduling and decision-making capability.Meanwhile, this modeling method have calculating intensity low, Adaptable feature, is more suitable for the scheduling institution popularization and application of the various scale of China.

The ultimate principle of the present invention and principal character and the present invention excellent has more than been shown and described Point.Skilled person will appreciate that of the industry, the present invention is not restricted to the described embodiments, on State the principle that the present invention is simply described described in embodiment and description, without departing from the present invention On the premise of spirit and scope, the present invention also has various changes and modifications, and these change and change Enter to both fall within scope of the claimed invention.Claimed scope is by appended power Profit claim and equivalent thereof define.

Claims (5)

1. the generation schedule optimization method considering unit regulation dead band, it is characterised in that Specifically include following step:

(1) safeguard and the service data needed for generation schedule Optimized model is provided, safeguarding simultaneously The minimum adjustment amount data of unit；

(2) obtain up-to-date electrical network physical model and real-time mode data by data-interface, open Dynamic Calculation of Sensitivity, obtains full-mesh network and monitors the element sensitivity results information to unit；

(3) generation schedule Optimized model by optimize data-interface read system-based data, The minimum adjustment amount data of described step (1) and described step (2) calculated sensitivity Object information；

(4) all input data read according to described step (3), in generation schedule optimization On the basis of model, introduce the minimum adjustment amount constraint of unit, i.e. minimum adjustment amount is retrained into Row modeling, forms the Optimized model considering unit regulation dead band, finally exports object information；Institute State minimum adjustment amount constraint to refer to if the variable quantity absolute value of exerting oneself of adjacent two periods is more than Zero, then should be greater than minimum adjustment amount data；

In step (4), the modeling method of described minimum adjustment amount constraint is as follows:

Owing to the change direction of exerting oneself of unit adjacent two periods is indefinite, it may be possible to upper adjustment, also It is probably lower adjustment, therefore, introduces the adjustment amount up and down of non-negative, then adjustment amount and adjacent up and down The exert oneself relationship description of variable quantity of two periods is:

$p_{i,t}-p_{i,t-1}=p_{i,t}^{+}-p_{i,t}^{-}$

Wherein, p_i,tFor exerting oneself of conventional power unit i moment t,Exert oneself for unit i moment t Upper adjustment amount,Exert oneself lower adjustment amount for unit i moment t；

If upper adjustment amount or lower adjustment amount are more than zero, then should be greater than minimum adjustment amount data；

If upper adjustment amount or lower adjustment amount are zero, then without satisfied minimum adjustment amount data, need Whether be the flag bit of zero, i.e. 0/1 variable if introducing upper and lower adjustment amount, now, and minimum adjustment amount Constraint can be expressed as:

$p_{i,t}^{+}\≥h_{i,t}^{+}{m}_{i,t}$

$p_{i,t}^{-}\≥h_{i,t}^{-}{m}_{i,t}$

$p_{i,t}^{+}\≤h_{i,t}^{+}(p_{i,max}-p_{i,min})$

$p_{i,t}^{-}\≤h_{i,t}^{-}(p_{i,max}-p_{i,min})$

Wherein:Being 0/1 variable, unit i moment t exerts oneself rise integral denotation；For 0/1 variable, unit i moment t exerts oneself downward integral denotation；m_i,tExert oneself for unit i moment t Little adjustment amount；p_i,maxAnd p_i,minBe respectively unit i maximum, minimum technology is exerted oneself；

Above-mentioned minimum adjustment amount constraint expression formula associates upper and lower adjustment amount with adjustment mark up and down Getting up, when upper adjustment amount of exerting oneself is more than zero, raising integral denotation is 1, is otherwise zero；When going out When under power, adjustment amount is more than zero, lowering integral denotation is 1, is otherwise zero；When raising integral denotation it is When 1, then going up adjustment amount and be more than zero, when raising integral denotation and being 0, then going up adjustment amount is zero； Instantly adjust when being masked as 1, then descend adjustment amount more than zero, when adjustment is masked as 0 instantly, then Lower adjustment amount is zero；

For same unit, synchronization has and only one of which adjustment direction, i.e. raises Adjustment whole, lower and not adjusting, therefore, need to meet adjustment direction unique constraints:

$h_{i,t}^{+}+h_{i,t}^{-}\≤1$

So far, i.e. complete the modeling of the minimum adjustment amount constraint of unit, if adjacent two periods Variable quantity of exerting oneself more than zero, change adjustment amount must be more than regulation dead band.

The generation schedule optimization side in consideration unit the most according to claim 1 regulation dead band Method, it is characterised in that

In step (2), described Calculation of Sensitivity specifically uses PQ decoupling method.

The generation schedule optimization side in consideration unit the most according to claim 1 regulation dead band Method, it is characterised in that

In step (3), described system-based data include that system loading prediction, new forms of energy are pre- Survey, repair schedule, section limit, compressor emergency shutdown information, interconnection plan, unit may bring up Power, Unit Economic parameter information and unit reference plan.

The generation schedule optimization side in consideration unit the most according to claim 1 regulation dead band Method, it is characterised in that

Described generation schedule Optimized model uses security constraint Unit Combination and security constraint Economic load dispatching Optimized model.

The generation schedule optimization side in consideration unit the most according to claim 2 regulation dead band Method, it is characterised in that

Described generation schedule Optimized model uses security constraint Unit Combination and security constraint Economic load dispatching Optimized model.