CN108539781A - Improve the extension black-start scheme bi-level programming optimization method of recovery process safety - Google Patents
Improve the extension black-start scheme bi-level programming optimization method of recovery process safety Download PDFInfo
- Publication number
- CN108539781A CN108539781A CN201810269185.4A CN201810269185A CN108539781A CN 108539781 A CN108539781 A CN 108539781A CN 201810269185 A CN201810269185 A CN 201810269185A CN 108539781 A CN108539781 A CN 108539781A
- Authority
- CN
- China
- Prior art keywords
- unit
- black
- starting
- level programming
- recovery process
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses the extension black-start scheme bi-level programming optimization methods for improving recovery process safety, recovery control for initial stage after having a power failure on a large scale provides new thinking, to improve safety of the extension black-start scheme in practical recovery process, the safety factor for influencing recovery process is brought into optimization aim, in view of principal and subordinate's logical relation that unit restores and path is restored, extension black-start scheme bi-level programming Optimized model is established；With scheme, overall efficiency is up to target under the conditions of high security for upper layer optimization, decision variable is to be activated unit, lower layer, which optimizes to seek, to be included critical circuits and restores safe while taking into account the restoration path restored conducive to important load as target, and decision variable is circuit to be restored；Model is solved with the hybrid algorithm that the critical path method (CPM) based on Dijkstra's algorithm is combined using bacterial foraging algorithm is improved.Recovery process has higher safety, improves recovery effects, has practical engineering application value.
Description
Technical field
The present invention relates to the extension black-start scheme bi-level programming optimization methods for improving recovery process safety, belong to electric power
Systems technology field.
Background technology
In recent years, intelligent grid construction has played important function to improving operation of power networks reliability, but power grid is various interior
Being still within the bounds of possibility for large-scale blackout occurs under the influence of portion's failure and external disturbance.The research of power system restoration
It can accelerate the recovering process of system after having a power failure on a large scale, reduce the have a power failure huge economic losses brought and social influence, be electric system
One of important topic of Prevention-Security.
It, usually can be the system recovery procedure after having a power failure on a large scale according to the task and feature of different times in system recovery procedure
It is divided into black starting-up, rack reconstruct and load restoration three phases.The black starting-up stage is the starting stage restored, and is activated unit
Power can be provided after successfully starting up for follow-up recovery and supports that the small network in part of formation is the basis subsequently restored, but must be with
Be activated premised on the successful recovery of unit, if be activated unit restore failure, can delay system recovering process, cause more
Big loss.In the entire recovery process in black starting-up stage, line over-voltage, black starting-up power supply self-excitation, sky are filled by sky
It fills transformer generation excitation surge current and overvoltage, set auxiliary machinery startup causes the operational risks such as system voltage declines and frequency is fallen
It restricts, any link, which is unsatisfactory for limitation, can then cause to restore unsuccessfully to prevent to be activated unit from successfully starting up.Therefore, extensive in formulation
When compound case, it is necessary to ensure that recovery process has higher-security, so that unit successfully restores, but can not still accomplish at present
The requirement of compromise between security while ensureing to optimize recovery efficiency.
Invention content
The technical problem to be solved by the present invention is to overcome the deficiencies of existing technologies, provides and improve recovery process safety
Black-start scheme bi-level programming optimization method is extended, brings the correlative factor for influencing recovery process safety into extension black starting-up side
In case optimization aim, the extension black-start scheme bilevel programming model for considering recovery process safety is established, program decisions is made
The requirement of compromise between security while pursuing unit starting rear weight generating efficiency maximum.Using improve bacterial foraging algorithm and
The hybrid algorithm that critical path method (CPM) based on Dijkstra's algorithm is combined is solved, and is utilized and is changed to upper layer Non-Linear Programming
Optimize into bacterial foraging algorithm, using the minimum support tree optimization lower layer restoration path based on Dijkstra's algorithm, fully sends out
The advantage of two kinds of algorithms is waved.
In order to solve the above technical problems, the present invention uses following technical scheme.
Improve the extension black-start scheme bi-level programming optimization method of recovery process safety, characterized in that establish extension
Black-start scheme bi-level programming Optimized model, respectively using unit starting and restoration path as the change of the upper and lower Optimal Decision-making
Amount, establishes bi-level programming Optimized model constraints, most using improvement bacterial foraging algorithm and based on Dijkstra's algorithm
The hybrid solving algorithm that short-circuit shot is combined solves bi-level programming Optimized model.
Upper layer Optimized model establishment step is：
If the unit number being activated is NG, it is P to be activated unit i subsidiary engine capacityGL.i, PLItem is constrained to meet system operation
Recoverable maximal workload under part, consider unit location prominence, unit generation amount and efficiency, subsidiary engine start influence and
The feedback of lower layer's restoration path optimization optimizes unit starting, and object function is：
In formula：F is upper layer target, including set optimization and lower layer's object function penalty term two parts；T1To optimize the time；
t0For the recovery time in extension black starting-up path；aiThe pitch point importance where unit i, with the network cohesion degree after node contraction
It indicates；W (t) is the weighting coefficient of unit output in different periods；PGi(t) active power sent out in t moment for unit i；γ
For transformation coefficient；η is penalty coefficient；F is lower layer's target function value.
Upper layer Optimized model constraints include startup power constraint, unit starting time-constrain, set auxiliary machinery capacity about
Beam and unit output constraint.
The establishment step of lower layer's Optimized model is：
The object function of lower layer restoration path optimization is：
In formula：NlFor the circuitry number of recovery；BjFor the susceptance of branch j；Binary decision variable λjIndicate whether branch j is to become
Depressor branch is to take 1, no to take 0；JjFor the betweenness after branch j normalization；θ(Lj) be the both ends branch j plant stand load importance,
It is set according to load actual conditions by dispatching of power netwoks department, β1、β2、β3、β4For conversion coefficient.
Lower layer's Optimized model constraints include black starting-up power supply self-excitation magnetic confinement, System Reactive Power constraint, line power and
Node voltage constrains.
Establishing bi-level programming Optimized model constraints expression formula is：
In formula：Pcr.iTo be activated the startup power of unit i；K2For coefficient of reliability；P0It is opened for what black starting-up power supply provided
Dynamic power；TS.iFor the startup time of unit i；TCH.iFor the maximum critical thermal starting time of unit i；PGi、QGiFor having for unit i
Work(, idle output；For unit active power output upper and lower limit；For the idle output upper and lower limit of unit；QLj
It is generated for circuit j idle；KCBFor black starting-up unit short-circuit ratio；SBFor black starting-up power supply capacity；QB.maxFor black starting-up power supply energy
The maximum of absorption is idle；PjIt is active on circuit j；It is maximum active on circuit j；UiFor node i voltage,For
Node i voltage upper and lower limit；NlFor the circuit number of recovery.
The hybrid algorithm being combined with the critical path method (CPM) based on Dijkstra's algorithm using bacterial foraging algorithm is improved
When solution, using the unit starting improved in bacterial foraging algorithm solution upper layer Optimal Decision-making, calculated using based on Di Jiesitela
The Minimal Spanning Tree of method solves the optimal restoration path in lower layer's Optimal Decision-making.
Solve unit starting the step of be：
Bacterial foraging algorithm step is：
1) chemotactic operates：If bacterial population scale is S, search space is tieed up for D, and the position of bacterium k can use D dimension space vectorsIt indicates, the position of each bacterium represents a solution of problem；θk(l, m, n) indicates that bacterium k becomes at the l times
Change the m times position replicated after operation and n-th migration operation of operation, the chemotactic operation of each steps of bacterium k is expressed as：
θk(l+1, m, n)=θk(l,m,n)+c(d)φ(r) (9)
In formula：C (d) indicates the one step to move about forward；φ (r) indicates the direction vector of chemotactic；
2) operation is replicated：It replicates operation simulation bacterium to look for food middle survival of the fittest behavior, bacterium completes predetermined number of times chemotactic behaviour
After work, fitness is come subsequent S/2 bacterium and is eliminated, remaining S/2 excellent bacterium self-replacations, generate with it is respective complete
Exactly the same new individual keeps population scale constant；
3) migration operation：The organisms for meeting migration occurrence condition in population are deleted, and are arbitrarily generated in solution space
One new individual；
The optimization for extending black-start scheme has multiple constraintss, will be met by unit pre-selection and starts time and subsidiary engine
The unit of capacity-constrained is alternately activated unit, and black starting-up power supply self-excitation is to restoration path with System Reactive Power constraint
Idle constraint of charging is generated, if Qb=min (KCBSB,QB.max), two constraintss merge into idle constraintFormula
In, KCBFor black starting-up unit short-circuit ratio, SBFor black starting-up power supply capacity, QB.maxIt is idle for the absorbent maximum of black starting-up power supply,
QLjIdle, the N generated for circuit jlFor the circuit number of recovery；
In conjunction with the bi-level programming Optimized model established, bacterial foraging algorithm is improved：
1) bacterium position vector dimension takes the alternative of the condition of satisfaction to be activated unit number, and position vector, which becomes, measures range of variables
Interior arbitrary real number takes 1 or 0, the 1 corresponding unit starting of expression, and 0 indicates that corresponding unit does not start, and fitness takes unit to weight
Generating efficiency negative value and subsidiary engine start influence value two parts and, by the solution space of the position vector space reflection of bacterium to scheme；
2) chemotactic operation is improved：Take two positions, the variable among two positions at random on bacterium position vector
It is remained unchanged in chemotactic operation, and the variable of both sides changes at random.
Use critical path method (CPM) solve optimal restoration path the step of for：
Step 1：According to do not restore unit node serial number sequence call successively Dijkstra's algorithm be each machine group searching most
Short restoration path；
Step 2：The unit of routine weight value minimum is set to search sign, restoration path center line right of way value is set to minimum
Value ε；
Step 3：Judge whether not set also search sign is activated unit, and step 1 is turned to if having, and otherwise output is each
The optimal restoration path of unit；
Step 4：The target function value of lower layer's restoration path optimization is calculated.
The advantageous effect that the present invention is reached：
The factor that the application will influence black starting-up stage recovery safety brings optimization aim concentration into, establishes and extends black open
Dynamic scheme bi-level programming Optimized model, respectively using unit starting and restoration path as the variable of levels Optimal Decision-making, the side of making
Case has higher safety while pursuing unit starting rear weight generating efficiency maximum in recovery process.It proposes
The hybrid solving algorithm that bacterial foraging algorithm is combined with the critical path method (CPM) based on Dijkstra's algorithm is improved, the application's
The extension black-start scheme that method obtains restores compared with the extension black-start scheme for being only up to target with unit generation amount
Process has higher safety, improves recovery effects, has the practical engineering application value of bigger.
Description of the drawings
Fig. 1 is the calculation flow chart of hybrid optimization algorithm.
Specific implementation mode
The invention will be further described below in conjunction with the accompanying drawings.Following embodiment is only used for clearly illustrating the present invention
Technical solution, and not intended to limit the protection scope of the present invention.
1 considers to restore the extension black-start scheme bilevel programming model of safety factor
1.1 bilevel programming model basic principles
Bilevel programming model was proposed by Bracken and McGill in 1973 earliest, for solving that there are two layers to be passed rank knot
The system optimization decision problem of structure, upper layer decision and lower layer's decision respectively have object function and constraints, upper layer decision problem
Object function and constraints are not only related with the decision variable on upper layer, but also depend on the optimal solution or optimal value of lower layer,
The optimal solution of lower layer's decision problem is influenced by upper layer decision variable, and optimal solution is fed back to upper layer and influences upper layer decision most
Excellent solution is an optimization algorithm with wide application prospect, is ground in Electric Power Network Planning, power distribution network electricity optimization, electricity market etc.
Study carefully field to be applied, the universal model of bi-level programming is expressed as：
MinF=F (x, y) (1)
s.t.G(x)≤0 (2)
Minf=f (x, y) (3)
s.t.g(x,y)≤0 (4)
In formula：F () is the object function of upper layer decision problem；X is upper layer decision vector；G is that upper layer decision constrains item
Part；F is the object function of lower layer's decision problem；Y is lower layer's decision vector；G is lower layer's decision constraints.
1.2 extension black-start scheme bilevel programming models
In the black starting-up stage, the recovery of unit is basis, and unit provides power for follow-up recovery after successfully starting up and supports.Machine
On the one hand the optimization of group restoration path reduces the operational risk restored in unit starting path, improve the reliability of recovery, another party
Face will put into operation as early as possible is conducive to the critical circuits of follow-up back bone network structure, and is conducive to the recovery as early as possible of important load.It is being
In recovery process of uniting, the difference of unit to be launched results in the difference of restoration path, therefore unit restores and path recovery has
Principal and subordinate's logical relation.The application optimizes the recovery scheme for extending black starting-up using bilevel programming model.
1.2.1 upper layer Optimized model
When the subsidiary engine for being activated unit starts, since initial stages of restoration mini system is weaker, system voltage can be caused substantially
Degree declines and frequency is fallen and makes mini system by greater impact.Therefore, when optimizing unit starting, it is necessary to consider that subsidiary engine starts
Influence.In extension black starting-up restores, the stagger the time mode of startup of subsidiary engine that different units can be used reduces impact to system,
If the unit number being activated is NG, it is P to be activated unit i subsidiary engine capacityGL.i, PLIt can be extensive under system operation constraints to meet
Multiple maximal workload calculates P using piecewise-linear techniquesL, the maximum subsidiary engine of capacity in unit is activated with difference and is started
It indicates influence of the set auxiliary machinery startup to system in extension black-start scheme, the influence that subsidiary engine starts is defined as：
IFValue is bigger, when indicating that subsidiary engine starts, cause voltage decline and frequency to fall value bigger, initial stages of restoration system by
Impact it is bigger.
The influence that set auxiliary machinery starts in black starting-up is extended to reduce, it is smaller subsidiary engine capacity can preferably to be chosen by unit
Unit.It is in network topology critical positions and the unit for being capable of providing larger generated energy after starting starts as early as possible, after being more advantageous to
It is continuous to restore, therefore, upper layer Optimum Synthesis consider unit location prominence, unit generation amount and efficiency, subsidiary engine start influence and under
The feedback of layer restoration path optimization optimizes unit starting, and object function is：
In formula：F is upper layer target, including set optimization and lower layer's object function penalty term two parts；T1To optimize the time；
t0For the recovery time in extension black starting-up path；aiThe pitch point importance where unit i, with the network cohesion degree after node contraction
It indicates；W (t) is the weighting coefficient of unit output in different periods, and value reduces as time goes by；PGi(t) be unit i in t
The active power that moment sends out；γ is transformation coefficient；η is penalty coefficient；F is lower layer's target function value.
1.2.2 lower layer's Optimized model
When sky fills restoration path, it is related to circuit grid switching operation and no-load transformer input, line no-load switching will produce behaviour
Make overvoltage and power-frequency overvoltage, sky, which fills transformer, can cause resonance overvoltage and generate excitation surge current, include in excitation surge current
The prodigious higher hamonic wave of numerical value, a large amount of reactive powers that line mutual-ground capacitor generates easily cause black starting-up power supply self-excitation, on
The failure that the black starting-up stage restores may be caused by stating problem all.
Line loop operation overvoltage is related with the factors such as system structure and breaker performance, can be reduced using low pressure charging modes
The influence of line loop operation overvoltage and harmonic wave overvoltage, power-frequency overvoltage and black starting-up power supply self-excitation and reactive power in system
Related, and line mutual-ground capacitor positive correlation in restoration path, it is black that the excessive reactive power generated on restoration path limits extension
The scale for starting mini system therefore can be by the relatively small circuit of restoration path optimum option direct-to-ground capacitance to reduce charging nothing
Work(power can equally be chosen the few path of transformation number to reduce the empty influence for filling transformer by path optimization.It is crucial in system
The recovery of circuit is conducive to the foundation of back bone network, to subsequently restoring to play an important role；The substation passed through in restoration path
Including load it is more important, be more conducive to reduce the loss that brings of having a power failure.Therefore, the optimization of lower layer's restoration path is with comprising important
Circuit and important load and charge power it is smaller, be target by the few circuit of transformer, object function is：
In formula：NlFor the circuitry number of recovery；BjFor the susceptance of branch j；Binary decision variable λjIndicate whether branch j is to become
Depressor branch is to take 1, no to take 0；JjFor the betweenness after branch j normalization；θ(Lj) be the both ends branch j plant stand load importance,
It is set according to load actual conditions by dispatching of power netwoks department, β1、β2、β3、β4For conversion coefficient.
Model constraints is：Upper layer constraint includes startup power constraint, unit starting time-constrain, set auxiliary machinery capacity
Constraint, unit output constraint；Lower layer's constraint includes black starting-up power supply self-excitation magnetic confinement, System Reactive Power constraint, line power and section
The operations constraint such as point voltage, embodies as follows：
In formula：Pcr.iTo be activated the startup power of unit i；K2For coefficient of reliability；P0It is opened for what black starting-up power supply provided
Dynamic power；TS.iFor the startup time of unit i；TCH.iFor the maximum critical thermal starting time of unit i；PGi、QGiFor having for unit i
Work(, idle output；For the active and reactive output bound of unit；QLjThe nothing generated for circuit j
Work(；KCBFor black starting-up unit short-circuit ratio；SBFor black starting-up power supply capacity；QB.maxIt is idle for the absorbent maximum of black starting-up power supply；
PjIt is active on circuit j；It is maximum active on circuit j；UiFor node i voltage,It is upper and lower for node i voltage
Limit.
In above-mentioned bilevel programming model, upper layer decision variable is to be activated unit, is produced to path optimization in lower layer's decision
It is raw to influence, and lower layer path optimization in reflection to upper layer target, plays upper layer decision in the form of the optimal penalty function of object function
To feedback effect, model reflects that upper layer unit starting optimizes and what lower layer's restoration path optimized influences each other, excellent by upper layer
Change and the reciprocation of lower layer's optimization finally obtains comprehensive optimal extension black-start scheme.
2 hybrid solving algorithms and flow
Above-mentioned model belongs to complicated nonlinear multivariate regression analysis problem, and single method is difficult to solve, for institute's established model
Feature, the application are calculated using bacterial foraging algorithm is improved with the mixing that the critical path method (CPM) based on Dijkstra's algorithm is combined
Method solves.Bacterial foraging algorithm is improved, using in the strong advantage Optimization Solution upper layer decision of its ability of searching optimum
Unit starting solves the routing problem of lower layer using the Minimal Spanning Tree based on Dijkstra's algorithm.
2.1, which improve bacterial foraging algorithm, solves unit starting optimization
Bacterial foraging algorithm (Bacterial foraging algorithm, BFA) be by K.M.Passino professor in
The novel colony intelligence bionic Algorithm of the simulation Escherichia coli physiology sexual behaviour in foraging behavior proposed in 2002, mainly by thin
Three kinds of operation of bacterium chemotactic, duplication operation and migration operation operation iterative calculation optimization problem solvings, there is distributed parallel to search for, is complete
The advantages that office's optimizing ability is strong, is one of the hot spot of current intelligent algorithm research.
1) chemotactic operates.The behavior that bacterium is moved about by the swing of flagellum to eutrophy area, including move about and overturn two kinds
Behavior, bacterium are defined as overturning to random direction Moving Unit step-length, and travelling is to maintain unidirectional movement, if bacterial population
Scale is S, and search space is tieed up for D, and the position of bacterium k can use D dimension space vectorsIt indicates, each bacterium
Position represents a solution of problem.θk(l, m, n) indicates that bacterium k operates the m times duplication operation in the l times chemotactic and n-th is moved
The position after operation is moved, the chemotactic operation of each steps of bacterium k is expressed as：
θk(l+1, m, n)=θk(l,m,n)+c(d)φ(r) (9)
In formula：C (d) indicates the one step to move about forward；φ (r) indicates the direction vector of chemotactic.
2) operation is replicated.It replicates operation simulation bacterium to look for food middle survival of the fittest behavior, bacterium completes predetermined number of times chemotactic behaviour
After work, fitness is come subsequent S/2 bacterium and is eliminated, remaining S/2 excellent bacterium self-replacations, generate with it is respective complete
Exactly the same new individual keeps population scale constant.
3) migration operation.Organisms escape behavior caused by the mutation of migration operation simulated environment, is sent out with certain probability
Raw, the organisms that migration occurrence condition is met in population are deleted, and a new individual is arbitrarily generated in solution space, are migrated
Operation contributes to organisms to jump out local optimum search globally optimal solution.
The optimization for extending black-start scheme has multiple constraintss, can preselect meet by unit and starts time and auxiliary
The unit of machine capacity-constrained is alternately activated unit, and black starting-up power supply self-excitation is to restoring road with System Reactive Power constraint
Diameter generates idle constraint of charging, and takes Qb=min (KCBSB,QB.max), two constraintss can merge into idle constraintThe operations such as unit output constraint, line power and node voltage constraint can be to the scheme Load flow calculation of generation and imitative
It really verifies, then unit starting optimization can relax to constrain the knapsack problem constituted by startup power, and path optimization is to meet
To unit restoration path optimizing under idle constraint.
When bacterial foraging algorithm is applied to optimization extension black-start scheme, in conjunction with the specific of established bilevel programming model
Actual features have carried out following improvement to bacterial foraging algorithm：
1) bacterium position vector dimension takes the alternative of the condition of satisfaction to be activated unit number, and position vector variable generally takes variable
Arbitrary real number in range has startup or does not start two states due to being activated unit, and the application, which becomes, to measure 1 or 0,1 and indicate phase
The unit starting answered, 0 indicates that corresponding unit does not start, and fitness, which takes unit weighting generating efficiency negative value and subsidiary engine to start, to be influenced
Be worth two parts and, by the solution space of the position vector space reflection of bacterium to scheme.
2) chemotactic operation determines the direction of advance of bacterium, is the core operation of bacterial foraging algorithm, determines on optimization upper layer
When plan unit starting, following improve is carried out to chemotactic operation：Two positions are taken at random on bacterium position vector, among two positions
Variable remained unchanged in chemotactic operation, and the variable of both sides changes at random.
2.2 solve optimal restoration path using critical path method (CPM)
Unit restoration path optimizes the minimum support tree problem for being considered as figure, and the application is used to be calculated based on Di Jiesitela
The critical path method (CPM) of method seeks the Minimal Spanning Tree that connection is activated unit and black starting-up power supply, and circuit weights are taken asTo obtain the restoration path of total weight value minimum, it is as follows：
Step 1：According to do not restore unit node serial number sequence call successively Dijkstra's algorithm be each machine group searching most
Short restoration path.
Step 2：The unit of routine weight value minimum is set to search sign, restoration path center line right of way value is set to minimum
Value ε.
Step 3：Judge whether not set also search sign is activated unit, is to turn to step 1, otherwise exports each machine
The optimal restoration path of group.
Step 4：Lower layer's optimization object function value is calculated.
2.3 model solution flows
Above-mentioned improvement bacterial foraging algorithm and the critical path method (CPM) based on Dijkstra's algorithm are combined applied to expansion
The solution of black-start scheme bilevel programming model is opened up, specific calculation process is as shown in Figure 1.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, several improvement and deformations can also be made, these improvement and deformations
Also it should be regarded as protection scope of the present invention.
Claims (9)
1. improving the extension black-start scheme bi-level programming optimization method of recovery process safety, characterized in that it is black to establish extension
Startup scheme bi-level programming Optimized model, respectively using unit starting and restoration path as the change of the upper and lower Optimal Decision-making
Amount, establishes bi-level programming Optimized model constraints, most using improvement bacterial foraging algorithm and based on Dijkstra's algorithm
The hybrid algorithm that short-circuit shot is combined solves bi-level programming Optimized model.
2. the extension black-start scheme bi-level programming optimization method according to claim 1 for improving recovery process safety,
It is characterized in that upper layer Optimized model establishment step is：
If the unit number being activated is NG, it is P to be activated unit i subsidiary engine capacityGL.i, PLTo meet under system operation constraints
Recoverable maximal workload considers unit location prominence, unit generation amount and efficiency, subsidiary engine starts influence and lower layer
The feedback of restoration path optimization optimizes unit starting, and object function is：
In formula：F is upper layer target, including set optimization and lower layer's object function penalty term two parts；T1To optimize the time；t0For
Extend the recovery time in black starting-up path；aiThe pitch point importance where unit i, with the network cohesion degree table after node contraction
Show；W (t) is the weighting coefficient of unit output in different periods；PGi(t) active power sent out in t moment for unit i；γ is
Transformation coefficient；η is penalty coefficient；F is lower layer's target function value.
3. the extension black-start scheme bi-level programming optimization method according to claim 2 for improving recovery process safety,
It is characterized in that upper layer Optimized model constraints include startup power constraint, unit starting time-constrain, set auxiliary machinery capacity about
Beam and unit output constraint.
4. the extension black-start scheme bi-level programming optimization method according to claim 1 for improving recovery process safety,
It is characterized in that the establishment step of lower layer's Optimized model is：
The object function of lower layer restoration path optimization is：
In formula：NlFor the circuitry number of recovery；BjFor the susceptance of branch j；Binary decision variable λjIndicate whether branch j is transformer
Branch is to take 1, no to take 0；JjFor the betweenness after branch j normalization；θ(Lj) be the both ends branch j plant stand load importance, by electricity
It nets traffic department to be set according to load actual conditions, β1、β2、β3、β4For conversion coefficient.
5. the extension black-start scheme bi-level programming optimization method according to claim 4 for improving recovery process safety,
It is characterized in that lower layer's Optimized model constraints include black starting-up power supply self-excitation magnetic confinement, System Reactive Power constraint, line power and
Node voltage constrains.
6. the extension black-start scheme bi-level programming optimization method according to claim 1 for improving recovery process safety,
It is characterized in that establishing bi-level programming Optimized model constraints expression formula and being：
In formula：Pcr.iTo be activated the startup power of unit i；K2For coefficient of reliability；P0The startup work(provided for black starting-up power supply
Rate；TS.iFor the startup time of unit i；TCH.iFor the maximum critical thermal starting time of unit i；PGi、QGiFor active, the nothing of unit i
Work(is contributed；For unit active power output upper and lower limit；For the idle output upper and lower limit of unit；QLjFor line
Road j is generated idle；KCBFor black starting-up unit short-circuit ratio；SBFor black starting-up power supply capacity；QB.maxIt can be absorbed for black starting-up power supply
Maximum it is idle；PjIt is active on circuit j；It is maximum active on circuit j；UiFor node i voltage,For node
I voltage upper and lower limits；NlFor the circuit number of recovery.
7. the extension black-start scheme bi-level programming optimization method according to claim 1 for improving recovery process safety,
It is characterized in that the hybrid algorithm being combined with the critical path method (CPM) based on Dijkstra's algorithm using bacterial foraging algorithm is improved
When solution, using the unit starting improved in bacterial foraging algorithm solution upper layer Optimal Decision-making, calculated using based on Di Jiesitela
The Minimal Spanning Tree of method solves the optimal restoration path in lower layer's Optimal Decision-making.
8. the extension black-start scheme bi-level programming optimization method according to claim 7 for improving recovery process safety,
It is characterized in that the step of solving unit starting is：
Bacterial foraging algorithm step is：
1) chemotactic operates：If bacterial population scale is S, search space is tieed up for D, and the position of bacterium k can use D dimension space vectorsIt indicates, the position of each bacterium represents a solution of problem；θk(l, m, n) indicates that bacterium k becomes at the l times
Change the m times position replicated after operation and n-th migration operation of operation, the chemotactic operation of each steps of bacterium k is expressed as：
θk(l+1, m, n)=θk(l,m,n)+c(d)φ(r) (9)
In formula：C (d) indicates the one step to move about forward；φ (r) indicates the direction vector of chemotactic；
2) operation is replicated：Operation simulation bacterium is replicated to look for food middle survival of the fittest behavior, after bacterium completes the operation of predetermined number of times chemotactic,
Fitness is come subsequent S/2 bacterium to eliminate, remaining S/2 excellent bacterium self-replacations, be generated and respective complete phase
Same new individual keeps population scale constant；
3) migration operation：The organisms for meeting migration occurrence condition in population are deleted, and one is arbitrarily generated in solution space
New individual；
The optimization for extending black-start scheme has multiple constraintss, will be met by unit pre-selection and starts time and subsidiary engine capacity
The unit of constraint is alternately activated unit, and black starting-up power supply self-excitation with System Reactive Power constraint is generated to restoration path
It charges idle constraint, takes Qb=min (KCBSB,QB.max), two constraintss merge into idle constraintIn formula,
KCBFor black starting-up unit short-circuit ratio, SBFor black starting-up power supply capacity, QB.maxIdle, the Q for the absorbent maximum of black starting-up power supplyLj
Idle, the N generated for circuit jlFor the circuit number of recovery；
In conjunction with the bi-level programming Optimized model established, bacterial foraging algorithm is improved：
1) bacterium position vector dimension takes the alternative of the condition of satisfaction to be activated unit number, and position vector, which becomes to measure in range of variables, appoints
Meaning real number takes 1 or 0, the 1 corresponding unit starting of expression, and 0 indicates that corresponding unit does not start, and fitness takes unit weighting power generation
Efficiency negative value and subsidiary engine start influence value two parts and, by the solution space of the position vector space reflection of bacterium to scheme；
2) chemotactic operation is improved：Two positions, the variable among two positions are taken to become at random on bacterium position vector
Change and is remained unchanged in operation, and the variable of both sides changes at random.
9. the extension black-start scheme bi-level programming optimization method according to claim 7 for improving recovery process safety,
It is characterized in that use critical path method (CPM) solve optimal restoration path the step of for：
Step 1：Call Dijkstra's algorithm most short extensive for each machine group searching successively according to unit node serial number sequence is not restored
Multiple path；
Step 2：The unit of routine weight value minimum is set to search sign, restoration path center line right of way value is set to minimum ε；
Step 3：Judge whether not set also search sign is activated unit, and step 1 is turned to if having, otherwise exports each unit
Optimal restoration path；
Step 4：The target function value of lower layer's restoration path optimization is calculated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810269185.4A CN108539781B (en) | 2018-03-29 | 2018-03-29 | Extended black start scheme two-layer planning optimization method for improving safety of recovery process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810269185.4A CN108539781B (en) | 2018-03-29 | 2018-03-29 | Extended black start scheme two-layer planning optimization method for improving safety of recovery process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108539781A true CN108539781A (en) | 2018-09-14 |
| CN108539781B CN108539781B (en) | 2020-03-10 |
Family
ID=63482316
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810269185.4A Active CN108539781B (en) | 2018-03-29 | 2018-03-29 | Extended black start scheme two-layer planning optimization method for improving safety of recovery process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108539781B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109861297A (en) * | 2019-04-11 | 2019-06-07 | 上海电机学院 | A kind of power system blackstart method based on grey wolf optimization algorithm |
| CN111064218A (en) * | 2018-10-17 | 2020-04-24 | 云南电网有限责任公司怒江供电局 | Power grid black start multi-target unit recovery optimization method based on multi-target optimization algorithm |
| CN111293683A (en) * | 2020-02-13 | 2020-06-16 | 东方电子股份有限公司 | Distribution network self-healing optimization method considering safety and economy |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102904251A (en) * | 2012-10-23 | 2013-01-30 | 广东电网公司电力调度控制中心 | Method for system for configuring thermal power generating units to carry out self-healing of power grid |
| CN102904256A (en) * | 2012-10-23 | 2013-01-30 | 广东电网公司电力调度控制中心 | Method and system for rapid self-healing of power grid |
| US20130130231A1 (en) * | 2002-11-26 | 2013-05-23 | Isaac Bentwich | Bioinformatically detectable group of novel viral regulatory genes and uses thereof |
| CN103746406A (en) * | 2013-12-24 | 2014-04-23 | 广东电网公司电力调度控制中心 | Configuration method for fast cut back (FCB) thermo power set |
| CN103762589A (en) * | 2014-01-08 | 2014-04-30 | 河海大学 | Method for optimizing new energy capacity ratio in layers in power grid |
| CN104463375A (en) * | 2014-12-24 | 2015-03-25 | 贵州电网公司电力调度控制中心 | Power grid disaster recovery control model modeling method based on CIM standard |
| CN104638672A (en) * | 2015-01-26 | 2015-05-20 | 东南大学 | Determining method of photovoltaic transmission power limit considering variable correlation |
| CN105470945A (en) * | 2015-11-09 | 2016-04-06 | 南京理工大学 | Breadth-first direct-current converter station recovery path generation method |
| CN105529701A (en) * | 2015-12-30 | 2016-04-27 | 南京理工大学 | Direct current convertor station containing power supply recovery route optimization method based on artificial bee colony algorithm |
| CN106159944A (en) * | 2016-08-04 | 2016-11-23 | 上海电力学院 | Multi-stage transmission expansion planning method based on bilevel programming model under low-carbon (LC) environment |
| CN106209459A (en) * | 2016-08-11 | 2016-12-07 | 国网江苏省电力公司电力科学研究院 | A kind of network connectivty modification method of the system recovery path intelligent optimization that has a power failure |
| CN106815657A (en) * | 2017-01-05 | 2017-06-09 | 国网福建省电力有限公司 | A kind of power distribution network bi-level programming method for considering timing and reliability |
-
2018
- 2018-03-29 CN CN201810269185.4A patent/CN108539781B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130130231A1 (en) * | 2002-11-26 | 2013-05-23 | Isaac Bentwich | Bioinformatically detectable group of novel viral regulatory genes and uses thereof |
| CN102904251A (en) * | 2012-10-23 | 2013-01-30 | 广东电网公司电力调度控制中心 | Method for system for configuring thermal power generating units to carry out self-healing of power grid |
| CN102904256A (en) * | 2012-10-23 | 2013-01-30 | 广东电网公司电力调度控制中心 | Method and system for rapid self-healing of power grid |
| CN103746406A (en) * | 2013-12-24 | 2014-04-23 | 广东电网公司电力调度控制中心 | Configuration method for fast cut back (FCB) thermo power set |
| CN103762589A (en) * | 2014-01-08 | 2014-04-30 | 河海大学 | Method for optimizing new energy capacity ratio in layers in power grid |
| CN104463375A (en) * | 2014-12-24 | 2015-03-25 | 贵州电网公司电力调度控制中心 | Power grid disaster recovery control model modeling method based on CIM standard |
| CN104638672A (en) * | 2015-01-26 | 2015-05-20 | 东南大学 | Determining method of photovoltaic transmission power limit considering variable correlation |
| CN105470945A (en) * | 2015-11-09 | 2016-04-06 | 南京理工大学 | Breadth-first direct-current converter station recovery path generation method |
| CN105529701A (en) * | 2015-12-30 | 2016-04-27 | 南京理工大学 | Direct current convertor station containing power supply recovery route optimization method based on artificial bee colony algorithm |
| CN106159944A (en) * | 2016-08-04 | 2016-11-23 | 上海电力学院 | Multi-stage transmission expansion planning method based on bilevel programming model under low-carbon (LC) environment |
| CN106209459A (en) * | 2016-08-11 | 2016-12-07 | 国网江苏省电力公司电力科学研究院 | A kind of network connectivty modification method of the system recovery path intelligent optimization that has a power failure |
| CN106815657A (en) * | 2017-01-05 | 2017-06-09 | 国网福建省电力有限公司 | A kind of power distribution network bi-level programming method for considering timing and reliability |
Non-Patent Citations (5)
| Title |
|---|
| ZEYANG TANG等: "《Comparison of the Weibull and the Crow-AMSAA》", 《IEEE TRANSACTIONS ON POWER DELIVERY》 * |
| 刘刚，李永树: "《一种面向城市复杂路网最短路径》", 《计算机应用研究》 * |
| 李珺，党建武: "《改进细菌觅食算法在高维优化问题中的应用》", 《计算机学报》 * |
| 王大江等: "《The Bi-level Programming Optimization Method of》", 《2016 CHINA INTERNATIONAL CONFERENCE ON ELECTRICITY DISTRIBUTION (CICED 2016) XI’AN, 10-13 AUG, 2016》 * |
| 顾雪平等: "《电力系统扩展黑启动方案的后效性研究》", 《电工技术学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111064218A (en) * | 2018-10-17 | 2020-04-24 | 云南电网有限责任公司怒江供电局 | Power grid black start multi-target unit recovery optimization method based on multi-target optimization algorithm |
| CN109861297A (en) * | 2019-04-11 | 2019-06-07 | 上海电机学院 | A kind of power system blackstart method based on grey wolf optimization algorithm |
| CN109861297B (en) * | 2019-04-11 | 2022-11-01 | 上海电机学院 | Black start method of power system based on grey wolf optimization algorithm |
| CN111293683A (en) * | 2020-02-13 | 2020-06-16 | 东方电子股份有限公司 | Distribution network self-healing optimization method considering safety and economy |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108539781B (en) | 2020-03-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sharma et al. | Impact of ultracapacitor and redox flow battery with JAYA optimization for frequency stabilization in linked photovoltaic‐thermal system | |
| Khezri et al. | An intelligent coordinator design for GCSC and AGC in a two-area hybrid power system | |
| Ashari et al. | Power loss reduction strategy of distribution network with distributed generator integration | |
| Sogabe et al. | Smart grid optimization by deep reinforcement learning over discrete and continuous action space | |
| CN108539781A (en) | Improve the extension black-start scheme bi-level programming optimization method of recovery process safety | |
| Dixit et al. | Economic load dispatch using artificial bee colony optimization | |
| CN109274136A (en) | A kind of photovoltaic system idle work optimization method based on quanta particle swarm optimization | |
| CN110363363A (en) | A kind of dual-stage scheduling model optimization method for integrated energy system | |
| Ngamroo | Application of electrolyzer to alleviate power fluctuation in a stand alone microgrid based on an optimal fuzzy PID control | |
| Parmar | Load frequency control of multi-source power system with redox flow batteries: an analysis | |
| Dekaraja et al. | Impact of energy storage and flexible alternating current transmission devices in combined voltage and frequency regulation of multiarea multisource interconnected power system | |
| Liang | Application of grey linear programming to short-term hydro scheduling | |
| CN114881346A (en) | Double-layer optimization scheduling method for dynamic reconstruction and reactive voltage adjustment based on improved whale optimization algorithm | |
| Kalyan et al. | Donkey and smuggler optimization algorithm-based degree of freedom controller for stability of two area power system with AC-DC links | |
| Verma et al. | A state of art review on the opportunities in automatic generation control of hybrid power system | |
| CN109672223A (en) | A kind of probabilistic electric system back bone network recovery scheme of consideration output of wind electric field | |
| CN117833273A (en) | Power system look-ahead scheduling model solving method considering frequency constraint based on neural network assistance | |
| CN116865271A (en) | Digital twin-drive-based micro-grid multi-agent coordination optimization control strategy | |
| CN114330113B (en) | New energy power system elasticity optimization method based on deep reinforcement learning | |
| CN116316812A (en) | Multi-time-scale risk scheduling method for high-proportion distributed energy access transmission and distribution network | |
| CN115765035A (en) | Flexible power distribution network disturbance recovery method suitable for full-time dynamic reconstruction | |
| Shaheen et al. | A chaos game optimization algorithm-based optimal control strategy for performance enhancement of offshore wind farms | |
| CN114943448A (en) | Method and system for constructing micro-grid optimized scheduling model | |
| CN115021241A (en) | Self-recovery method of power distribution network based on distributed resources | |
| Kumar et al. | Improvement in LFC Performance of Dual Area Thermal Hydro System with Territory Control of TCPS and Redox Flow Battery Units |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |