CN108539781B - Extended black start scheme two-layer planning optimization method for improving safety of recovery process - Google Patents
Extended black start scheme two-layer planning optimization method for improving safety of recovery process Download PDFInfo
- Publication number
- CN108539781B CN108539781B CN201810269185.4A CN201810269185A CN108539781B CN 108539781 B CN108539781 B CN 108539781B CN 201810269185 A CN201810269185 A CN 201810269185A CN 108539781 B CN108539781 B CN 108539781B
- Authority
- CN
- China
- Prior art keywords
- unit
- layer
- recovery
- black start
- optimization
- 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.)
- Active
Links
- 238000011084 recovery Methods 0.000 title claims abstract description 98
- 238000005457 optimization Methods 0.000 title claims abstract description 82
- 241001672018 Cercomela melanura Species 0.000 title claims abstract description 70
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 54
- 230000001580 bacterial effect Effects 0.000 claims abstract description 39
- 230000002431 foraging effect Effects 0.000 claims abstract description 26
- 241000894006 Bacteria Species 0.000 claims description 31
- 230000035605 chemotaxis Effects 0.000 claims description 20
- 230000005012 migration Effects 0.000 claims description 11
- 238000013508 migration Methods 0.000 claims description 11
- 230000010076 replication Effects 0.000 claims description 6
- 230000009471 action Effects 0.000 claims description 3
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 3
- 231100000817 safety factor Toxicity 0.000 abstract 1
- 230000008901 benefit Effects 0.000 description 4
- 241000196324 Embryophyta Species 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000005284 excitation Effects 0.000 description 3
- 230000033001 locomotion Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000003399 chemotactic effect Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 241000588724 Escherichia coli Species 0.000 description 1
- 240000005893 Pteridium aquilinum Species 0.000 description 1
- 235000009936 Pteridium aquilinum Nutrition 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 210000003495 flagella Anatomy 0.000 description 1
- 230000019637 foraging behavior Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 239000011664 nicotinic acid Substances 0.000 description 1
- 230000008560 physiological behavior Effects 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
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]
Abstract
The invention discloses an extended black start scheme two-layer planning optimization method for improving the safety of a recovery process, which provides a new thought for recovery control at an initial stage after a major power failure, brings safety factors influencing the recovery process into an optimization target for improving the safety of the extended black start scheme in the actual recovery process, and establishes an extended black start scheme two-layer planning optimization model by considering the master-slave logic relationship of unit recovery and path recovery; the upper-layer optimization aims at achieving the highest comprehensive efficiency of the scheme under the high-safety condition, the decision variable is a started unit, the lower-layer optimization aims at solving a recovery path which contains a key line and is high in recovery safety and beneficial to important load recovery, and the decision variable is a line to be recovered; and solving the model by adopting a hybrid algorithm combining an improved bacterial foraging algorithm and a shortest path method based on a Dijkstra algorithm. The recovery process has higher safety, improves the recovery effect and has practical engineering application value.
Description
Technical Field
The invention relates to an extended black start scheme two-layer planning optimization method for improving the safety of a recovery process, and belongs to the technical field of power systems.
Background
In recent years, smart grid construction plays an important role in improving grid operation reliability, but the possibility of major power failure accidents of a grid under the influence of various internal faults and external interference still exists. Research on recovery control of the power system can accelerate the recovery process of the system after a major power failure and reduce huge economic loss and social influence caused by the power failure, and is one of important subjects of power system safety defense.
According to tasks and characteristics of different periods in the system recovery process, the system recovery process after major power failure can be generally divided into three stages of black start, grid reconstruction and load recovery. The black start phase is an initial phase of recovery, power support can be provided for subsequent recovery after the started unit is successfully started, the formed local small network is a basis of the subsequent recovery, but the successful recovery of the started unit is required to be taken as a premise, and if the started unit fails in recovery, a system recovery process is delayed, so that greater loss is caused. In the whole recovery process of the black start stage, operation risks such as overvoltage of an empty charge line, self-excitation of a black start power supply, excitation surge current and overvoltage generated by an empty charge transformer, system voltage reduction and frequency drop caused by starting of auxiliary machines of the unit are restricted, and if any link does not satisfy the restriction, recovery failure is caused, so that the started unit cannot be started successfully. Therefore, when a recovery scheme is prepared, it is necessary to ensure that the recovery process has higher safety so as to be beneficial to the successful recovery of the unit, but at present, the requirement of safety is not taken into consideration while the optimal recovery efficiency is ensured.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, provides an extended black start scheme two-layer planning optimization method for improving the safety of a recovery process, brings relevant factors influencing the safety of the recovery process into an extended black start scheme optimization target, establishes an extended black start scheme two-layer planning model considering the safety of the recovery process, and ensures that the scheme decision gives consideration to the requirement of safety while pursuing the maximum weighted power generation efficiency after a unit is started. The hybrid algorithm combining the improved bacterial foraging algorithm and the shortest path method based on the Dijkstra algorithm is adopted for solving, the improved bacterial foraging algorithm is utilized for optimizing the upper-layer nonlinear programming, the minimum support tree based on the Dijkstra algorithm is adopted for optimizing the lower-layer recovery path, and the advantages of the two algorithms are fully exerted.
In order to solve the technical problems, the invention adopts the following technical scheme.
The two-layer planning optimization method for the extended black start scheme is characterized by establishing an extended black start scheme two-layer planning optimization model, establishing two-layer planning optimization model constraint conditions by respectively taking unit start and recovery paths as variables of upper-layer and lower-layer optimization decisions, and solving the two-layer planning optimization model by adopting a hybrid solution algorithm combining an improved bacterial foraging algorithm and a shortest path method based on a Dijkstra algorithm.
The upper layer optimization model establishment steps are as follows:
setting the number of the started sets to be NGThe auxiliary machine capacity of the started unit i is PGL.i，PLIn order to meet the maximum load amount which can be recovered under the constraint condition of system operation, the unit starting is optimized by comprehensively considering the unit position importance, the unit generating capacity and efficiency, the auxiliary engine starting influence and the feedback of the lower layer recovery path optimization, and the objective function is as follows:
in the formula: f is an upper layer target which comprises two parts of unit optimization and a lower layer target function penalty item; t is1To optimize the time; t is t0To extend the recovery time of the black start path; a isiRepresenting the importance of the node where the unit i is located by the network aggregation degree after the node is contracted; w (t) is a weighting coefficient of unit output in different periods; pGiAnd (t) the active power of the unit i at the moment t, gamma is a transformation coefficient, η is a penalty coefficient, and f is a lower-layer objective function value.
The constraint conditions of the upper-layer optimization model comprise starting power constraint, unit starting time constraint, unit auxiliary machine capacity constraint and unit output constraint.
The lower layer optimization model is established by the following steps:
the objective function of the lower layer restoration path optimization is as follows:
in the formula: n is a radical oflThe number of branches recovered; b isjIs the susceptance of branch j; binary decision variable lambdajWhether the branch j is a transformer branch or not is represented, 1 is selected, and 0 is not selected; j. the design is a squarejThe number of medians after normalization for the branch j; theta (L)j) The importance of the plant loads at the two ends of the branch j is set by a power grid dispatching department according to the actual load conditions, β1、β2、β3、β4Are conversion coefficients.
The lower-layer optimization model constraint conditions comprise black-start power supply self-excitation constraint, system reactive power constraint, line power and node voltage constraint.
Establishing a constraint condition expression of a two-layer planning optimization model as follows:
in the formula: pcr.iStarting power of a started unit i; k2To be reliableA coefficient of sex; p0Starting power provided for the black start power supply; t isS.iThe starting time of the unit i is set; t isCH.iThe maximum critical hot start time of the unit i is obtained; pGi、QGiThe active and reactive power output of the unit i is obtained;
the upper limit and the lower limit of the active output of the unit are set;
the upper limit and the lower limit of the reactive power output of the unit are set; qLjReactive power generated for line j; kCBThe short circuit ratio of the unit is black start; sBBlack start power supply capacity; qB.maxThe maximum reactive power which can be absorbed by the black start power supply; pjActive on line j;
is the maximum active on line j; u shapeiIs the voltage at the node i and is,
the voltage upper limit and the voltage lower limit of a node i are set; n is a radical oflIs the number of lines recovered.
When a hybrid algorithm combining an improved bacterial foraging algorithm and a shortest path method based on a Dijkstra algorithm is adopted for solving, the improved bacterial foraging algorithm is adopted for solving the unit start in the upper-layer optimization decision, and the minimum support tree based on the Dijkstra algorithm is adopted for solving the optimal recovery path in the lower-layer optimization decision.
The method for solving the starting of the unit comprises the following steps:
the bacterial foraging algorithm comprises the following steps:
1) chemotaxis operation: setting the size of the bacterial population as S, the search space as D dimension and the position of the bacteria k as D dimension space vector
Indicating that the location of each bacterium represents a solution to the problem; thetak(l, m, n) indicates that bacterium k replicates at the m-th operation in the l-th chemotaxis operationAnd position after the nth migration operation, the chemotactic operation of bacterium k at each step is expressed as:
θk(l+1,m,n)＝θk(l,m,n)+c(d)φ(r) (9)
in the formula: c (d) represents a unit step size of forward walking; phi (r) denotes the direction vector of chemotaxis;
2) copying operation: the replication operation simulates the action of excellence and elimination in bacterial foraging, S/2 bacteria with fitness arranged behind are eliminated after the bacteria complete specified times of chemotaxis operation, the remaining S/2 excellent bacteria are replicated by themselves to generate new individuals completely the same as the respective ones, and the population scale is kept unchanged;
3) and (3) migration operation: deleting the bacterial individuals in the population which meet the migration occurrence condition, and randomly generating a new individual in the solution space;
the optimization of the extended black start scheme has a plurality of constraint conditions, the unit meeting the constraint of start time and auxiliary machine capacity is used as a standby started unit through unit preselection, the self-excitation of the black start power supply and the reactive power constraint of the system are both constraints for generating charging reactive power on a recovery path, and Q is setb＝min(KCBSB,QB.max) The two constraints are combined into a reactive constraint
In the formula, KCBShort-circuit ratio of black start unit, SBFor black start power supply capacity, QB.maxMaximum reactive power, Q, absorbed by the black start power supplyLjReactive power generated for line j, NlThe number of lines to be recovered;
and (3) improving a bacterial foraging algorithm by combining the established two-layer programming optimization model:
1) the method comprises the steps that the number of alternative started sets meeting conditions is selected in the position vector dimension of bacteria, any real number in a variable range is selected as a position vector variable, 1 or 0 is selected, 1 represents that a corresponding set is started, 0 represents that the corresponding set is not started, the fitness is the sum of a set weighted generating efficiency negative value and an auxiliary machine starting influence value, and the position vector space of the bacteria is mapped to a solution space of a scheme;
2) the chemotaxis operation is improved: two positions are randomly selected on the bacterial position vector, the variable in the middle of the two positions is kept unchanged in chemotaxis operation, and the variables on the two sides are randomly changed.
The steps of solving the optimal recovery path by adopting the shortest path method are as follows:
step 1: sequentially calling Dijkstra algorithm according to the serial number sequence of the unrecovered unit nodes to search the shortest recovery path for each unit;
step 2: setting the unit with the minimum path weight as a searched mark, and setting the line weight in a recovery path as a minimum value epsilon;
and step 3: judging whether a started unit without a search mark exists, if so, turning to the step 1, otherwise, outputting an optimal recovery path of each unit;
and 4, step 4: and calculating to obtain an objective function value of the lower layer restoration path optimization.
The invention achieves the following beneficial effects:
the method and the device have the advantages that factors influencing the recovery safety of the black start stage are brought into the optimization target set, the two-layer planning optimization model of the extended black start scheme is established, the unit start path and the recovery path are respectively used as variables of upper-layer and lower-layer optimization decisions, the weighted power generation efficiency of the scheme after the unit start is pursued is maximized, and meanwhile the scheme has high safety in the recovery process. The hybrid solving algorithm combining the improved bacterial foraging algorithm and the shortest path method based on the Dijkstra algorithm is provided, compared with the extended black start scheme which only uses the maximum generating capacity of the unit as the target, the extended black start scheme obtained by the method has higher safety in the recovery process, improves the recovery effect and has higher practical engineering application value.
Drawings
FIG. 1 is a computational flow diagram of a hybrid optimization algorithm.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
1 expansion black start scheme two-layer planning model considering recovery safety factor
1.1 basic principles of two-layer planning model
The two-layer planning model is firstly proposed by Bracken and McGill in 1973, and is used for solving the system optimization decision problem with a two-layer hierarchical structure, an upper-layer decision and a lower-layer decision respectively have an objective function and a constraint condition, the objective function and the constraint condition of the upper-layer decision problem are not only related to a decision variable of an upper layer, but also depend on an optimal solution or an optimal value of a lower layer, the optimal solution of the lower-layer decision problem is influenced by the decision variable of the upper layer, the optimal solution is fed back to the upper layer to influence the optimal solution of the upper-layer decision, the two-layer planning model is an optimization algorithm with wide application prospect, and the two-layer planning model is applied in the research fields of power grid planning, power supply optimization of a power distribution network, power market and the like, and:
minF＝F(x,y) (1)
s.t.G(x)≤0 (2)
minf＝f(x,y) (3)
s.t.g(x,y)≤0 (4)
in the formula: f (-) is an objective function of an upper-layer decision problem; x is an upper layer decision vector; g is an upper-layer decision constraint condition; f is an objective function of a lower-layer decision problem; y is a lower layer decision vector; g is the lower layer decision constraint.
1.2 two-layer programming model of extended Black Start scheme
In the black start stage, the recovery of the unit is the basis, and the unit provides power support for subsequent recovery after being successfully started. The optimization of the unit recovery path reduces the operation risk of unit starting path recovery and improves the reliability of recovery, and on the other hand, the optimization of the unit recovery path is beneficial to the early recovery of important loads and the early operation of key lines which are beneficial to the construction of subsequent skeleton networks. In the system recovery process, the unit recovery and the path recovery have a master-slave logic relationship because the recovery paths are different due to the difference of the units to be started. The recovery scheme of the extended black start is optimized by adopting a two-layer planning model.
1.2.1 Upper layer optimization model
When the auxiliary machine of the started unit is started, the small system is weak at the initial recovery stage, so that the voltage of the system is greatly reduced, the frequency is dropped, and the small system is subjected to large impact. Therefore, when optimizing the start-up of the plant, the effect of the start-up of the auxiliary machines must be taken into account. In the recovery of the extended black start, the impact on the system can be reduced by adopting a mode that auxiliary machines of different machine sets are started in a time-staggered manner, and the number of the started machine sets is NGThe auxiliary machine capacity of the started unit i is PGL.i，PLIn order to meet the maximum recoverable load under the constraint condition of system operation, P is calculated by adopting a piecewise linearization methodLThe method comprises the following steps of expressing the influence of the starting of the auxiliary machine of the unit on a system in an extended black start scheme by starting the auxiliary machine with the maximum capacity in different started units, and defining the influence of the starting of the auxiliary machine as follows:
IFthe larger the value, the larger the voltage drop and frequency drop value caused when the auxiliary machine is started, the larger the shock applied to the system at the initial recovery stage.
In order to reduce the influence of the starting of the auxiliary machine of the unit in the expansion black start, the unit with smaller auxiliary machine capacity can be preferably selected by the unit. The unit which is located at the important position of the network topology and can provide larger generating capacity after being started is started as early as possible, and subsequent recovery is facilitated, so that the upper-layer optimization comprehensively considers the importance of the unit position, the generating capacity and the efficiency of the unit, the starting influence of an auxiliary machine and the feedback of the lower-layer recovery path optimization to optimize the starting of the unit, and the objective function is as follows:
in the formula: f is an upper layer target which comprises two parts of unit optimization and a lower layer target function penalty item; t is1To optimize the time; t is t0To extend the recovery time of the black start path; a isiRepresenting the importance of the node where the unit i is located by the network aggregation degree after the node is contracted; w (t) is a weighting system of unit output in different periodsThe value of the number is reduced along with the time; pGiAnd (t) the active power of the unit i at the moment t, gamma is a transformation coefficient, η is a penalty coefficient, and f is a lower-layer objective function value.
1.2.2 lower layer optimization model
During the empty charging recovery path, the operation of circuit switching and the no-load input of a transformer are involved, the no-load switching of the circuit can generate operation overvoltage and power frequency overvoltage, the empty charging transformer can cause resonance overvoltage and generate excitation surge current, the excitation surge current contains higher harmonics with large values, a large amount of reactive power generated by the circuit to ground capacitance easily causes the self-excitation of a black start power supply, and the problem can cause the failure of the black start stage recovery.
The circuit operation overvoltage is related to factors such as a system structure, the performance of a circuit breaker and the like, the influence of the circuit operation overvoltage and harmonic overvoltage can be reduced by adopting a low-voltage charging mode, the power frequency overvoltage and the self-excitation of the black start power supply are related to reactive power in a system and are positively related to the ground capacitance of a circuit in a recovery path, and the scale of the expansion black start small system is limited by excessive reactive power generated on the recovery path. The recovery of the key line in the system is beneficial to the establishment of a skeleton network, and plays an important role in subsequent recovery; the more important the load contained in the substation passing through the restoration path is, the more advantageous the reduction of the loss due to the power outage is. Therefore, the lower layer restoration path optimization aims at the line which comprises important lines and important loads, has small charging power and passes few transformers, and the objective function is as follows:
in the formula: n is a radical oflThe number of branches recovered; b isjIs the susceptance of branch j; binary decision variable lambdajWhether the branch j is a transformer branch or not is represented, 1 is selected, and 0 is not selected; j. the design is a squarejThe number of medians after normalization for the branch j; theta (L)j) The importance of the plant loads at the two ends of the branch j is set by a power grid dispatching department according to the actual load conditions, β1、β2、β3、β4Are conversion coefficients.
The model constraint conditions are as follows: the upper layer constraint comprises a starting power constraint, a unit starting time constraint, a unit auxiliary machine capacity constraint and a unit output constraint; the lower layer constraint comprises self-excitation constraint of a black-start power supply, system reactive constraint, line power, node voltage and other operation constraints, and is specifically expressed as follows:
in the formula: pcr.iStarting power of a started unit i; k2Is a reliability coefficient; p0Starting power provided for the black start power supply; t isS.iThe starting time of the unit i is set; t isCH.iThe maximum critical hot start time of the unit i is obtained; pGi、QGiThe active and reactive power output of the unit i is obtained;
the active and reactive output upper and lower limits of the unit are set; qLjReactive power generated for line j; kCBThe short circuit ratio of the unit is black start; sBBlack start power supply capacity; qB.maxThe maximum reactive power which can be absorbed by the black start power supply; pjActive on line j;
is the maximum active on line j; u shapeiIs the voltage at the node i and is,
the upper and lower limits of the voltage of the node i.
In the two-layer planning model, the upper-layer decision variable is a started unit and influences the path optimization in the lower-layer decision, the lower-layer path optimization is reflected in an upper-layer target in a penalty function mode with an optimal target function and plays a feedback role on the upper-layer decision, the model reflects the interaction of the upper-layer unit start optimization and the lower-layer recovery path optimization, and the comprehensive optimal extended black start scheme is finally obtained through the interaction of the upper-layer optimization and the lower-layer optimization.
2 hybrid solving algorithm and flow
The model belongs to a complex nonlinear two-layer programming problem, a single method is difficult to solve, and aiming at the characteristics of the built model, the hybrid algorithm combining the improved bacterial foraging algorithm and the shortest path method based on the Dijkstra algorithm is adopted for solving. The bacterial foraging algorithm is improved, the advantage of strong global search capability is utilized to optimize and solve unit starting in upper-layer decision-making, and a minimum support tree based on the Dijkstra algorithm is adopted to solve the path optimization problem of the lower layer.
2.1 improved bacterial foraging Algorithm solving unit startup optimization
The Bacterial Foraging Algorithm (BFA) is a novel swarm intelligent bionic algorithm which is provided by k.m. passino in 2002 and simulates the physiological behavior of escherichia coli in foraging behavior, and mainly solves the problem by iterative calculation and optimization through three operations, namely Bacterial chemotaxis operation, replication operation and migration operation, so that the Bacterial foraging algorithm has the advantages of distributed parallel search, strong global optimization capability and the like, and is one of hot spots of current intelligent algorithm research.
1) And (5) chemotactic operation. The method comprises the following steps that the bacteria move to a eutrophic area through the swinging of flagella, the movement of the bacteria to a random direction is defined as overturning in unit step length, the movement is the movement of keeping the same direction, the bacterial population scale is set as S, the search space is D-dimension, and the position of the bacteria k can use a D-dimension space vector
It is shown that the location of each bacterium represents a solution to the problem. Thetak(l, m, n) represents the position of bacterium k after the mth replication operation and the nth migration operation in the l chemotaxis operation, and the chemotaxis operation of bacterium k at each step is represented as:
θk(l+1,m,n)＝θk(l,m,n)+c(d)φ(r) (9)
in the formula: c (d) represents a unit step size of forward walking; φ (r) represents the direction vector of chemotaxis.
2) And (4) copying operation. The replication operation simulates the action of excellence and elimination in bacterial foraging, S/2 bacteria with fitness arranged behind are eliminated after the bacteria complete specified times of chemotaxis operation, the remaining S/2 excellent bacteria are replicated by themselves to generate new individuals which are completely the same as the respective excellent bacteria, and the population scale is kept unchanged.
3) And (5) migrating the operation. The migration operation simulates the bacterial individual death escaping behavior caused by environmental mutation, the bacterial individual is deleted according to a certain probability, the bacterial individual meeting the migration occurrence condition in the population is deleted, a new individual is randomly generated in the solution space, and the migration operation is favorable for the bacterial individual to jump out of the local optimal search global optimal solution.
The optimization of the extended black start scheme has a plurality of constraint conditions, the unit meeting the constraint of the start time and the auxiliary machine capacity can be used as an alternative started unit through unit preselection, the self-excitation of the black start power supply and the reactive power constraint of the system are both constraints for generating charging reactive power on a recovery path, and Q is selectedb＝min(KCBSB,QB.max) The two constraints can be combined into a reactive constraint
The unit output constraint, the line power, the node voltage and other operation constraints can be verified through calculation and simulation of generated scheme load flow, unit starting optimization can be relaxed to a knapsack problem formed by starting power constraints, and path optimization is to optimize a unit recovery path under the condition of meeting reactive power constraints.
When the bacterial foraging algorithm is applied to the optimized extended black start scheme, the following improvements are made on the bacterial foraging algorithm by combining the specific practical characteristics of the established two-layer programming model:
1) the method includes the steps that the number of alternative started units meeting conditions is selected according to the vector dimension of bacteria positions, any real number in a variable range is generally selected according to the position vector variable, the started units are in a starting state or a non-starting state, the variable quantity is 1 or 0, 1 represents that the corresponding units are started, 0 represents that the corresponding units are not started, the fitness is the sum of a negative value of the weighted generating efficiency of the units and a starting influence value of an auxiliary machine, and the position vector space of the bacteria is mapped to a solution space of a scheme.
2) The chemotaxis operation determines the advancing direction of bacteria, is the core operation of a bacterial foraging algorithm, and is improved as follows when an upper-layer decision unit is optimized to start: two positions are randomly selected on the bacterial position vector, the variable in the middle of the two positions is kept unchanged in chemotaxis operation, and the variables on the two sides are randomly changed.
2.2 solving the optimal restoration Path by the shortest Path method
The method adopts a shortest path method based on Dijkstra algorithm to obtain the minimum support tree connecting the started unit and the black start power supply, and the line weight is taken as
To obtain the recovery path with the minimum total weight, the specific steps are as follows:
step 1: and sequentially calling Dijkstra algorithm according to the serial number sequence of the nodes of the unrecovered unit to search the shortest recovery path for each unit.
Step 2: and setting the unit with the minimum path weight as a searched mark, and setting the line weight in the recovery path as a minimum value epsilon.
And step 3: and (4) judging whether a started unit without a search mark exists, if so, turning to the step 1, otherwise, outputting the optimal recovery path of each unit.
And 4, step 4: and calculating to obtain a lower-layer optimization objective function value.
2.3 model solution flow
The improved bacterial foraging algorithm and the shortest path method based on the dijkstra algorithm are combined and applied to the solution of the two-layer programming model of the extended black start scheme, and the specific calculation flow is shown in fig. 1.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.
Claims (8)
1. The extended black start scheme two-layer planning optimization method for improving the safety of the recovery process is characterized by establishing an extended black start scheme two-layer planning optimization model, establishing two-layer planning optimization model constraint conditions by respectively taking unit start and recovery paths as variables of upper-layer and lower-layer optimization decisions, and solving the two-layer planning optimization model by adopting a hybrid algorithm combining an improved bacterial foraging algorithm and a shortest path method based on a Dijkstra algorithm;
the lower layer optimization model is established by the following steps:
the objective function of the lower layer restoration path optimization is as follows:
in the formula: n is a radical oflThe number of branches recovered; b isjIs the susceptance of branch j; binary decision variable lambdajWhether the branch j is a transformer branch or not is represented, 1 is selected, and 0 is not selected; j. the design is a squarejThe number of medians after normalization for the branch j; theta (L)j) The importance of the plant loads at the two ends of the branch j is set by a power grid dispatching department according to the actual load conditions, β1、β2、β3、β4Are conversion coefficients.
2. The extended black start scheme two-layer programming optimization method for improving the security of the recovery process according to claim 1, wherein the upper layer optimization model establishing step is as follows:
setting the number of the started sets to be NGThe auxiliary machine capacity of the started unit i is PGL.i，PLIn order to meet the maximum load amount which can be recovered under the constraint condition of system operation, the unit starting is optimized by comprehensively considering the unit position importance, the unit generating capacity and efficiency, the auxiliary engine starting influence and the feedback of the lower layer recovery path optimization, and the objective function is as follows:
in the formula: f is an upper layer target which comprises two parts of unit optimization and a lower layer target function penalty item; t is1To optimize the time; t is t0To extend the recovery time of the black start path αiRepresenting the importance of the node where the unit i is located by the network aggregation degree after the node is contracted; w (t) is a weighting coefficient of unit output in different periods; pGiAnd (t) the active power of the unit i at the moment t, gamma is a transformation coefficient, η is a penalty coefficient, and f is a lower-layer objective function value.
3. The extended black-start scheme two-layer programming optimization method for improving recovery process security of claim 2, wherein the upper layer optimization model constraints include start power constraints, unit start time constraints, unit auxiliary capacity constraints, and unit output constraints.
4. The extended black start scheme two-level programming optimization method for improving recovery process security of claim 1, wherein the lower level optimization model constraints include black start power supply self-excitation constraints, system reactive constraints, line power and node voltage constraints.
5. The extended black start scheme two-layer programming optimization method for improving the security of the recovery process according to claim 1, wherein the two-layer programming optimization model is established by a constraint condition expression:
in the formula: pcr.iStarting power of a started unit i; k2Is a reliability coefficient; p0Starting power provided for the black start power supply; t isS.iThe starting time of the unit i is set; t isCH.iThe maximum critical hot start time of the unit i is obtained; pGi、QGiIs a unit iActive and reactive power output;
the upper limit and the lower limit of the active output of the unit are set;
the upper limit and the lower limit of the reactive power output of the unit are set; qLjReactive power generated for line j; kCBThe short circuit ratio of the unit is black start; sBBlack start power supply capacity; qB.maxThe maximum reactive power which can be absorbed by the black start power supply; pjActive on line j;
is the maximum active on line j; u shapeiIs the voltage at the node i and is,
the voltage upper limit and the voltage lower limit of a node i are set; n is a radical oflIs the number of lines recovered.
6. The extended black start scheme two-tier planning optimization method for improving the security of the recovery process according to claim 1, wherein when solving by using a hybrid algorithm combining an improved bacterial foraging algorithm and a shortest path method based on a dijkstra algorithm, the unit start in the upper-tier optimization decision is solved by using the improved bacterial foraging algorithm, and the optimal recovery path in the lower-tier optimization decision is solved by using a minimum support tree based on the dijkstra algorithm.
7. The extended black start scheme two-layer programming optimization method for improving recovery process security of claim 6, wherein the step of solving the unit start is:
the bacterial foraging algorithm comprises the following steps:
1) chemotaxis operation: setting the size of the bacterial population as S, the search space as D dimension and the position of the bacteria k as D dimension space vector
Indicating that the location of each bacterium represents a solution to the problem; thetak(l, m, n) represents the position of bacterium k after the mth replication operation and the nth migration operation in the l chemotaxis operation, and the chemotaxis operation of bacterium k at each step is represented as:
θk(l+1,m,n)＝θk(l,m,n)+c(d)φ(r) (9)
in the formula: c (d) represents a unit step size of forward walking; phi (r) denotes the direction vector of chemotaxis;
2) copying operation: the replication operation simulates the action of excellence and elimination in bacterial foraging, S/2 bacteria with fitness arranged behind are eliminated after the bacteria complete specified times of chemotaxis operation, the remaining S/2 excellent bacteria are replicated by themselves to generate new individuals completely the same as the respective ones, and the population scale is kept unchanged;
3) and (3) migration operation: deleting the bacterial individuals in the population which meet the migration occurrence condition, and randomly generating a new individual in the solution space;
the optimization of the extended black start scheme has a plurality of constraint conditions, the unit meeting the constraint of start time and auxiliary machine capacity is used as an alternative started unit through unit preselection, the self-excitation of the black start power supply and the reactive power constraint of the system are both constraints for generating charging reactive power on a recovery path, and the self-excitation of the black start power supply and the reactive power constraint of the system are taken
The two constraint conditions are combined into reactive constraint
In the formula, KCBShort-circuit ratio of black start unit, SBFor black start power supply capacity, QB.maxMaximum reactive power, Q, absorbed by the black start power supplyLjReactive power generated for line j, NlThe number of lines to be recovered;
and (3) improving a bacterial foraging algorithm by combining the established two-layer programming optimization model:
1) the method comprises the steps that the number of alternative started sets meeting conditions is selected in the position vector dimension of bacteria, any real number in a variable range is selected as a position vector variable, 1 or 0 is selected, 1 represents that a corresponding set is started, 0 represents that the corresponding set is not started, the fitness is the sum of a set weighted generating efficiency negative value and an auxiliary machine starting influence value, and the position vector space of the bacteria is mapped to a solution space of a scheme;
2) the chemotaxis operation is improved: two positions are randomly selected on the bacterial position vector, the variable in the middle of the two positions is kept unchanged in chemotaxis operation, and the variables on the two sides are randomly changed.
8. The extended black start scheme two-layer planning optimization method for improving the safety of the recovery process according to claim 7, wherein the step of solving the optimal recovery path by using a shortest path method comprises:
step 1: sequentially calling Dijkstra algorithm according to the serial number sequence of the unrecovered unit nodes to search the shortest recovery path for each unit;
step 2: setting the unit with the minimum path weight as a searched mark, and setting the line weight in a recovery path as a minimum value epsilon;
and step 3: judging whether a started unit without a search mark exists, if so, turning to the step 1, otherwise, outputting an optimal recovery path of each unit;
and 4, step 4: and calculating to obtain an objective function value of the lower layer restoration path optimization.
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 CN108539781A (en) | 2018-09-14 |
| CN108539781B true 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) |
Families Citing this family (3)
| 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 |
| 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 |
| CN111293683B (en) * | 2020-02-13 | 2021-05-18 | 东方电子股份有限公司 | Distribution network self-healing optimization method considering safety and economy |
Family Cites Families (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 |
| CN102904256B (en) * | 2012-10-23 | 2014-09-10 | 广东电网公司电力调度控制中心 | Method and system for rapid self-healing of power grid |
| CN102904251B (en) * | 2012-10-23 | 2014-09-10 | 广东电网公司电力调度控制中心 | Method and system for configuring thermal power generating units to carry out self-healing of power grid |
| CN103746406B (en) * | 2013-12-24 | 2016-08-17 | 广东电网公司电力调度控制中心 | The collocation method of fast cut back fired power generating unit |
| CN103762589B (en) * | 2014-01-08 | 2015-10-21 | 河海大学 | A kind of new forms of energy capacity ratio hierarchy optimization method in electrical network |
| 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 |
| CN104638672B (en) * | 2015-01-26 | 2017-02-01 | 东南大学 | Determining method of photovoltaic transmission power limit considering variable correlation |
| CN105470945B (en) * | 2015-11-09 | 2017-11-14 | 南京理工大学 | A kind of DC converter station restoration path generation method based on breadth First |
| CN105529701B (en) * | 2015-12-30 | 2018-01-05 | 南京理工大学 | A kind of method for optimizing route of power up containing DC converter station based on artificial bee colony algorithm |
| CN106159944B (en) * | 2016-08-04 | 2019-08-16 | 上海电力学院 | Multi-stage transmission expansion planning method under low-carbon environment based on bilevel programming model |
| CN106209459B (en) * | 2016-08-11 | 2019-06-11 | 国网江苏省电力公司电力科学研究院 | A kind of network connectivty modification method of power failure system restoration path intelligent optimization |
| CN106815657B (en) * | 2017-01-05 | 2020-08-14 | 国网福建省电力有限公司 | Power distribution network double-layer planning method considering time sequence and reliability |
-
2018
- 2018-03-29 CN CN201810269185.4A patent/CN108539781B/en active Active
Non-Patent Citations (2)
| Title |
|---|
| 《Comparison of the Weibull and the Crow-AMSAA》;Zeyang Tang等;《IEEE TRANSACTIONS ON POWER DELIVERY》;20151231;第30卷(第6期);第2410-2418页 * |
| 《一种面向城市复杂路网最短路径》;刘刚，李永树;《计算机应用研究》;20110630;第28卷(第6期);第2082-2084页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108539781A (en) | 2018-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mishra et al. | A comprehensive review on power distribution network reconfiguration | |
| Bouchekara et al. | Optimal power flow using the league championship algorithm: A case study of the Algerian power system | |
| Ashari et al. | Power loss reduction strategy of distribution network with distributed generator integration | |
| CN108539781B (en) | Extended black start scheme two-layer planning optimization method for improving safety of recovery process | |
| CN104820865B (en) | Intelligent distribution network fault recovery intelligent optimization method based on graph theory | |
| CN105932690B (en) | A kind of distribution running optimizatin method of comprehensive idle work optimization and network reconfiguration | |
| CN104037765B (en) | The method of active power distribution network service restoration scheme is chosen based on improved adaptive GA-IAGA | |
| Sayah et al. | Optimal power flow solution of integrated AC‐DC power system using enhanced differential evolution algorithm | |
| Duman et al. | Economical operation of modern power grids incorporating uncertainties of renewable energy sources and load demand using the adaptive fitness-distance balance-based stochastic fractal search algorithm | |
| CN112271727A (en) | Fault recovery method for flexible power distribution network containing flexible soft switch | |
| Mahdad et al. | Security optimal power flow considering loading margin stability using hybrid FFA–PS assisted with brainstorming rules | |
| Anuranj et al. | Resiliency based power restoration in distribution systems using microgrids | |
| Ela et al. | Optimal corrective actions for power systems using multi-objective genetic algorithms | |
| Biswas et al. | Optimal power flow solutions using algorithm success history based adaptive differential evolution with linear population reduction | |
| Khetrapal | Distribution Network Reconfiguration of Radial Distribution Systems for Power Loss Minimization Using Improved Harmony Search Algorithm. | |
| Manikanta et al. | Distribution network reconfiguration with different load models using adaptive quantum inspired evolutionary algorithm | |
| Lotfi et al. | An optimal co-operation of distributed generators and capacitor banks in dynamic distribution feeder reconfiguration | |
| CN106487001B (en) | A kind of isolated power system intelligent reconstruction method | |
| Zhang et al. | Security-constrained optimal power flow solved with a dynamic multichain particle swarm Optimizer | |
| CN113507116B (en) | Power distribution network load transfer method, device, equipment and storage medium | |
| CN111313416B (en) | Multi-source collaborative intelligent power distribution network fault recovery sequence optimization decision method | |
| Fan et al. | An integrated power restoration method based on improved genetic algorithm for active distribution network | |
| CN110365006A (en) | A kind of sub-area division method based on nwbbo algorithm | |
| LI et al. | Islanding partition of distribution system with distributed generations based on community structure | |
| Zhang et al. | Coordinated restoration model of renewable energy and conventional generator units after faults |
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 |