CN108923427B - Method for reconstructing ship power distribution network based on queue intelligent algorithm - Google Patents
Method for reconstructing ship power distribution network based on queue intelligent algorithm Download PDFInfo
- Publication number
- CN108923427B CN108923427B CN201810916768.1A CN201810916768A CN108923427B CN 108923427 B CN108923427 B CN 108923427B CN 201810916768 A CN201810916768 A CN 201810916768A CN 108923427 B CN108923427 B CN 108923427B
- Authority
- CN
- China
- Prior art keywords
- power
- load
- candidate
- ship
- candidates
- 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
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
-
- 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
A reconstruction method of a ship power distribution network based on a queue intelligent algorithm belongs to the technical field of ship power distribution network reconstruction. The invention aims to solve the problems of large calculation amount and slow power supply recovery in the process of recovering power supply after a ship distribution network fault, firstly, relevant parameters of a ship distribution network are obtained, state information of a generator set, a distribution board, a contact bus, a load, a breaker and the like is determined, when a power system has a fault, the fault of the distribution network is judged, a power loss area is isolated, whether the load power requirement of the power loss area can be met by controlling the output power of a generator and the breaker is judged, the ship power network is reconstructed based on a queue intelligent algorithm according to the conditions, whether non-critical loads need to be unloaded is judged, the load power maximization is realized, and the power is supplied again to the power loss area so as to enable the power loss area to.
Description
Technical Field
The invention belongs to the technical field of ship distribution network reconstruction, and particularly relates to a ship distribution network reconstruction method based on a queue intelligent algorithm.
Background
The problem of power distribution network reconstruction is a typical problem in a power system, a power distribution network comprises section switches and interconnection switches and generally has the characteristics of closed-loop design and open-loop operation, and the power distribution network reconstruction is to change a network structure by changing the states of the switches so as to achieve the purpose of optimizing operation or safe operation. Distribution network reconfiguration typically involves two cases: the first condition is that the network structure is adjusted according to the operation condition when the power distribution system operates normally, the power supply voltage quality of the power distribution system is improved, and the power flow direction in the power distribution system is changed, so that the purpose of reducing the power transmission loss of the system network is achieved, and the condition is summarized as the network optimization problem in power distribution network reconstruction; the second condition is that when the power distribution system is in fault, the fault area is isolated by changing the switch state according to the fault information, and the power supply is rapidly recovered to the non-fault area as far as possible. Generally, the power system on land has many nodes, long transmission distance, large network loss and limited influence of load change on the system, so the reconstruction of the distribution network of the power system on land usually refers to the first situation, i.e. the network optimization problem.
Different from a land power system, the ship power system has short transmission distance, unobvious voltage drop and small network loss, and has the basic task of ensuring continuous and reliable power supply service of a ship.
The requirement of rapid power restoration after the failure of the ship comprehensive electric propulsion system is also the key problem to be solved in the construction of the energy management of the full electric ship, and is an important basis for improving the task execution capacity and the survival capacity of the electric ship. The traditional conventional ship adopts alternating current power distribution, a nonlinear power flow calculation equation meeting system operation constraint needs to be solved in a reconstruction process, and the switch state in a power distribution network is solved by adopting artificial intelligent algorithms such as a genetic algorithm and a particle swarm algorithm group to achieve the aim of load recovery power supply maximization after system failure, so that the high coupling of the switch state and the load power supply maximization is realized, but the defects of large calculation amount and slow power supply recovery exist.
Disclosure of Invention
The invention aims to solve the problems of large calculated amount and slow power supply recovery in the process of recovering power supply after a ship power distribution network fault, and provides a ship power distribution network reconstruction method based on a queue intelligent algorithm.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a ship power distribution network reconstruction method based on a queue intelligent algorithm comprises the following specific steps:
the method comprises the following steps: carrying out graph theory representation on the ship power distribution network aiming at different ship types and power distribution networks;
step two: analyzing faults of the ship power distribution network: after partial area loses power, firstly isolating a fault area, secondly judging whether the existing power supply power of the ship can compensate the power shortage of the area which does not lose power, namely whether the load power requirement of the area which does not lose power can be met by controlling the output power of a generator and a breaker, and if the current line generating capacity is met, not needing to reconstruct a power grid; if the current line power generation capacity is insufficient, power grid reconstruction is needed, namely, load power supply is maximized;
step three: an objective function is established, the reconstruction of the ship power grid is abstracted into a 0-1 knapsack problem,
when the system power generation capacity is insufficient, the load that plays an important role in ship vitality and task execution should be recovered first; the load power supply maximization problem is mathematically described as:
wherein f (x) represents the overall value of the load; x is the number ofiWhether power supply is restored for the load i is shown, 1 represents power supply, and 0 represents no power supply; r isiIndicates the priority of the load i, i.e. the importance of the load, satisfies
PLiRepresents the power demand of load i; pGRepresenting the power generation capacity of the system; and the constraint condition that the sum of the consumed power of all the loads cannot exceed the maximum power generation capacity of the system is realized as an objective function;
decomposing each propulsion load into n equal discrete loads, wherein the corresponding priority coefficient is the propulsion load priority coefficient, and the decision variables are expanded to 2n + 3;
step four: reconstructing a ship power distribution network based on a queue intelligent algorithm, and solving a 0-1 knapsack problem by using the queue intelligent algorithm; firstly, defining the total number of candidate cases C of a queue and the execution times t of each generation of optimization algorithm, and initially randomly selecting each group of candidate cases C (C is 1A set of object loads, each set of candidates representing a power supply scheme, all candidates forming a queue of problems, for each set of candidates, a relative value F is calculatedC＝{f(v1),...,f(vc),...,f(vC) And weight FCW＝{f(w1),...,f(wc),...,f(wC)}，
(1) To measure the similarity of each candidate output to the optimal solution, a similarity probability function p is definedcThe following were used:
wherein the content of the first and second substances,
the probability of a value is represented by,
representing the weight probability, and W representing the maximum value of the system capacity;
based on the wheel selection method, each set of candidates C (C1.., C) will select one target candidate f (v)c[？]) To evolve itself, superscript [?]The target candidate is randomly selected by a roulette method based on a similar probability function, and each candidate c cannot be determined in advance;
(2) each candidate performs the above process t times so that each candidate c can obtain an associated value output set Fc,t＝{f(vc)1,...,f(vc)j,...,f(vc)t}, (C ═ 1.., C); after each iteration is completed, selecting the optimal value f (v) of each group of candidates as the initial input of the next iteration; in this way, the value output F of each set of candidates in the queue is updatedC＝{f*(v1),...,f*(vc),...,f*(vC)}，
The schemes are classified as follows:
A. if a solution of the candidate C (C ═ 1., C) is feasible, i.e., satisfies the weight constraint given by equation (1), then one of the following modifications is randomly selected:
1. adding a randomly selected object from the target candidates, requiring that the object is not included in the current candidate c and still satisfies the weight constraint given by equation (1);
2. replacing a randomly selected object in the original candidates with a randomly selected object in the target candidates, and satisfying the formula (1);
B. if a solution for candidate C (C1., C) is not feasible, then one of the following modifications is randomly selected:
1. randomly removing an object from the candidates;
2. replacing a randomly selected object from the original candidate c with a randomly selected object from the target candidate to reduce the total weight f (w) of the candidate cc) (ii) a Each candidate performs the above process t times; this allows each candidate c to obtain a relevant set of value outputs Fc,t＝{f(vc)1,...,f(vc)j,...,f(vc)t}, (C ═ 1.., C); after each iteration is completed, each candidate selects the optimal value f (v) as the initial input of the next iteration, and since infeasible results may occur in each iteration evolution, the optimal value is selected according to the following conditions:
2.1. selecting the result with the greatest value if the evolved result is feasible;
2.2. selecting the result with the least weight if the post-evolution result is not feasible;
2.3. if there is an infeasible and feasible change, selecting the feasible result with the greatest value;
in this way, the value output F for each candidate in the queue is updatedC＝{f*(v1),...,f*(vc),...,f*(vC)}
(3) Each group of candidates optimizes the solution thereof in a mutual learning mode; this iterative process continues until the algorithm saturates (converges), i.e., the results of each candidate output tend to be consistent and do not change in a large number of successive evolution attempts, i.e., the algorithm is considered to converge;
step five: and completing power grid reconstruction, and re-supplying power to the load in the area without power loss to recover the operation of the load.
Compared with the prior art, the invention has the beneficial effects that: the method improves the process of realizing the reconstruction of the ship distribution network, provides a ship power distribution reconstruction method based on a queue intelligent algorithm, and evaluates the reconstruction result of each time in real time, thereby taking corresponding measures and ensuring the whole distribution reconstruction process to be complete and reliable. On the other hand, by the roulette method, the searching precision of the algorithm is improved, and the speed and the efficiency of reconstructing the ship distribution network are improved.
Drawings
FIG. 1 is an overall flow chart of a method for reconstructing a ship distribution network based on a queue intelligence algorithm according to the present invention;
FIG. 2 is a power distribution block diagram of the medium voltage DC integrated power system of the present invention;
FIG. 3 is a flowchart of a queue intelligence algorithm employed by the method of reconfiguring a ship power distribution network of the present invention;
FIG. 4 is a diagram of an example of a structure of a medium-voltage DC integrated electric propulsion system before a fault occurs;
FIG. 5 is a graphical representation of an example of a marine MVDC integrated electric propulsion system fault;
FIG. 6 is a fault instance reconstruction result diagram for implementing ship distribution network reconstruction based on a queue intelligence algorithm;
fig. 7 is a probability distribution diagram.
Detailed Description
The technical solution of the present invention is further described below with reference to the drawings and the embodiments, but the present invention is not limited thereto, and modifications or equivalent substitutions may be made to the technical solution of the present invention without departing from the spirit of the technical solution of the present invention, and the technical solution of the present invention is covered by the protection scope of the present invention.
As a special type of the integrated electric propulsion system applied to ships, the medium-voltage direct-current integrated electric power system adopts a direct-current system for power supply and distribution, and the power flow constraint is an active power flow linear constraint, so that the requirement of real-time reconstruction after the ship electric power system is in fault can be better met, and therefore, a faster reconstruction method and a more effective reconstruction strategy are researched, and a new thought is developed for realizing ship electric power recovery.
The method of the invention comprises the following steps: the method comprises the steps of obtaining relevant parameters of a ship power distribution network, judging faults of the power distribution network, isolating a power loss area, formulating a scheme for reconstructing the ship power distribution network based on a queue intelligent algorithm, and supplying power again to a load area without power loss so as to enable the load area to continue to work normally. According to the method, on one hand, a ship power distribution network is reconstructed based on a queue intelligent algorithm, the convergence speed generated by a power distribution reconstruction scheme is improved, the situation that the power distribution network is trapped in local optimization is avoided, on the other hand, a selection algorithm in searching is adjusted, a roulette method is utilized, the searching precision of the algorithm is improved, the algorithm is used for reconstructing a ship power grid, a power redistribution result is given, and finally, the effectiveness and the superiority of the proposed method are verified based on example analysis.
A method for reconstructing a ship power distribution network based on a queue intelligent algorithm is disclosed, and the overall implementation flow is as shown in figure 1: the method comprises the steps of firstly obtaining relevant parameters of a ship power distribution network, determining state information of a generator set, a distribution board, a contact bus, a load, a breaker and the like, judging the fault of the power distribution network and isolating a power loss area when a power system has a fault, judging whether the load power requirement of the power loss area can be met by controlling the output power of a generator and the breaker, reconstructing the ship power distribution network based on a queue intelligent algorithm according to the conditions, judging whether non-critical loads need to be unloaded, realizing load power maximization, supplying power again to the power loss area, and enabling the power loss area to continue to work normally.
The first embodiment is as follows: the embodiment describes a method for reconstructing a ship power distribution network based on a queue intelligent algorithm, which comprises the following specific steps:
the method comprises the following steps: carrying out graph theory representation on the ship power distribution network aiming at different ship types and power distribution networks; determining a circuit configuration for a vessel integrated electric propulsion system power distribution network: the MVDC power distribution structure of the medium-voltage direct-current integrated power system is shown in fig. 2 (for convenience of description, converter and inverter links are omitted), and includes 2 main gas turbine generator sets (MTG1, MTG2), 2 auxiliary gas turbine generator sets (ATG1, ATG2), 6 main distribution boards (Bus1, Bus2, Bus3, Bus5, Bus6 and Bus7), 2 interconnection buses (Bus4 and Bus8), 3 regional Load distribution boards (Bus9, Bus10 and Bus11), 2 propulsion loads (Load1 and Load2), 3 regional loads (Load 3-Load 5), and 23 circuit breakers (BK 1-BK 23); the ship service loads are distributed in four areas from the bow to the stern and are supplied with power by port and starboard direct current buses, the loads in each area can be divided into critical loads and non-critical loads from the mission angle of a ship task, the critical loads are provided with two power supply lines, the power can be supplied by the port or starboard area buses, the reliability of the critical loads can be improved, the non-critical loads are provided with only one power supply line, and the non-critical loads can be immediately unloaded without influencing the execution and vitality of the ship task; the 23 circuit breakers are divided into four types, wherein the grey squares in fig. 2 represent the circuit breakers, and the circuit breakers BK 1-BK 4 represent the circuit breakers connected with the medium-voltage direct-current rectifying generator set, and the power flow on the circuit breakers can only flow from the generator to the bus; the breakers BK5 to BK9 indicate low-voltage ac breakers connected to a load, and power flows to the breakers only from the bus to the load; the circuit breakers BK10, BK11, BK14, BK15 represent medium voltage dc bus breakers connected to the power supply grid, on which the power can flow bidirectionally; the other circuit breakers represent circuit breakers of a direct-current regional distribution network, and the power flow direction of the circuit breakers can only flow into a regional distribution bus from a main distribution network bus; dividing the integrated electric propulsion system into 11 protection areas by taking each bus as a center, wherein the areas are overlapped to avoid unprotected areas as shown by a dotted line box in FIG. 1; the solid lines indicate that the circuit breakers on the path are closed and the dashed lines indicate that the circuit breakers on the path are open, for example, Bus4 and Bus8 are tie switches for port and starboard dc buses, and at least one of them is normally open to ensure the power distribution network structure is radial; the regional loads Load 3-Load 5 are provided with two power supply paths, and the dotted lines can be understood as standby paths;
step two: analyzing faults of the ship power distribution network: after partial area loses power, firstly isolating a fault area, secondly judging whether the existing power supply power of the ship can compensate the power shortage of the area which does not lose power, namely whether the load power requirement of the area which does not lose power can be met by controlling the output power of a generator and a breaker, and if the current line generating capacity is met, not needing to reconstruct a power grid; if the current line generating capacity is insufficient, for example, the maximum capacity of the generating unit is 44MW, and the load power demand is 48MW, so that the critical load cannot be started, the grid reconfiguration is required, that is, the load power supply is maximized;
step three: establishing an objective function, abstracting the reconstruction of the ship power grid into a 0-1 knapsack problem, wherein the 0-1 knapsack problem is a classic NP-hard combination problem, and the concrete steps are as follows: in a given set N, each object i, i 1iAnd a weight attribute wiSolving which objects in the set are loaded into the backpack can ensure that the value is maximized on the premise that the weight does not exceed the given capacity; the mathematics are described as follows:
Subject to f(w)≤W
wherein the content of the first and second substances,
(W) represents the overall weight, f (v) represents the overall value, W represents the constraint value that the weight maximum cannot exceed, xiWhether the object is taken or not is indicated as 0 or 1, 0 indicates not to be taken, and 1 indicates to be taken;
when the system power generation capacity is insufficient, the load that plays an important role in ship vitality and task execution should be recovered first; the load power supply maximization problem is mathematically described as:
xi0 or 1
In the formula (I), the compound is shown in the specification,(x) represents the overall value of the load; x is the number ofiWhether power supply is restored for the load i is shown, 1 represents power supply, and 0 represents no power supply; r isiIndicates the priority of the load i, i.e. the importance of the load, satisfies
PLiRepresents the power demand of load i; pGRepresenting the power generation capacity of the system; and the constraint condition that the sum of the consumed power of all the loads cannot exceed the maximum power generation capacity of the system is realized as an objective function;
when considering the problem of load recovery, a conventional ship usually considers the load as a discrete variable, namely, if the load can be provided with the required power, the load is supplied, otherwise, the load is not supplied, in the case that the required power of the load in the conventional ship is small compared with the total installed capacity, the assumption is reasonable, the regional load in the full-power ship can be assumed to be processed, but the size of the propelling load is related to the ship speed, the variation range is large, when the system cannot meet the whole power requirement of the propelling load, the system can select to recover partial power supply for the propelling load instead of completely not supplying power, and therefore, the assumption is not reasonable for the actual operation of the ship;
processing the demanded power of the propulsion load into a combination of several discrete loads, one priority for each discrete load, for example: the power demand of the propulsion Load1 is P1The priority coefficient is w1The propulsion power demand may be expressed as P1 1,P1 2,…,P1 kAnd corresponding priority coefficients
Combinations of discrete loads of (a); decomposing each propulsion load into n equal discrete loads, wherein the corresponding priority coefficient is the propulsion load priority coefficient, the mathematical description of the problem is not changed essentially, and only the number of the loads is expanded to (n + n +3), so that the decision variables are also expanded to 2n + 3; from the above analysis assumptions and equation 1, it can be seen that the load supply maximization problem is essentially a "0-1"Knapsack problem, power generation capacity P in the systemGKnapsack Capacity W, load demand Power P, seen as in the "0-1" knapsack problemLiWeight w of item to be selectedi，riPLiValue v for the item to be selectedi，xiWhether the articles are put into the backpack or not is shown, and x is solved to enable the value of the articles in the backpack to be maximum;
step four: reconstructing the ship power distribution network based on the queue intelligent algorithm is shown in fig. 3, and solving a 0-1 knapsack problem by using the queue intelligent algorithm; applying a CI algorithm to the knapsack problem, the characteristics of each object i, i 1.., N directly determine the overall value f (v) and overall weight f (w) of the knapsack; firstly, defining the total number C of candidate cases of a queue and the execution times t of each generation of optimization algorithm, initially randomly selecting a group of object loads for each group of candidates C (C1.., C), wherein each group of candidates represents a power supply scheme, all the candidates form a problem queue, and calculating the related value F for each group of candidatesC＝{f(v1),...,f(vc),...,f(vC) And weight FCW＝{f(w1),...,f(wc),...,f(wC) The CI algorithm, discussed further below, is shown in fig. 3;
(1) in order to measure the similarity degree of each candidate output and the optimal solution, a probability basis is provided for the roulette method to select an evolution target, and a similarity probability function p is definedcThe following were used:
wherein the content of the first and second substances,
the probability of a value is represented by,
representing the weight probability, and W representing the maximum value of the system capacity;
based on the wheel selection method, each set of candidates C (C1.., C) will select one target candidate f (v)c[？]) To evolve itself, in an algorithm,evolution refers to the goal of optimizing its own solution by combining some objects from the target candidates; the superscript [?]The target candidate is randomly selected by a roulette method based on a similar probability function, and each candidate c cannot be determined in advance;
to bias the solution towards feasibility, FIG. 7 designs a probability distribution; probability of
Linearly increasing with increasing overall backpack weight, peaking at maximum capacity W, with rapid decrease in probability with increasing weight, the slope being twice that before; therefore, the probability is highest around the maximum capacity W, i.e. at this moment, from the capacity point of view, the candidate output is closest to the optimal solution;
(2) each candidate performs the above process t times so that each candidate c can obtain an associated value output set Fc,t＝{f(vc)1,...,f(vc)j,...,f(vc)t}, (C ═ 1.., C); after each iteration is completed, selecting the optimal value f (v) of each group of candidates as the initial input of the next iteration; in this way, the value output F of each set of candidates in the queue is updatedC＝{f*(v1),...,f*(vc),...,f*(vC)}。
Typical schemes are classified as follows:
A. if a solution of the candidate C (C ═ 1., C) is feasible, i.e., satisfies the weight constraint given by equation (1), then one of the following modifications is randomly selected:
1. adding a randomly selected object from the target candidates, requiring that the object is not included in the current candidate c and still satisfies the weight constraint given by equation (1);
2. replacing a randomly selected object in the original candidates with a randomly selected object in the target candidates, and satisfying the formula (1);
B. if a solution for candidate C (C1., C) is not feasible, then one of the following modifications is randomly selected:
1. randomly removing an object from the candidates;
2. replacing a randomly selected object from the original candidate c with a randomly selected object from the target candidate to reduce the total weight f (w) of the candidate cc) (ii) a Each candidate performs the above process t times; this allows each candidate c to obtain a relevant set of value outputs Fc,t＝{f(vc)1,...,f(vc)j,...,f(vc)t}, (C ═ 1.., C); after each iteration is completed, each candidate selects the optimal value f (v) as the initial input of the next iteration, and since infeasible results may occur in each iteration evolution, the optimal value is selected according to the following conditions:
2.1. selecting the result with the greatest value if the evolved result is feasible;
2.2. selecting the result with the least weight if the post-evolution result is not feasible;
2.3. if there is an infeasible and feasible change, selecting the feasible result with the greatest value;
in this way, the value output F for each candidate in the queue is updatedC＝{f*(v1),...,f*(vc),...,f*(vC)}
(3) Each group of candidates optimizes the solution thereof in a mutual learning mode; this iterative process continues until the algorithm saturates (converges), i.e., the results of each candidate output tend to be consistent and do not change in a large number of successive evolution attempts, i.e., the algorithm is considered to converge;
step five: and completing power grid reconstruction, and re-supplying power to the load in the area without power loss to recover the operation of the load.
Example 1:
fig. 4 is an example of the structure of the MVDC integrated electric propulsion system of the ship before failure, the power of each generator and load is shown in the figure, and the failure occurs in each protection area.
Fig. 5 is a graphical representation of an example of a failure of the MVDC integrated electric propulsion system of a ship, a fault of a regional Bus connected with a generator is one of the most prone situations causing insufficient power in the system, and assuming that a fault of a Bus1 connected with MTG1 occurs, circuit breakers BK1, BK10, BK17 and BK18 must be in an open state in order to isolate the fault. The graphical representation of the system at this point is shown in fig. 5, with the black dashed path indicating that the circuit breaker cannot be used for reconfiguration operations.
Fig. 6 is a fault example reconstruction result of realizing reconstruction of a ship distribution network based on a queue intelligent algorithm, a ship power grid reconstruction scheme is generated based on the queue intelligent algorithm, and the problem that a Bus1 meets load requirements to the maximum extent after a fault is specifically described as follows:
object:f(x)＝0.3×5(x1+x2+x3+x4)+0.3×5(x5+x6+x7+x8)+0.15×2x9+0.05×2x10+0.2×4x11
xi0 or 1, i ∈ [1,2, …,11]；
The program first initializes candidate C and variable t of the queue: the candidate C is 5 and the execution time t is 12.
The CI algorithm is executed until saturation (convergence), and as with the previous results, a loop can be skipped, outputting the final result as:
fmax(ri·PLi)＝12.8，fmax(PLi) 44. This shows that when the power generation capacity of the system is insufficient, the loads Load3 and Load4 with lower priority are unloaded to satisfy the power balance of the system, and the power generation capacity of the system is 44MW and the Load is 44MW, thereby proving the effectiveness of the method in the invention.
Claims (1)
1. A ship power distribution network reconstruction method based on a queue intelligent algorithm is characterized by comprising the following steps: the method comprises the following specific steps:
the method comprises the following steps: carrying out graph theory representation on the ship power distribution network aiming at different ship types and power distribution networks;
step two: analyzing faults of the ship power distribution network: after partial area loses power, firstly isolating a fault area, secondly judging whether the existing power supply power of the ship can compensate the power shortage of the area which does not lose power, namely whether the load power requirement of the area which does not lose power can be met by controlling the output power of a generator and a breaker, and if the current line generating capacity is met, not needing to reconstruct a power grid; if the current line power generation capacity is insufficient, power grid reconstruction is needed, namely, load power supply is maximized;
step three: an objective function is established, the reconstruction of the ship power grid is abstracted into a 0-1 knapsack problem,
when the system power generation capacity is insufficient, the load that plays an important role in ship vitality and task execution should be recovered first; the load power supply maximization problem is mathematically described as:
wherein f (x) represents the overall value of the load; x is the number ofiWhether power supply is restored for the load i is shown, 1 represents power supply, and 0 represents no power supply; r isiIndicates the priority of the load i, i.e. the importance of the load, satisfies
PLiRepresents the power demand of load i; pGRepresenting the power generation capacity of the system; and the constraint condition that the sum of the consumed power of all the loads cannot exceed the maximum power generation capacity of the system is realized as an objective function;
decomposing each propulsion load into n equal discrete loads, wherein the corresponding priority coefficient is the propulsion load priority coefficient, and the decision variables are expanded to 2n + 3;
step four: reconstructing a ship power distribution network based on a queue intelligent algorithm, and solving by using the queue intelligent algorithm0-1 backpack problem; firstly, defining the total number C of candidate cases of a queue and the execution times t of each generation of optimization algorithm, initially randomly selecting a group of object loads for each group of candidates C (C1.., C), wherein each group of candidates represents a power supply scheme, all the candidates form a problem queue, and calculating the related value F for each group of candidatesC＝{f(v1),...,f(vc),...,f(vC) And weight FCW＝{f(w1),...,f(wc),...,f(wC)}，
(1) To measure the similarity of each candidate output to the optimal solution, a similarity probability function p is definedcThe following were used:
wherein the content of the first and second substances,
the probability of a value is represented by,
representing the weight probability, and W representing the maximum value of the system capacity;
based on the wheel selection method, each set of candidates C (C1.., C) will select one target candidate f (v)c[？]) To evolve itself, superscript [?]The target candidate is randomly selected by a roulette method based on a similar probability function, and each candidate c cannot be determined in advance;
(2) each candidate performs the above process t times so that each candidate c can obtain an associated value output set Fc,t＝{f(vc)1,...,f(vc)j,...,f(vc)t}, (C ═ 1.., C); after each iteration is completed, selecting the optimal value f (v) of each group of candidates as the initial input of the next iteration; in this way, the value output F of each set of candidates in the queue is updatedC＝{f*(v1),...,f*(vc),...,f*(vC) The scheme is classified as follows:
A. if a solution of the candidate C (C ═ 1., C) is feasible, i.e., satisfies the weight constraint given by equation (1), then one of the following modifications is randomly selected:
(1) adding a randomly selected object from the target candidates, requiring that the object is not included in the current candidate c and still satisfies the weight constraint given by equation (1);
(2) replacing a randomly selected object in the original candidates with a randomly selected object in the target candidates, and satisfying the formula (1);
B. if a solution for candidate C (C1., C) is not feasible, then one of the following modifications is randomly selected:
(1) randomly removing an object from the candidates;
(2) replacing a randomly selected object from the original candidate c with a randomly selected object from the target candidate to reduce the total weight f (w) of the candidate cc) (ii) a Each candidate performs the above process t times; this allows each candidate c to obtain a relevant set of value outputs Fc,t＝{f(vc)1,...,f(vc)j,...,f(vc)t}, (C ═ 1.., C); after each iteration is completed, each candidate selects the optimal value f (v) as the initial input of the next iteration, and since infeasible results may occur in each iteration evolution, the optimal value is selected according to the following conditions:
(2.1.) if the post-evolution result is feasible, selecting the result with the greatest value;
(2.2.) if the post-evolution result is not feasible, selecting the result with the least weight;
(2.3.) if there is an infeasible and feasible change, selecting the feasible result with the greatest value;
in this way, the value output F for each candidate in the queue is updatedC＝{f*(v1),...,f*(vc),...,f*(vC)}
(3) Each group of candidates optimizes the solution thereof in a mutual learning mode; this iterative process continues until the algorithm saturates, i.e., converges, i.e., the result of each candidate output tends to be consistent and does not change in a large number of successive evolution attempts, i.e., the algorithm is considered to converge;
step five: and completing power grid reconstruction, and re-supplying power to the load in the area without power loss to recover the operation of the load.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810916768.1A CN108923427B (en) | 2018-08-13 | 2018-08-13 | Method for reconstructing ship power distribution network based on queue intelligent algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810916768.1A CN108923427B (en) | 2018-08-13 | 2018-08-13 | Method for reconstructing ship power distribution network based on queue intelligent algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108923427A CN108923427A (en) | 2018-11-30 |
| CN108923427B true CN108923427B (en) | 2020-11-03 |
Family
ID=64404500
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810916768.1A Active CN108923427B (en) | 2018-08-13 | 2018-08-13 | Method for reconstructing ship power distribution network based on queue intelligent algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108923427B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112580936A (en) * | 2020-11-30 | 2021-03-30 | 江苏科技大学 | Fault isolation and power grid recovery method in medium-voltage direct-current networking ship power grid |
| CN114142471B (en) * | 2021-11-29 | 2023-08-18 | 江苏科技大学 | Ship comprehensive power system reconstruction method considering communication faults |
| CN116307650B (en) * | 2023-05-24 | 2023-07-25 | 东南大学溧阳研究院 | Novel power distribution network source network load coordination random optimization operation method oriented to flexibility |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093097A (en) * | 2013-01-15 | 2013-05-08 | 天津大学 | Electrical power system fragile section identification method based on normalized-cut |
| CN104820865A (en) * | 2015-03-31 | 2015-08-05 | 浙江工业大学 | Graph-theory-based intelligent optimization method for failure recovery of smart distribution grid |
| CN106451439A (en) * | 2016-11-11 | 2017-02-22 | 哈尔滨工程大学 | Two-stage reconstruction method for power distribution network of comprehensive power propulsion system of ship |
| CN106487001A (en) * | 2015-08-28 | 2017-03-08 | 中国电力科学研究院 | A kind of isolated power system intelligent reconstruction method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9876356B2 (en) * | 2014-10-02 | 2018-01-23 | Mitsubishi Electric Research Laboratories, Inc. | Dynamic and adaptive configurable power distribution system |
-
2018
- 2018-08-13 CN CN201810916768.1A patent/CN108923427B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093097A (en) * | 2013-01-15 | 2013-05-08 | 天津大学 | Electrical power system fragile section identification method based on normalized-cut |
| CN104820865A (en) * | 2015-03-31 | 2015-08-05 | 浙江工业大学 | Graph-theory-based intelligent optimization method for failure recovery of smart distribution grid |
| CN106487001A (en) * | 2015-08-28 | 2017-03-08 | 中国电力科学研究院 | A kind of isolated power system intelligent reconstruction method |
| CN106451439A (en) * | 2016-11-11 | 2017-02-22 | 哈尔滨工程大学 | Two-stage reconstruction method for power distribution network of comprehensive power propulsion system of ship |
Non-Patent Citations (1)
| Title |
|---|
| 基于改进粒子群算法的船舶电力系统故障恢复;杨晨晖;《中国优秀硕士学位论文全文数据库》;20140831;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108923427A (en) | 2018-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108923427B (en) | Method for reconstructing ship power distribution network based on queue intelligent algorithm | |
| Al-Falahi et al. | Power management optimization of hybrid power systems in electric ferries | |
| CN107862405B (en) | Power system grid reconstruction optimization method taking microgrid as black-start power supply | |
| Zohrabi et al. | An overview of design specifications and requirements for the MVDC shipboard power system | |
| Ordono et al. | Interlinking converters and their contribution to primary regulation: a review | |
| Kanellos et al. | Onboard DC grid employing smart grid technology: challenges, state of the art and future prospects | |
| Xu et al. | Optimal power management for failure mode of MVDC microgrids in all-electric ships | |
| CN110690730B (en) | Power and energy control method for hybrid power ship | |
| CN106451439B (en) | A kind of two stages reconstructing method of Marine Synthesize Electric Propulsion System distribution network | |
| Wang et al. | Unintentional islanding transition control strategy for three-/single-phase multimicrogrids based on artificial emotional reinforcement learning | |
| Tang et al. | Multi-agent based power and energy management system for hybrid ships | |
| Cavallo et al. | Supervised energy management in advanced aircraft applications | |
| Nejad et al. | Enhancing active distribution systems resilience by fully distributed self-healing strategy | |
| Zhang et al. | BDI-agent-based quantum-behaved PSO for shipboard power system reconfiguration | |
| Spagnolo et al. | Smart controller design for safety operation of the MEA electrical distribution system | |
| Zhang et al. | Distributed power allocation and scheduling for electrical power system in more electric aircraft | |
| Fobes et al. | Optimal microgrid networking for maximal load delivery in phase unbalanced distribution grids: A declarative modeling approach | |
| Babaei et al. | Real-time implementation of MVDC shipboard power system reconfiguration | |
| Ni et al. | An overview of design, control, power management, system stability and reliability in electric ships | |
| Gaber et al. | An intelligent energy management system for ship hybrid power system based on renewable energy resources | |
| CN113010988B (en) | Post-disaster optimization recovery method and system for AC/DC hybrid power distribution network | |
| Patsios et al. | Discussion on adopting intelligent power management and control techniques in integrated power systems of all-electric ships | |
| Zhang et al. | A novel multi-objective discrete particle swarm optimization with elitist perturbation for reconfiguration of ship power system | |
| CN113541138A (en) | Fan switching method and device suitable for mode conversion of direct-current power transmission system | |
| Siddique et al. | Control strategy for a smart grid-hybrid controller for renewable energy using artificial neuro and fuzzy intelligent system |
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 |