Topology reconstruction and visualization method of ship DC regional power grid based on balance optimizer

Through a balance optimizer-based method and visual simulator, the rapid reconstruction of load power supply in case of ship grid failure is solved, the reliability and recovery speed of the power grid are improved, and the power supply needs of large ships are met.

CN114565179BActive Publication Date: 2024-11-26JIMEI UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210260551.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-16
Publication Date
2024-11-26
Estimated Expiration
2042-03-16

AI Technical Summary

Technical Problem

The prior art is difficult to quickly and effectively reconstruct the load power supply in case of a ship's power grid failure, resulting in a decrease in system reliability and safety. The existing reconstruction methods are slow to converge and poorly result in quality, which cannot meet the power supply needs of large ships.

Method used

Using a balance optimizer-based method, the power supply path from the power to the load is judged through the law of connectivity and the shortest path method, priority unloading according to the importance of the load and power constraints, and use an improved balance optimizer to operate with maximum power supply limit and minimum switching. The number of times is the target for optimization, and the grid reconstruction is designed and visualized by the Qt5 software.

Benefits of technology

It improves the rapid recovery capability of the ship's regional power grid in the event of failure, shortens the reconstruction convergence time, improves the accuracy and reliability of the reconstruction results, realizes timely power supply to important loads, and meets the power supply needs of large ships.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114565179B_ABST
    Figure CN114565179B_ABST
Patent Text Reader

Abstract

The present invention relates to a ship DC regional power grid topology reconstruction and visualization method based on a balanced optimizer. Taking the ship regional power grid as a carrier, the connectivity rule and the shortest path method are used to determine the shortest power supply path from the power supply to the load, so as to determine the corresponding load power supply; determine whether the total power of the load connected to the power supply is overloaded, and if overloaded, unload the load according to the priority level according to the importance of the load; after unloading the load to meet the power constraint, if the corresponding regional power supply still has redundancy to supply power to part of the unloaded load, then use the improved balanced optimizer to optimize and solve with the maximum power supply limit and the minimum number of switch operations as the goal; after unloading the load to meet the power constraint, if each power supply does not meet the power supply to any load, the reconstruction result is directly output. The present invention can resist the impact on the power supply of the ship regional power grid under various emergency conditions, and provide strong power support for offshore ships to meet the actual needs of ships.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of intelligent reconstruction of ship DC regional power grid faults and its visualized operation, and specifically relates to a ship DC regional power grid topology reconstruction and visualization method based on a balancing optimizer. Background Art

[0002] Most of the early ship power grids adopted the dry-feed power supply, in which the front and rear power stations serve as backup for each other. The main power supply and the backup power supply are connected to the same bus (bus) to supply power to all loads. However, when a short circuit occurs near the bus, all loads will lose power, and the redundancy is 100%. Therefore, the power supply reliability is low, and there is no good redistribution of the priority of the loads. Therefore, the traditional radial power system and reconstruction technology can no longer meet the power supply requirements of large and medium-sized ships. With the development of high-power power electronic devices, frequency conversion technology and ship integrated power systems, the ship regional DC power supply system came into being. It takes the power supply of each area as the starting point, divides two lines through the DC / DC converter, and connects to the port DC bus and the starboard DC bus of the ship to supply power to the corresponding regional loads. Among them, the important loads are connected to the two buses at the same time as backup for each other. This arrangement is simple and has higher reliability, so it is gradually becoming the development direction of large ships. At the same time, higher requirements are put forward for the reconstruction technology of ships. With it comes the corresponding intelligent management and control methods for the ship regional power grid, and an intelligent reconstruction technology is indispensable. The ship's power system is the only power supply for the ship. In line with the policy of energy conservation and emission reduction, the ship's power grid reconstruction technology can improve the power supply reliability of the load by changing the switch state without additional cost when the power grid fails. Therefore, we can restore the normal operation of the entire power system by reasonably redistributing the load of the power system at the time of failure. It can be seen that the power grid reconstruction technology plays a pivotal role in ships. When a ship is performing a mission at sea, due to the failure of some equipment, improper operation of the crew or attack, the power system is in an abnormal state, which affects the power supply of important loads, resulting in a decrease in system reliability and safety capabilities and even injuring the crew. It is required that the ship detect the fault area in time when a fault occurs, isolate it from the normal power supply area, and then supply power to the load in the fault area that cannot work due to the damage to the power supply line in time from other lines or backup power supplies. When the power grid fails or is damaged, we require that the power supply to the important loads in the fault area be restored quickly and as much as possible. Therefore, an intelligent fault reconstruction technology is needed to maintain the power supply reliability of the ship's power grid.

[0003] With the development of integrated power systems on ships, the loads carried have increased exponentially in both quantity and power. Faced with increasingly complex working conditions, how to quickly reconstruct the loads when encountering sudden power outages in power grid lines or power supplies has become increasingly important and difficult. On the one hand, the existing technology generally unloads the loads in sequence according to their priority level to provide power for important loads. Although this method is relatively simple, it will obviously cause great power losses and waste time. There are some existing methods based on particle swarm optimization algorithms that have alleviated the fault reconstruction of some ship power grids, such as ship ring power grids. However, the convergence speed is long and the quality of the local and global optimal solutions obtained due to the randomness of the algorithm is poor, which cannot be well applied to ship power grids. When the ship power grid is damaged, the normal power supply to important loads cannot be responded to in time, which cannot meet the actual working needs of the ship. On the other hand, most reconstruction methods are still based on code simulation, which makes the results relatively monotonous and may not be effectively used in the power grid. Summary of the invention

[0004] The purpose of the present invention is to provide a method for topological reconstruction and visualization of a ship DC regional power grid based on a balanced optimizer, taking the ship regional power grid as the control object, formulating a general fault reconstruction strategy and proposing an intelligent reconstruction method, and designing a visual ship fault simulator through Qt5 software to simulate the power grid reconstruction process, so as to provide powerful power support for offshore ships to meet the actual needs of ships.

[0005] To achieve the above-mentioned purpose, the technical scheme of the present invention is: a method for reconstructing and visualizing the topology of a ship DC regional power grid based on a balancing optimizer, taking the ship regional power grid as a carrier, firstly using the connectivity rule and the shortest path method to determine the shortest power supply path from the power supply to the load without considering the power constraint, thereby determining the corresponding load power supply; then according to the power upper limit of each regional power supply, determine whether the total power of the load connected to the power supply is overloaded, and if overloaded, unload the load according to the priority level according to the importance of the load; finally, after unloading the load to meet the power constraint, if the corresponding regional power supply still has redundancy to supply power to part of the unloaded load, then use the improved balancing optimizer to optimize and solve with the maximum power supply limit and the minimum number of switch operations as the goal; after unloading the load to meet the power constraint, if each power supply does not meet the power supply to any load, the reconstruction result is directly output.

[0006] In one embodiment of the present invention, the method is implemented in the following steps:

[0007] Step S1: for a known ship area power grid, the connectivity rule is used to record the connectivity status of all loads and power supplies in their respective connectivity matrices;

[0008] Step S2: using the shortest path processing method according to the connectivity matrix of each load, obtaining the shortest power supply path power supply to which the load is connected;

[0009] Step S3: Check whether the upper limit of each power supply is greater than the total power of the connected loads according to the power constraints of the power supply and the loads. If so, unload the loads according to the priority. Otherwise, directly output the reconstruction result.

[0010] Step S4: After the load is unloaded, check whether the power supply still has a large redundancy;

[0011] Step S5: If the power supply is redundant, the improved balance optimizer is used to re-power part of the unloaded load, and check whether the power constraint is met. If so, the reconstruction result is output;

[0012] Step S6: If the power supply is insufficient to support normal power supply of any load, directly output the reconstruction result;

[0013] Step S7: Use Qt5 software to design a visualized ship area power grid reconstruction simulator, convert the improved balancing optimizer reconstruction method into a DLL dynamic link library, a LIB static link library and a C++ header file, and import them into the simulator to verify the feasibility of the method.

[0014] In one embodiment of the present invention, in step S1, the load connectivity matrix is ​​constructed as follows: first, the circuit diagram is converted into a topological structure according to the number of ship area power grids and loads Load, each line node node is digitally marked in sequence, and then all loads and all power supply sources G are obtained based on the connectivity rule. i The power supply connectivity of each load and power supply path is obtained, Load j =[A, B, C, D, ...]; the dimensions {A, B, C, D, ...} in the matrix represent the connection path from the power supply of each area to each load; in the topological structure, the node node as the starting point represents the power supply, and the remaining nodes node represent the line nodes and loads.

[0015] In one embodiment of the present invention, in step S2, the shortest power supply path is to set the length quantization value between nodes in the connectivity matrix to 1, calculate the power supply of the shortest power supply path of each load, and thereby determine the power supply of the corresponding load:

[0016]

[0017] Above is the power supply path distance matrix obtained by summing the connectivity matrix, Represents the state length value of the power supply line from each power source to the load. Among them, the important load can be powered by four power supplies from the left and right busbars, and the third-level load can only be powered by four power supplies from one side of the busbar. Therefore, the power supply path of the important load is eight power supply lines, and the power supply path of the third-level load is four power supply lines. The distance matrix of the power supply path of the power grid under the fault-free state is compared. The power supply lines that are not equal to the latter will be eliminated, indicating that there is a line disconnection, and then the shortest power supply path is obtained by comparison.

[0018] In one embodiment of the present invention, in step S3, the power constraint is calculated: after determining the power supply status of the power supply source and the load, the total power of the connected load is calculated by the power upper limit of the power supply source to see whether it is overloaded, that is, In the formula Represents the power value of each power supply, Represents the total power value of the load connected to the corresponding power supply.

[0019] In one embodiment of the present invention, in step S5, if the power supply source still has redundancy to supply power to the unloaded part of the load, that is, In the formula Represents the remaining power value after the power supply provides power. Represents the power value of the partially unloaded load. Start the improved balance optimizer to calculate the optimal reconstruction result according to the objective function Load Recon_result .

[0020] In one embodiment of the present invention, in step S6, if the power supply does not have redundancy to supply power to the unloaded part of the load, that is, Then directly output the reconstruction result Load Recon_result Where Represents the remaining power value after the power supply provides power. The power value representing a partially unloaded load.

[0021] In one embodiment of the present invention, the specific implementation method of step S7 is: first, a visual power system structure environment is designed according to the ship area power grid topology map; then, a corresponding output environment is designed according to the required test results; finally, the improved balance optimizer reconstruction method is packaged and converted into a DLL dynamic link library, a LIB static link library and a C++ header file, and the designed simulator is imported, and the electronic switch on the line is clicked to set the fault point, so as to realize the fault reconstruction of the power grid and verify the feasibility of the method.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] The present invention uses the ship as the carrier for the ship regional power system topology reconstruction based on the balance optimizer, and adopts a visualization software platform to realize the ship regional power grid fault reconstruction function. The regional power grid topology structure is processed based on the connectivity graph and the shortest path, and the power grid is intelligently reconstructed by improving the balance optimizer and the objective function. According to the relationship between the power supply and the load, the power constraint is used to select whether to directly output the reconstruction result or to output the result after optimization by the improved balance optimizer. This method saves running memory and shortens the reconstruction convergence time. And the initial candidate solution is determined through steps S3 to S5, thereby avoiding the uncertain solution and slow convergence speed caused by the random initial solution.

[0024] The present invention solves a series of problems in quickly restoring loads and stabilizing power supply when the ship's regional power grid is damaged, so that the ship's regional power grid has a strong reconstruction function. Compared with some methods that completely use random initial solutions to solve the uncertain solutions and local optimal problems caused by ship power grid reconstruction, the accuracy and convergence speed of the optimal solution for power grid reconstruction in this method are significantly improved, effectively improving the ship's regional power grid reconstruction capability. The effectiveness of the visual operation platform test method is tested. By simulating various fault environments, the proposed method can respond quickly, realize the reconstruction of the faulty power grid, and make the power grid stable and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is the topological diagram of the ship area power system.

[0026] Figure 2 It is the process of reconfiguring the ship area power grid.

[0027] Figure 3 It is the connection topology diagram of power supply and load.

[0028] Figure 4 It is an improved balance optimizer reconstruction flowchart.

[0029] Figure 5 It is a flow chart of improving the combination of balance optimizer and simulator.

[0030] Figure 6 This is the test diagram of the ship area power grid structure simulator.

[0031] Figure 7 It is the output environment map of the ship area power grid simulator.

[0032] Figure 8 This is the interface diagram of the ship area power grid simulator software.

[0033] Fig. 9 It is a schematic diagram of a fault point of an example of the present invention.

[0034] Fig.10 This is a simulation result of a ship area power grid fault reconstruction according to an example of the present invention.

[0035] Fig.11 It is the output result of ship area power grid fault reconstruction in one example of the present invention. DETAILED DESCRIPTION

[0036] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings.

[0037] The present invention provides a method and a visual simulator for fault reconstruction of a ship DC regional power system using an improved balancing optimizer. The fault reconstruction function implementation method uses the ship regional power grid as a carrier, such as Figure 1 As shown. A four-step solution is designed. First, the connectivity rule and the shortest path method are used to determine the shortest power supply path from the power supply to the load without considering the power constraint, so as to determine which power supply supplies the load. Then, according to the power upper limit of the power supply in each area, if the total power of the load connected to the power supply is overloaded, the third-level load is first unloaded according to the importance of the load, and the second-level load is continued to be unloaded if the constraint is still not met. Finally, in order to save memory and time, it will be divided into two small steps. The first small step is that after the load is unloaded in the second step to meet the power constraint, if the power supply in the corresponding area still has redundancy to supply power to part of the unloaded load, the improved balance optimizer is used to optimize the solution with the maximum power supply limit and the minimum number of switching operations as the goal. The second small step is that after unloading the load, if each power supply does not meet the power supply of any load, the reconstruction result is directly output. The effectiveness of the reconstruction method is tested by packaging the reconstruction method and importing it into the simulator. Figure 2 The framework of the ship area power grid fault reconstruction process is presented. This method effectively improves the reconstruction capability under the failure of the ship area power grid line or power supply, so that the ship can quickly restore the network power supply to meet the actual needs.

[0038] Reference for the topology reconstruction process of the ship DC regional power system Figure 2 Here are the steps:

[0039] Step 1: For a known regional power grid, the connectivity rules are used to record the connectivity status of all loads and power sources in their respective connectivity matrices;

[0040] Step 2, using the shortest path processing method according to the connectivity matrix of each load, to obtain the closest path power supply to which the load is connected;

[0041] Step 3: Based on the power constraints of the power supply and the load, check whether the upper limit of each power supply is greater than the total power of the connected load. If so, unload the load according to the priority. Otherwise, directly output the reconstruction result.

[0042] Step 4: After the load is unloaded, check whether the power supply still has a large redundancy.

[0043] Step 5: If the power supply is redundant, the improved balancing optimizer is used to partially re-power the unloaded load and check whether the power constraint is met. If so, the reconstruction result is output;

[0044] Step 6: If the power of the power supply is insufficient to support normal power supply of any load, the reconstruction result is directly output.

[0045] Step 7: Use Qt5 software to design a visual ship area power grid reconstruction simulator, convert the improved balance optimizer reconstruction method into a DLL dynamic link library, a LIB static link library, and a C++ header file, and import it into the simulator to verify the feasibility of the method.

[0046] For step 1, load connectivity matrix. First, the circuit diagram must be converted into a topological structure according to the number of regional power grids and loads, such as Figure 1 As shown, each line node is digitally marked in turn, and then all loads and all power sources G are obtained based on the connectivity rule. i The power supply connectivity of each load and power supply path is obtained, Load j =[A, B, C, D, ...]. The dimensions {A, B, C, D, ...} in the matrix represent the connection path from the power supply of each area to each load. The topology is as follows: Figure 3 As shown, the node node (for example, node 1~4 ) represents the power supply, and the rest represent line nodes and loads.

[0047] For step 2, the shortest power supply path. By setting the length between nodes in the connectivity matrix to 1, the shortest power supply path power supply for each load can be calculated, thereby determining which power supply the load will be powered by:

[0048]

[0049] Above is the power supply path distance matrix obtained by summing the connectivity matrix, Represents the state length value of the power supply line from each power source to the load. Among them, the important load can be powered by four power supplies from the left and right busbars, and the third-level load can only be powered by four power supplies from one side of the busbar. Therefore, the power supply path of the important load is eight power supply lines, and the power supply path of the third-level load is four power supply lines. The distance matrix of the power supply path of the power grid under the fault-free state is compared. The power supply lines that are not equal to the latter will be eliminated, indicating that there is a line disconnection, and then the shortest power supply path is obtained by comparison.

[0050] For step 3, power constraint calculation. After determining the power supply status of the power supply and the load, calculate whether the total power of the connected load is overloaded by using the power upper limit of the power supply. In the formula Represents the power value of each power supply, Represents the total power value of the load connected to the corresponding power supply. If overloaded, first unload the third-level load according to the load priority, then the second-level load. Then proceed to step 4, otherwise directly output the reconstruction result Load Recon_result .

[0051] For step 4, after the operation in step 3, a preliminary reconstruction result is obtained, but there is still some room for improvement.

[0052] For step 5, if the power supply still has a certain redundancy to supply part of the unloaded load, In the formula Represents the remaining power value after the power supply provides power. Represents the power value of the partially unloaded load. Start the improved balance optimizer to calculate the optimal reconstruction result according to the objective function Load Recon_result ,like Figure 4 The figure shows the solution flow chart of EO. m is the number of power-off loads.

[0053] For step 6, if the power supply does not have redundancy to supply the unloaded part of the load, Then directly output the reconstruction result Load Recon_result Where Represents the remaining power value after the power supply provides power. The power value representing a partially unloaded load.

[0054] For step 7, a visual simulator is designed using Qt5.14.2 software according to the ship area power grid structure diagram to test the feasibility of improving the balance optimizer. Figure 5 As shown in Figure 1, firstly, a visual power system structure environment is designed based on the ship area power grid topology diagram, as shown in Figure 1. Figure 6 Then design the corresponding output environment according to the required test results, such as Figure 7 Finally, the improved balance optimizer reconstruction method is packaged and converted into a DLL dynamic link library, a LIB static link library and a C++ header file, imported into the designed simulator, and the electronic switch on the line is clicked to set the fault point to achieve the fault reconstruction of the power grid and verify the feasibility of the method. The overall structure of the simulator environment is shown in Figure 8 shown.

[0055] The following is a specific application example of the present invention.

[0056] Equilibrium optimizer (EO) sometimes encounters the shortcomings of high randomness and falling into local optimal solutions. By setting the initial candidate solution according to the power loss load data that needs to be redistributed, the infeasible solutions are reduced. Grid redistribution belongs to discrete switch reconstruction, so EO is improved to discretized DEO.

[0057] As we all know, when a ship is performing a mission at sea, due to some equipment failure, improper operation of the crew or attack, the power system is in an abnormal state, which affects the power supply of important loads, reduces the reliability and safety of the system, and even injures the crew. It is crucial to install the reconstruction technology of the power grid. It requires the ship to detect the fault area in time when a fault occurs, and isolate it from the normal power supply area, and then the load in the fault area that cannot work due to the damage of the power supply line is promptly powered by other backup power sources. At present, the ship's regional power grid is still in development and its reconstruction technology is not perfect. Therefore, a reconstruction technology for the ship's regional power grid is invented. In addition, most reconstruction methods are still based on code simulation, which may cause many inoperable situations in practical applications. Therefore, in order to enable the crew to intuitively operate and observe the reconstruction state of the power grid, this paper designs a visual and operable simulator based on the ship's regional power grid structure to verify the feasibility and effectiveness of the proposed reconstruction method. EO is an optimization algorithm inspired by the physical phenomenon of control volume mass balance. By estimating the control volume mass balance model of dynamic and equilibrium states, the solution concentration is used as a search agent to update the optimal solution. The most outstanding feature of the optimizer is that it will obtain four best solutions so far when solving, and then update the agent through the improved mass balance equation to obtain better local and global optimal solutions. EO has a high randomness, which makes its working efficiency low and it is not suitable for directly using it for discrete problems because it is a continuous problem. Therefore, it is necessary to improve EO into a discrete form and reduce the randomness of the solution to improve the working efficiency of EO. The reconstruction objective function is integrated into DEO, and the function makes DEO to re-distribute the power supply to the power-off load according to the maximum power value. On the one hand, it realizes the power supply of the maximum number of loads, and on the other hand, it improves the power supply efficiency of the power supply and reduces the loss caused by no-load. The initial candidate solution is generated by re-powering the load after power failure, avoiding the randomness of the solution, and greatly reducing the amount of calculation when searching the global space, thereby improving the convergence speed.

[0058] By adopting the improved balancing optimizer to realize the reconstruction of the ship area power grid, the load power supply state is taken as the feasible solution, and the objective function is combined for calculation. The software simulator of the ship area power grid is visualized through the QT5 software. The process of self-repair of the power grid can be clearly seen during the fault reconstruction process. The specific process is as follows:

[0059] Step 1: According to the ship area power grid topology diagram, through connectivity and the shortest path rule of load power supply, a connectivity matrix is ​​set for each load, where important loads have eight power supply paths and other loads have four power supply paths. For example, load L13:

[0060]

[0061] The above formula is the connectivity matrix Load of the three-level load L13 13 , the path between the power supply and the load L13 is in the brackets. i Represents the power supply, x j Represents the status value of the power supply line between nodes, which consists of 0 and 1. 0 represents that the line is disconnected, and 1 represents that the line is normal.

[0062] The order of the rows determines the priority power supply path of the load. The length between nodes is quantized to 1. The connectivity matrix is ​​summed to obtain the path length of each power supply to the load in the default state. in is the sum of the connectivity matrices, It is the power supply path distance matrix between the power source and the load in the fault-free power grid.

[0063] Step 2: When diagnosing a power grid fault, the current power supply path length of the power load can be used and The shortest path for powering the load can be determined, and the power supply that powers the load can be determined, and the reconstruction result of the load can be obtained at the same time.

[0064] Step 3: After determining the power supply status and power source of the load, calculate the total power of the load connected to the power supply in each area. If the power constraint is met, the reconstruction result Load is directly output Recon_result (k,:). If If the power supply is overloaded, the load under the overloaded power supply will be unloaded according to the load power supply priority until the power constraint is met.

[0065] Step 4: After unloading the load, if the remaining power of the power supply is not enough to re-power any power-off load, Then directly output the reconstruction result Load Recon_result (k, :).

[0066] Step 5: If Start DEO to optimize and reconstruct the power-off load. According to the above reconstruction results, the number and length of the power-off load are determined as:

[0067] len lose=length(find(Load Recon_result (k,:)<1))

[0068] In the formula, find() is used to find the number less than 1 in the reconstruction result, and length() is used to get the number of power-off loads in the reconstruction result. lose is the total number of power-off loads. Recon_result (k,:) is the reconstruction result, consisting of 0 and 1.

[0069] Then, the candidate solution is initialized according to the power-off load length, and the logical mapping chaos model is adopted as follows:

[0070] C(num×len lose ,length(Load Recon_result ))=2×rand()×(1-rand())

[0071] Where rand() is a random number in (0, 1), num is a constant, and is calculated by comparing it with the number of power-off loads len. lose Multiplication increases the number of initial candidate solutions and improves the globality of DEO's search space. Recon_result ) is the total number of loads in the power grid. C represents the initial candidate solution.

[0072] The candidate solution reflects the power supply state of the load according to the switch state, so it is discretized to consist of 0 and 1. That is, if C>num1, then C=1, otherwise C=0. C represents the candidate solution, and num1 represents a constant.

[0073] Power constraint calculation is performed based on the obtained initial candidate solution

[0074]

[0075] Where PLoad i Represents the total power of the load connected to the power supply in each area. Set the DEO optimization parameters, including the volume constant y, weights a1, a2, the mass generation rate G inside the control volume, the number of iterations Maxiter, the dimension Dim, and the upper and lower limits ub, lb of the candidate solution.

[0076] Entering the iteration process, first check whether C exceeds the limit. If so, set C equal to the corresponding maximum value of the candidate solution closed interval. Then calculate the objective function:

[0077]

[0078] Where x represents the candidate solution array, L represents the power value of the load, and M represents the load weight coefficient matrix. According to the load priority, the first-level, second-level, and third-level load weight coefficients are num2, num3, and num4 respectively. num5 represents the function The function sum(x.*L.*M) represents the maximum restoration of load power supply, and the function Represents the minimum number of switching times.

[0079] So the fitness function is

[0080]

[0081] If the total power of the load connected to the power supply is less than the total power of the load, the fitness value is output normally, otherwise fitness = 0, so that the maximum optimal solution can be quickly calculated.

[0082] The initial optimal fitness value is

[0083] best_fit=[best_fit1, best_fit2, best_fit3, best_fit4]=0

[0084] Where best_fit in the matrix i Represents the four optimal fitness values ​​set initially. Since we want to find the maximum value, we set them all to 0. This is for subsequent comparison and assignment.

[0085] By comparing best_fit with fitness, fitness is assigned to best_fit.

[0086]

[0087] Then best = [best1, best2, best3, best4] and best_fit values ​​can be output in sequence. i is the optimal candidate solution corresponding to the optimal fitness value, and best is the current optimal candidate solution matrix.

[0088] By averaging best to get best_ave, we get an optimal candidate solution balance pool, best_pool = [best, best_ave]. We discretize it, best_pool = round (best_pool). In the formula, round() rounds off best_pool to discretize it.

[0089] The candidate solution is updated by controlling the first-order differential equation of the dynamic mass balance of the volume:

[0090]

[0091] Where V is the volume constant, G is the mass generation rate inside the control volume, Q is the volume flow rate into or out of the control volume, C is the solution density, i.e., the candidate solution, C eq represents the concentration in the equilibrium state, that is, the candidate solution taken from the equilibrium pool.

[0092] By solving the equation we can get:

[0093] C(j,:)=C eq +(C(j,:)-C eq )×F+(G. / λV)×(1-F)

[0094] In the formula, F represents the exponential term coefficient and λ represents the flow rate. From the balancing pool best_pool, a current optimal candidate solution is taken with the same probability and assigned to C eq Update the candidate solution.

[0095] Then discretize C(j,:):

[0096] C(j,:)=round(C(j,:))

[0097] The candidate solution is updated. The total load power connected to the current power source is updated and calculated by the updated candidate solution. Thus, an iterative update is completed, and an optimal candidate solution and optimal fitness value that meet the requirements are obtained.

[0098] According to the improved reconstruction method, it is packaged and converted into dynamic link library DLL, static link library LIB and h header file based on C++. Then, according to the ship area power grid structure diagram, Qt5.14.2 is used to design the simulator interface. Finally, the packaged DEO file is imported into the simulator. By clicking the line button to create a fault point, the power grid will immediately detect the fault section and isolate it, and start the reconstruction algorithm DEO to perform fault reconstruction. Figure 1 In the example, assume that line 23 and line 8 are disconnected. Figure 6 The simulator x23 and x8 are disconnected, such as Fig. 9 As shown. Click the switch at x23 and x8, and the button will change from closed to open. The power grid system detects a system line fault and immediately starts DEO for reconstruction. The reconstruction result is shown in Fig.10 As shown in Figure 1, only load L8 is unloaded. The output result is Fig.11 As shown in the figure, the load reconstruction results, active utilization and operation time of the four power supply groups meet the power constraint requirements and achieve the purpose of restoring the power supply of important loads in microseconds. Therefore, the results of the simulator and the reconstruction method verify the feasibility of this method.

[0099] The above are preferred embodiments of the present invention. Any changes made according to the technical solution of the present invention, as long as the resulting functions do not exceed the scope of the technical solution of the present invention, belong to the protection scope of the present invention.

Claims

1. A method for topology reconstruction and visualization of a ship DC regional power grid based on a balanced optimizer, characterized in that: Taking the ship regional power grid as the carrier, the connectivity rule and the shortest path method are first used to determine the shortest power supply path from the power source to the load without considering the power constraint, so as to determine the power supply source of the corresponding load; then, according to the power upper limit of each regional power supply, it is determined whether the total power of the load connected to the power supply is overloaded. If it is overloaded, the load is unloaded according to the priority level according to the importance of the load; finally, after the load is unloaded to meet the power constraint, if the corresponding regional power supply still has redundancy to supply power to part of the unloaded load, the improved balance optimizer is used to optimize and solve with the maximum power supply limit and the minimum number of switch operations as the goal; after the load is unloaded to meet the power constraint, if each power supply does not meet the power supply of any load, the reconstruction result is directly output; The method realizes the reconstruction of the ship area power grid by adopting an improved balancing optimizer, takes the load power supply state as a feasible solution, and performs calculations in combination with the objective function. The ship area power grid is visualized through QT5 software, and the process of power grid self-repair can be clearly seen during the fault reconstruction process; the specific process is as follows: Step 1: According to the ship area power grid topology diagram, a connectivity matrix is ​​set for each load through connectivity and the shortest path rule of load power supply, where important loads have eight power supply paths and other loads have four power supply paths; for the third-level load L13: The above formula is the connectivity matrix Load of the three-level load L13 13 , the path between the power supply and the load L13 is in the brackets, where G i Represents the power supply, x j Represents the status value of the power supply line between nodes, which consists of 0 and 1. 0 represents disconnection and 1 represents normal operation. The order of the rows determines the priority power supply path for the load; The length between nodes is quantified to 1; The connectivity matrix is ​​summed to obtain the path length of each power supply to the load in the default state. in is the sum of the connectivity matrices, The power supply path distance matrix between the power source and the load of the fault-free power grid; Step 2: When diagnosing a power grid fault, use the current power supply path length of the power load and Determine the shortest path for powering the load, the power supply for the load, and obtain the reconstruction result of the load; Step 3: After determining the power supply status and power supply of the load, calculate the total power of the load connected to each regional power supply. If the power value of each power supply is Represents the total power value of the load connected to the corresponding power supply. If it meets the power constraint, the reconstruction result Load is directly output. Recon_result (k,:); if If the power supply is overloaded, the load under the overloaded power supply will be unloaded according to the load power supply priority until the power constraint is met; Step 4: If the remaining power of the power supply after unloading the load is not enough to re-power any power-off load, the remaining power value after the power supply provides power Then directly output the reconstruction result Load Recon_result (k,:); Step 5: If Start DEO to optimize and reconstruct the power-off loads; according to the reconstruction results, the number and length of the power-off loads are determined as: len lose =length(find(Load Recon_result (k,:)<1)) In the formula, find() is used to find the number less than 1 in the reconstruction result, and length() is used to get the number of power-off loads in the reconstruction result. lose is the total number of power-off loads, Load Recon_result (k,:) is the reconstruction result, consisting of 0 and 1; Then, the candidate solution is initialized according to the power-off load length, and the logical mapping chaos model is adopted as follows: C(num×len lose ,length(Load Recon_result ))=2×rand()×(1-rand()) Where rand() is a random number in (0,1), num is a constant, and is calculated by comparing it with the number of power-off loads len. lose Multiply to increase the number of initial candidate solutions, length(Load Recon_result ) is the total number of loads in the power grid, and C represents the candidate solution; The candidate solution reflects the power supply state of the load according to the switch state, so it is discretized to consist of 0 and 1; that is, if C>num1, then C=1, otherwise C=0, num1 represents a constant; Power constraint calculation is performed based on the obtained initial candidate solution Where PLoad i represents the total power of the load connected to the power supply in each area, and sets the DEO optimization parameters, including the volume constant V, and the mass generation rate G inside the control volume; Entering the iteration process, first check whether C exceeds the limit. If so, set C equal to the corresponding maximum value of the candidate solution closed interval, and then calculate the objective function: Where x represents the candidate solution array, L represents the power value of the load, and M represents the load weight coefficient matrix. According to the load priority, the first-level, second-level, and third-level load weight coefficients are num2, num3, num4, and num5, respectively, representing the function The weight coefficient of the function sum(x.*L.*M) represents the maximum restoration of load power supply, and the function Represents the minimum number of switching times; So the fitness function is If the total power of the load connected to the power supply is less than the total power of the load, the fitness value is output normally, otherwise fitness = 0, so as to quickly calculate the maximum optimal solution; The initial optimal fitness value is best_fit=[best_fit1, best_fit2, best_fit3, best_fit4]=0 Where best_fit in the matrix i Represents the four optimal fitness values ​​initially set. Since the maximum value is required, they are all set to 0 for subsequent comparison and assignment; By comparing best_fit with fitness, fitness is assigned to best_fit. If Then output best = [best1, best2, best3, best4] and best_fit value in turn; among which best i is the optimal candidate solution corresponding to the optimal fitness value, and best is the current optimal candidate solution matrix; By averaging best to get best_ave, we get an optimal candidate solution balance pool best_pool = [best, best_ave]; discretize it, best_pool = round (best_pool), where round() rounds off best_pool for discretization; The candidate solution is updated by controlling the first-order differential equation of the dynamic mass balance of the volume: Where V is the volume constant, G is the mass generation rate inside the control volume, Q is the volume flow rate into or out of the control volume, and C is the solution density, i.e., the candidate solution. eq represents the concentration in the equilibrium state, that is, the candidate solution taken from the equilibrium pool; By solving the equation we can get: C(j,:)=C eq +(C(j,:)-C eq )×F+(G. / λV)×(1-F) In the formula, F represents the exponential term coefficient, λ represents the flow rate, and a current optimal candidate solution is taken from the balancing pool best_pool with the same probability and assigned to C eq Update candidate solutions; Then discretize C(j,:): C(j,:)=round(C(j,:)) The candidate solution is updated; the total load power connected to the current power source is updated and calculated by the updated candidate solution, and an iterative update is completed to obtain an optimal candidate solution and an optimal fitness value that meet the requirements; According to the improved reconstruction method, it is packaged and converted into dynamic link library DLL, static link library LIB and C++ based h header file; then the simulator interface is designed using Qt5.14.2 according to the ship area power grid structure diagram, and finally the packaged DEO file is imported into the simulator, and the fault point is created by clicking the line button. At this time, the power grid will immediately detect the faulty section and isolate it, and start the reconstruction algorithm DEO to perform fault reconstruction, verifying the feasibility of the improved reconstruction method.

2. The method for topology reconstruction and visualization of ship DC regional power grid based on balanced optimizer according to claim 1 is characterized in that: The steps to implement this method are as follows: Step S1: for a known ship area power grid, the connectivity rule is used to record the connectivity status of all loads and power supplies in their respective connectivity matrices; Step S2: using the shortest path processing method according to the connectivity matrix of each load, obtaining the shortest power supply path power supply to which the load is connected; Step S3: Check whether the upper limit of each power supply is greater than the total power of the connected loads according to the power constraints of the power supply and the loads. If so, unload the loads according to the priority. Otherwise, directly output the reconstruction result. Step S4: After the load is unloaded, check whether the power supply still has a large redundancy; Step S5: If the power supply is redundant, the improved balance optimizer is used to re-power part of the unloaded load, and check whether the power constraint is met. If so, the reconstruction result is output; Step S6: If the power supply is insufficient to support normal power supply of any load, the reconstruction result is directly output; Step S7: Use Qt5 software to design a visualized ship area power grid reconstruction simulator, convert the improved balancing optimizer reconstruction method into a DLL dynamic link library, a LIB static link library and a C++ header file, and import them into the simulator to verify the feasibility of the method.

3. The method for topology reconstruction and visualization of ship DC regional power grid based on balanced optimizer according to claim 2 is characterized in that: In step S1, the load connectivity matrix is ​​constructed: first, the circuit diagram is converted into a topological structure according to the number of ship area power grids and loads, each line node is digitally marked in sequence, and then all loads and all power supplies G are obtained based on the connectivity rule. i The power supply connectivity of each load and power supply path is obtained, Load j =[A; B; C; D; ...]; the dimensions {A; B; C; D, ...} in the matrix represent the connection path from the power supply of each area to each load; in the topological structure, the node node as the starting point represents the power supply, and the remaining nodes node represent the line nodes and loads.

4. The method for topology reconstruction and visualization of a ship DC regional power grid based on a balanced optimizer according to claim 2 is characterized in that: In step S2, the shortest power supply path is to set the length quantization value between nodes in the connectivity matrix to 1, calculate the power supply of the shortest power supply path of each load, and thereby determine the power supply of the corresponding load: In the formula, is the power supply path distance matrix obtained by summing the connectivity matrix, Represents the state length value of the power supply line from each power supply source to the load; the important load is powered by the port and starboard busbars through four power supplies, and the tertiary load is powered by the corresponding side busbar through four power supplies. Therefore, the power supply path for the important load is eight power supply lines, and the power supply path for the tertiary load is four power supply lines; By calculating the The distance matrix of the power supply path of the power grid under the fault-free state is compared. The power supply lines that are not equal to the latter will be eliminated, indicating that there is a line disconnection, and then the shortest power supply path is obtained by comparison.

5. The method for topology reconstruction and visualization of ship DC regional power grid based on balanced optimizer according to claim 2 is characterized in that: In step S3, the power constraint is calculated: after determining the power supply status of the power supply source and the load, the total power of the connected load is calculated by the power upper limit of the power supply source to see whether it is overloaded, that is, In the formula Represents the power value of each power supply, Represents the total power value of the load connected to the corresponding power supply.

6. The method for topology reconstruction and visualization of ship DC regional power grid based on balanced optimizer according to claim 2 is characterized in that: In step S5, if the power supply still has redundancy to supply power to the unloaded part of the load, that is, Start the improved balance optimizer to calculate the optimal reconstruction result based on the objective function Load Recon_result , where Represents the remaining power value after the power supply provides electricity. The power value representing a partially unloaded load.

7. The method for topology reconstruction and visualization of ship DC regional power grid based on balanced optimizer according to claim 2 is characterized in that: In step S6, if the power supply does not have redundancy to supply power to the unloaded part of the load, that is, Then directly output the reconstruction result Load Recon_result , where Represents the remaining power value after the power supply provides electricity. The power value representing a partially unloaded load.

8. The method for topology reconstruction and visualization of ship DC regional power grid based on balanced optimizer according to claim 2 is characterized in that: The specific implementation method of step S7 is: first, design a visualized power system structure environment according to the ship area power grid topology diagram; then design a corresponding output environment according to the required test results; finally, package and convert the improved balance optimizer reconstruction method into a DLL dynamic link library, a LIB static link library and a C++ header file, import the designed simulator, click the electronic switch on the line to set the fault point, realize the fault reconstruction of the power grid, and verify the feasibility of the method.