Ship power grid fault reconstruction system

Abstract

The invention discloses a ship power grid fault reconstruction system, and belongs to the technical field of ship power systems. The system comprises a power grid modeling module, a fault feature extraction module, a fault positioning module, a mathematical model reconstruction module, a reconstruction optimization module and a real-time monitoring and execution module. A power grid topology model is constructed based on a graph theory and topology, and a basis is provided for subsequent analysis. Fractional Fourier transform (FRFT) is used to extract fault features, and a neural network is combined to accurately position a fault. And a reconstruction mathematical model is established by considering multiple targets and constraints, and an optimal reconstruction scheme is obtained by improving a sparrow algorithm to optimize and solve. And the real-time monitoring and execution module acquires data, receives a scheme and controls a switch to realize reconstruction and state monitoring. The system can quickly position and reconstruct ship power grid faults, gives consideration to multi-objective optimization and actual constraints, effectively improves the reliability and stability of a ship power system, and guarantees the safe and stable operation of a ship.

Description

Technical Field

[0001] The present invention relates to the technical field of ship power systems, and in particular to a ship power grid fault reconstruction system. Background Art

[0002] At present, with the development of ship integrated power systems, their complexity and importance have been continuously increasing. Once a fault occurs in the ship power grid, many deficiencies are exposed in traditional fault reconstruction methods. On the one hand, the fault reconstruction methods for land power grids are difficult to apply due to the differences in load characteristics, line parameters, etc. of ship power systems. On the other hand, current research on ship power grid fault reconstruction is mostly based on intelligent algorithms, but existing intelligent algorithms have defects such as slow convergence speed and easy to fall into local optimum, and do not fully consider multi-objective optimization and complex constraint conditions in ship power grid reconstruction, resulting in poor reconstruction effects and unable to meet the requirements of ships for high reliability and stability of power systems. Summary of the Invention

[0003] The object of the present invention is to provide a ship power grid fault reconstruction system that can accurately and efficiently reconstruct the faulted ship power grid. This system should not only achieve rapid restoration of power supply to ensure the normal operation of key ship loads, but also comprehensively consider various factors, optimize the operation state of the power grid, and improve the overall performance and reliability of the ship power system.

[0004] To achieve the above object, the present invention provides a ship power grid fault reconstruction system, including a power grid modeling module based on graph theory and topology, a fault feature extraction module based on FRFT, a fault location module integrating neural networks, a reconstruction mathematical model module considering multi-objectives and constraints, a reconstruction optimization module improving the sparrow algorithm, and a real-time monitoring and execution module. The power grid modeling module based on graph theory and topology is respectively connected to the fault feature extraction module based on FRFT, the fault location module integrating neural networks, and the reconstruction mathematical model module considering multi-objectives and constraints. The fault feature extraction module based on FRFT is connected to the fault location module integrating neural networks and the real-time monitoring and execution module. The fault location module integrating neural networks is connected to the real-time monitoring and execution module through the reconstruction mathematical model module considering multi-objectives and constraints and the reconstruction optimization module improving the sparrow algorithm.

[0005] Preferably, the power grid modeling module based on graph theory and topology constructs a topological model of the ship integrated power system distribution network based on graph theory and topology principles, and uses matrix representation to describe the connection relationship between nodes and edges. The formula of the matrix representation is:

[0006]

[0007] Preferably, the fault feature extraction module based on FRFT adopts the fractional Fourier transform (FRFT) technology, and its basic definition formula is:

[0008]

[0009] Among them, K p (t, u) is the kernel function of FRFT, and p is the transformation order;

[0010] By determining the optimal transformation order and adopting the sampling type DFRFT algorithm, characteristic parameters such as the power spectrum and energy rate of the fault signal are extracted.

[0011] Preferably, the fault location module of the fusion neural network combines the fault feature vector extracted by FRFT, and uses the BP neural network and RBF neural network for fault location. Among them, the BP neural network adjusts the weights and biases through the backpropagation algorithm, and its error calculation formula is:

[0012]

[0013] Among them, m is the number of samples, y k is the expected output, is the actual output.

[0014] Preferably, the fault reconstruction mathematical model established by the reconstruction mathematical model module considering multiple objectives and constraints includes multiple objective functions, including:

[0015] Power consumption load restoration objective:

[0016]

[0017] Among them, w i is the importance coefficient of the i-th load, P i is the power of the i-th load, S i is the state of the i-th switch after reconstruction;

[0018] Objective of reducing the number of switch operations:

[0019]

[0020] Among them, S i is the state of the i-th switch after reconstruction; is the state of the i-th switch before reconstruction, and m is the total number of switches;

[0021] Load balancing objective:

[0022]

[0023] Among them, l is the number of main distribution boards, J jIt is a set of loads connected to the first main distribution board.

[0024] Preferably, the reconstruction optimization module of the improved sparrow algorithm uses cubic mapping for population initialization and introduces a sine-cosine search strategy into the follower position update formula. The improved formula is:

[0025]

[0026] where a is a constant greater than 1, and S1, S2, S3, and S4 are random numbers following a uniform distribution. is the current global optimal position. is the position of the i-th sparrow at the j-th dimension in the t-th iteration.

[0027] Preferably, the real-time monitoring and execution module uses the power grid monitoring technology based on FTU to collect the operation data of the ship power grid in real time. When a fault is detected, it controls the switch actions according to the reconstruction scheme generated by other modules, realizes power grid reconstruction, and monitors the state of the reconstructed power grid in real time.

[0028] Preferably, the fault feature extraction module based on FRFT uses a two-dimensional peak search method to determine the optimal transformation order of FRFT and normalizes the extracted feature parameters.

[0029] Preferably, the fault location module integrating neural network determines the number of neurons in the input layer of the neural network according to the dimension of the fault feature vector, determines the number of neurons in the hidden layer according to the empirical formula and experimental debugging, and determines the number of neurons in the output layer according to the fault location target.

[0030] Preferably, the reconstruction optimization module of the improved sparrow algorithm initializes the sparrow population size, the maximum number of iterations, and the safety threshold; uses the improved sparrow search algorithm to solve the reconstruction mathematical model and outputs the optimal power grid reconstruction scheme.

[0031] Therefore, a ship power grid fault reconstruction system with the above structure of the present invention has the following beneficial effects:

[0032] The present invention provides a clear structural framework for fault location and reconstruction by constructing an accurate topological model, improving the accuracy and efficiency of analysis. The fault feature extraction method based on FRFT can effectively process non-stationary signals and extract accurate fault features, providing a reliable basis for fault location. The fault location method integrating neural networks utilizes the powerful learning ability of neural networks to achieve fast and accurate fault location, improving the timeliness of fault handling. The reconstruction mathematical model considering multiple objectives and constraints comprehensively takes into account various actual requirements and limiting conditions in the reconstruction of ship power grids, making the reconstruction scheme more reasonable and optimized. The improved sparrow algorithm demonstrates better global search ability and convergence speed when solving the reconstruction mathematical model, and can quickly find the optimal reconstruction scheme.

[0033] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Brief Description of the Drawings

[0034] Figure 1 It is a schematic structural diagram of a ship power grid fault reconstruction system of the present invention;

[0035] Figure 2 It is a schematic structural diagram of a ship power grid fault reconstruction system of the present invention. Detailed Embodiments

[0036] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0037] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left" and "right" are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0038] Embodiment

[0039] As Figure 1As shown in the figure, the present invention provides a ship power grid fault reconstruction system, including a power grid modeling module based on graph theory and topology, a fault feature extraction module based on FRFT, a fault location module integrating neural networks, a reconstruction mathematical model module considering multiple objectives and constraints, a reconstruction optimization module improving the sparrow algorithm, and a real-time monitoring and execution module. The power grid modeling module based on graph theory and topology is respectively connected to the fault feature extraction module based on FRFT, the fault location module integrating neural networks, and the reconstruction mathematical model module considering multiple objectives and constraints. The fault feature extraction module based on FRFT is connected to the fault location module integrating neural networks and the real-time monitoring and execution module. The fault location module integrating neural networks is connected to the real-time monitoring and execution module through the reconstruction mathematical model module considering multiple objectives and constraints and the reconstruction optimization module improving the sparrow algorithm.

[0040] The power grid modeling module based on graph theory and topology constructs a topological model of the ship integrated power system distribution network based on graph theory and topology principles, and uses the matrix representation method to describe the connection relationship between nodes and edges. The formula of the matrix representation method is:

[0041]

[0042] The fault feature extraction module based on FRFT uses the fractional Fourier transform (FRFT) technology, and its basic definition formula is:

[0043]

[0044] Among them, K p (t, u) is the kernel function of FRFT, and p is the transformation order;

[0045] By determining the optimal transformation order and using the sampling type DFRFT algorithm to extract characteristic parameters such as the power spectrum and energy rate of the fault signal.

[0046] The fault location module integrating neural networks combines the fault feature vectors extracted by FRFT, and uses the BP neural network and the RBF neural network for fault location. Among them, the BP neural network adjusts the weights and biases through the backpropagation algorithm, and its error calculation formula is:

[0047]

[0048] Among them, m is the number of samples, y k is the expected output, is the actual output.

[0049] The fault reconstruction mathematical model established by the reconstruction mathematical model module considering multiple objectives and constraints includes multiple objective functions, including:

[0050] Power consumption load restoration objective:

[0051]

[0052] Among them, w i is the importance coefficient of the i-th load, and P i is the power of the i-th load, and S i is the state of the i-th switch after reconstruction;

[0053] Objective of reducing the number of switch operations:

[0054]

[0055] Among them, S i is the state of the i-th switch after reconstruction; is the state of the i-th switch before reconstruction, and m is the total number of switches;

[0056] Load balancing objective:

[0057]

[0058] Among them, l is the number of main switchboards, and J j is the set of loads connected to the i-th main switchboard.

[0059] The reconstruction optimization module of the improved sparrow algorithm uses cubic mapping for population initialization and introduces a sine-cosine search strategy in the follower position update formula. The improved formula is:

[0060]

[0061] Among them, a is a constant greater than 1, S1, S2, S3, S4 are random numbers obeying a uniform distribution, is the current global optimal position, is the position of the i-th sparrow at the j-th dimension in the t-th iteration.

[0062] The real-time monitoring and execution module uses the power grid monitoring technology based on FTU to collect the operation data of the ship power grid in real time. When a fault is detected, it controls the switch actions according to the reconstruction scheme generated by other modules to realize power grid reconstruction and monitor the state of the reconstructed power grid in real time.

[0063] The fault feature extraction module based on FRFT uses a two-dimensional peak search method to determine the optimal transformation order of FRFT and normalizes the extracted feature parameters.

[0064] The fault location module integrating neural networks determines the number of neurons in the input layer of the neural network according to the dimension of the fault feature vector, determines the number of neurons in the hidden layer according to the empirical formula and experimental debugging, and determines the number of neurons in the output layer according to the fault location objective.

[0065] The reconstruction and optimization module based on the improved sparrow algorithm initializes the sparrow population size, maximum number of iterations, and safety threshold; uses the improved sparrow search algorithm to solve the reconstruction mathematical model, and outputs the optimal power grid reconstruction plan.

[0066] As Figure 2 described above, the working steps of a ship power grid fault reconstruction system provided by the present invention are as follows:

[0067] S1. Power grid modeling based on graph theory and topology: According to the actual layout of the ship integrated power system, components such as busbars, switches, and loads are transformed into nodes and edges in the topological graph. Using relevant concepts in graph theory, the adjacency relationship between nodes is determined. If two nodes are connected by a line, they are adjacent to each other. The matrix representation method is used to clearly describe the connection relationship of the power grid components and complete the construction of the topological model. This model can intuitively display the power grid structure and provide a basic framework for subsequent analysis. When node i is connected to node j, A ij = 1; otherwise, A ij = 0.

[0068] S2. Fault feature extraction based on FRFT: When a fault occurs in the ship power grid, the real-time monitoring and execution module collects the fault signal and transmits it to this module. The collected signal is preprocessed to remove noise interference. Using the two-dimensional peak search method, the order p of the fractional Fourier transform (FRFT) is selected one by one, the FRFT is performed and the (p, u) parameter distribution diagram is drawn, and the order p corresponding to the point with the highest energy peak is selected as the optimal order. The sampling type DFRFT (such as the Ozaktas sampling type algorithm) is used to process the signal, and the calculation is performed according to the basic definition formula of the FRFT:

[0069]

[0070] Among them, K p (t, u) is the kernel function of the FRFT. Calculate the power spectrum P(u) of the signal, and its expression is

[0071] P(u) = |F p (u)| 2 ;

[0072] N is the total number of sampling points;

[0073] When the fixed frequency band is divided into sub-bands, the energy rate Ei of the i-th sub-band is calculated, and the energy rate Ei is calculated, where These parameters are integrated into a feature vector and normalized to provide accurate data for fault location.

[0074] S3. Fault Location of the Integrated Neural Network: Input the fault feature vector obtained by the fault feature extraction module based on FRFT into the fault location module of the integrated neural network. Construct BP neural network and RBF neural network models. Determine the number of neurons in the input layer according to the dimension of the fault feature vector, determine the number of neurons in the hidden layer based on empirical formulas and experimental debugging, and determine the number of neurons in the output layer according to the goal of fault location (such as locating to a specific branch). Train the neural network with a large amount of fault sample data. The BP neural network adjusts the network weights and biases continuously through the backpropagation algorithm according to the error formula

[0075]

[0076] where m is the number of samples, y k is the expected output, is the actual output.

[0077] The RBF neural network uses radial basis functions for data mapping and processing. After training, select the neural network with better performance to locate the fault and output the fault location information.

[0078] S4. Establishment of the Reconfiguration Mathematical Model Considering Multiple Objectives and Constraints: The fault location information output by the fault location module of the integrated neural network is transmitted to this module. Determine the multi-objective function according to the operating characteristics of the ship integrated power system distribution network.

[0079] Including:

[0080] Objective of restoring power load:

[0081]

[0082] where w i is the importance coefficient of the i-th load, P i is the power of the i-th load, S i is the state of the i-th switch after reconfiguration;

[0083] Objective of reducing the number of switch operations:

[0084]

[0085] where S i is the state of the i-th switch after reconfiguration; is the state of the i-th switch before reconfiguration, and m is the total number of switches;

[0086] Objective of load balancing:

[0087]

[0088] where l is the number of main switchboards, J jIt is a set of loads connected to the first main distribution board.

[0089] Establish a fault reconstruction mathematical model through the above objective function.

[0090] S5. Reconstruction optimization of the improved sparrow algorithm: Obtain the mathematical model from the reconstruction mathematical model module considering multiple objectives and constraints. Initialize the sparrow population using cubic mapping, and the cubic mapping formula is y(n + 1) = 1 - 2y(n) 2 (n represents the number of mapping times, and y(n) represents the value of the nth mapping) to make the initial population distribution more uniform and enhance the global search ability. During the algorithm iteration process, according to the improved follower position update formula

[0091]

[0092] where a is a constant greater than 1, and S1, S2, S3, S4 are random numbers following a uniform distribution. is the current global optimal position. is the position of the jth dimension of the ith sparrow at the tth iteration. Calculate the fitness value of each sparrow, and select the optimal solution according to the fitness value. When the algorithm termination condition is met (such as reaching the maximum number of iterations, convergence of the optimal solution, etc.), output the optimal power grid reconstruction scheme, that is, the optimal combination of switch states.

[0093] S6. Real-time monitoring and execution: Use the power grid monitoring technology based on FTU to collect the operating data of the ship's power grid in real time, such as voltage, current, power, etc. When a power grid fault is detected, transmit the fault signal to the fault feature extraction module based on FRFT. Receive the reconstruction scheme output by the reconstruction optimization module of the improved sparrow algorithm, control the switch actions in the ship's power grid, and realize power grid reconstruction. During and after the reconstruction process, continuously monitor the operating state of the power grid, including parameters such as the voltage of each node, line current, and power distribution, and feedback this information to other modules for subsequent analysis and adjustment to ensure stable and reliable power supply.

[0094] Therefore, the present invention adopts the above-mentioned ship power grid fault reconstruction system, which can accurately and efficiently reconstruct the ship power grid after a fault. This system not only needs to achieve rapid power supply restoration to ensure the normal operation of key loads on the ship, but also comprehensively consider various factors, optimize the power grid operating state, and improve the overall performance and reliability of the ship power system.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A ship power grid fault reconstruction system, characterized in that, It includes a power grid modeling module based on graph theory and topology, a fault feature extraction module based on FRFT, a fault location module integrating neural networks, a reconstruction mathematical model module considering multiple objectives and constraints, a reconstruction optimization module improving the sparrow algorithm, and a real-time monitoring and execution module. The power grid modeling module based on graph theory and topology is respectively connected to the fault feature extraction module based on FRFT, the fault location module integrating neural networks, and the reconstruction mathematical model module considering multiple objectives and constraints. The fault feature extraction module based on FRFT is connected to the fault location module integrating neural networks and the real-time monitoring and execution module. The fault location module integrating neural networks is connected to the real-time monitoring and execution module through the reconstruction mathematical model module considering multiple objectives and constraints and the reconstruction optimization module improving the sparrow algorithm.

2. The ship power grid fault reconstruction system according to claim 1, characterized in that: The power grid modeling module based on graph theory and topology constructs a topological model of the ship integrated power system distribution network based on graph theory and topology principles, and uses matrix representation to describe the connection relationship between nodes and edges. The formula of the matrix representation is:

3. The shipboard power grid fault reconstruction system according to claim 1, wherein: The fault feature extraction module based on FRFT uses the fractional Fourier transform (FRFT) technology, and its basic definition formula is: where K p (t, u) is the kernel function of the FRFT, and p is the transformation order; By determining the optimal transformation order and using the sampling-type DFRFT algorithm to extract the power spectrum and energy rate of the fault signal.

4. A ship power grid fault reconstruction system according to claim 1, characterized in that: The fault location module integrating neural networks combines the fault feature vectors extracted by FRFT, and uses BP neural network and RBF neural network for fault location. Among them, the BP neural network adjusts the weights and biases through the backpropagation algorithm, and its error calculation formula is: where m is the number of samples, and y k is the expected output, and is the actual output.

5. A ship power grid fault reconstruction system according to claim 1, characterized in that: The fault reconstruction mathematical model established by the reconstruction mathematical model module considering multiple objectives and constraints includes multiple objective functions, including: The objective of restoring the power consumption load: where, w i is the importance coefficient of the i-th load, P i is the power of the i-th load, S i is the state of the i-th switch after reconstruction; The objective of reducing the number of switch operations: Among them, S i is the state of the i-th switch after reconstruction; is the state of the i-th switch before reconstruction, and m is the total number of switches; The objective of load balancing: where l is the number of main switchboards, and J j is the set of loads connected to the th main switchboard.

6. The shipboard power grid fault reconstruction system according to claim 1, characterized in that: The reconstruction optimization module improving the sparrow algorithm uses cubic mapping for population initialization, and introduces a sine-cosine search strategy in the follower position update formula. The improved formula is: where a is a constant greater than 1, and S1, S2, S3, S4 are random numbers following a uniform distribution. is the current global optimal position. is the position of the i-th sparrow in the j-th dimension at the t-th iteration.

7. The ship power grid fault reconstruction system according to claim 1, characterized in that: The real-time monitoring and execution module uses the power grid monitoring technology based on FTU to collect the operation data of the ship power grid in real time. When a fault is detected, it controls the switch actions according to the reconstruction scheme generated by other modules, realizes power grid reconstruction, and monitors the state of the reconstructed power grid in real time.

8. A ship power grid fault reconstruction system according to claim 3, characterized in that: The fault feature extraction module based on FRFT uses a two-dimensional peak search method to determine the optimal transformation order of FRFT, and normalizes the extracted feature parameters.

9. A ship power grid fault reconstruction system according to claim 4, characterized in that: The fault location module integrating neural networks determines the number of neurons in the input layer of the neural network according to the dimension of the fault feature vector, determines the number of neurons in the hidden layer according to the empirical formula and experimental debugging, and determines the number of neurons in the output layer according to the fault location objective.

10. A ship power grid fault reconstruction system according to claim 6, characterized in that: The reconstruction optimization module improving the sparrow algorithm initializes the sparrow population size, the maximum number of iterations, and the safety threshold; uses the improved sparrow search algorithm to solve the reconstruction mathematical model, and outputs the optimal power grid reconstruction scheme.

Images