Ship power failure early warning and diagnosis system and method thereof

Abstract

The invention provides a ship power failure early warning and diagnosis system and a method thereof, and relates to the technical field of power measurement, a power belief network of a target ship power grid is constructed according to a topological connection relationship among different cable sections in a target ship, and association confidence among nodes in the power belief network is determined; performing measurement deviation estimation on the first power data of different cable sections, when an estimation result is judged to be abnormal, correcting the first power data through a power belief network to obtain power correction data, and determining a transaction coefficient of the target ship power grid according to the association confidence and the power correction data; and when the transaction coefficient is greater than a preset transaction threshold, predicting a fault state of each cable section according to the second power data and the power correction data to obtain a fault response value of the cable section, and performing partition diagnosis on the target ship power based on each fault response value. Therefore, the power fault diagnosis precision is improved.

Description

Technical Field

[0001] The present application relates to the technical field of power measurement, and more specifically, to a ship power failure early warning and diagnosis system and method thereof. Background Art

[0002] Power measurement plays a vital role in the early warning and diagnosis of ship power failures. Ship power systems usually operate in complex and dynamic environments. The status of power equipment and cable segments directly affects the normal operation of the ship. Accurate power measurement can monitor key parameters such as voltage, current, and power in real time, and detect potential power failures or unstable factors in a timely manner by capturing abnormal fluctuations in power signals.

[0003] In ship power grids, accurate measurement and fault diagnosis of power data are the key to ensuring stable operation of the power grid. However, traditional power signal measurement methods are susceptible to noise and equipment errors, resulting in insufficient reliability of measurement data. Especially in complex ship power grids, such errors may accumulate and amplify, reducing the accuracy of fault identification. At the same time, current diagnostic methods do not adequately consider the correlation between cable segments and fail to fully utilize the intrinsic connection between power grid topology information and power signals, resulting in low accuracy in fault location and impact range division. In addition, in the face of multimodal data commonly seen in ship power grid operation (such as high-frequency and low-frequency power data), there is a lack of effective correlation fusion analysis mechanism, which makes the fault prediction model insufficient in processing the relationship between data, thereby affecting the ability to capture complex fault modes. Therefore, how to achieve multi-level confidence measurement of ship power data and improve the accuracy of power fault diagnosis has become a difficult problem faced by the industry. Summary of the invention

[0004] The present application provides a ship power fault early warning and diagnosis system and method thereof, which can realize multi-level confidence measurement of ship power data, thereby improving the accuracy of power fault diagnosis.

[0005] In a first aspect, the present application provides a ship power failure early warning and diagnosis method, the diagnosis method comprising the following steps:

[0006] synchronously measuring power signals of different cable segments in the target ship according to a first measurement frequency to obtain first power data of each cable segment;

[0007] Constructing a power confidence network of the target ship power grid according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and determining the correlation confidence between the nodes in the power confidence network;

[0008] Performing measurement deviation estimation on all first power data, when the measurement deviation estimation result is determined to be abnormal, correcting the first power data of each cable segment through the power confidence network to obtain all power correction data, and determining the abnormal coefficient of the target ship power grid according to the associated confidence and all power correction data based on a fuzzy evaluation mechanism;

[0009] When the abnormal coefficient is greater than a preset abnormal threshold, the second power data of different cable segments are synchronously measured at a second measurement frequency, and then the fault status of each cable segment is fused and predicted based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, and the target ship power is zoned for diagnosis based on the various fault response values.

[0010] In this embodiment, a smart meter is used to measure the power signal of the cable segment.

[0011] In this embodiment, the power signal includes a current signal and a voltage signal.

[0012] In this embodiment, constructing a power trust network of a target ship power grid according to the topological connection relationship between cable segments and the historical first power data of each cable segment specifically includes:

[0013] Establish the topological connection relationship between cable segments according to the design diagram of the ship power grid;

[0014] Perform correlation analysis on the historical first power data of each cable segment to obtain the power correlation between each cable segment;

[0015] Based on the topological connection relationship and the power correlation between each cable segment, a power confidence network of the target ship power grid is constructed using a Bayesian network.

[0016] In this embodiment, determining the association confidence between nodes in the power trust network specifically includes:

[0017] extracting trust between adjacent nodes from the power trust network;

[0018] The associated confidences between nodes in the power trusted network are determined by all the confidences.

[0019] In this embodiment, the first power data of each cable segment is corrected by the power trust network to obtain all the power correction data, specifically including:

[0020] Determining a deviation correction model based on the power confidence network;

[0021] The deviation correction model is used to correct the data that deviates from the normal range in each first power data to obtain all power correction data.

[0022] In this embodiment, the fault status of each cable segment is fused and predicted based on all the second power data and all the power correction data, and the fault response value of each cable segment is obtained, which specifically includes:

[0023] performing an alignment operation on all the second power data and all the power correction data to generate aligned power data of each cable segment;

[0024] Constructing a fusion prediction model for power failure based on twin neural networks;

[0025] The aligned power data of each cable segment is used as input parameters of the fusion prediction model, and the fault response value of each cable segment is outputted through the fusion prediction model.

[0026] In this embodiment, the first measurement frequency is set between 10 and 50 Hz.

[0027] In this embodiment, the second measurement frequency is set between 50 and 100 Hz.

[0028] In a second aspect, the present application provides a ship power fault early warning and diagnosis system for executing a ship power fault early warning and diagnosis method, the ship power fault early warning and diagnosis system comprising:

[0029] A synchronous measurement module, used to synchronously measure power signals of different cable segments in the target ship according to a first measurement frequency to obtain first power data of each cable segment;

[0030] A node association module, used to construct a power confidence network of the target ship power grid according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and determine the association confidence between the nodes in the power confidence network;

[0031] an abnormality evaluation module, used for estimating the measurement deviation of all the first power data, and when the measurement deviation estimation result is determined to be abnormal, correcting the first power data of each cable segment through the power confidence network to obtain all the power correction data, and determining the abnormality coefficient of the target ship power grid based on the fuzzy evaluation mechanism according to the associated confidence and all the power correction data;

[0032] The diagnostic module is used to synchronously measure the second power data of different cable segments through a second measurement frequency when the abnormal coefficient is greater than a preset abnormal threshold, and then perform a fusion prediction of the fault state of each cable segment based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, and perform a zoning diagnosis of the target ship's power based on each fault response value.

[0033] The technical solution provided by the embodiments disclosed in this application has the following beneficial effects:

[0034] First, the power signals of different cable segments in the target ship are synchronously measured at a first measurement frequency to obtain the first power data of each cable segment; a power confidence network of the target ship power grid is constructed according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and the associated confidence between the nodes in the power confidence network is determined; measurement deviation estimation is performed on all the first power data, and when the measurement deviation estimation result is determined to be abnormal, the first power data of each cable segment is corrected through the power confidence network to obtain all the power correction data, and based on the fuzzy evaluation mechanism, the abnormal coefficient of the target ship power grid is determined according to the associated confidence and all the power correction data; when the abnormal coefficient is greater than the preset abnormal threshold, the second power data of different cable segments are synchronously measured at a second measurement frequency, and then the fault state of each cable segment is fused and predicted according to all the second power data and all the power correction data to obtain the fault response value of each cable segment, and the target ship power is partitioned and diagnosed based on each fault response value.

[0035] It can be seen that the present application performs a fusion prediction of the fault status of each cable segment based on all the second power data and all the power correction data, obtains the fault response value of each cable segment, and performs a zoning diagnosis on the target ship's power based on each fault response value; first, by synchronously measuring the power signals of different cable segments, the real-time power data of the ship's power grid can be accurately obtained, avoiding noise interference and equipment error accumulation in traditional measurement methods, and improving the reliability and accuracy of the data from the source; secondly, a power confidence network is constructed based on the topological connection relationship and historical data of the cable segments, and the correlation confidence is used to further quantify the intrinsic relationship between the cable segments of the power grid, which can fully consider the correlation between the cable segments, which not only improves the correlation analysis capability of the power data, but also strengthens the application of power grid topology information, thereby locating the subsequent faults and affecting the scope. The division provides more accurate support; then, by correcting the power data in real time, the abnormal data in the measurement process is corrected, ensuring the consistency and reliability of all data, and further calculating the abnormal coefficient of the power correction data based on the fuzzy evaluation mechanism, which can effectively identify the abnormal state and potential fault risks of the power grid, avoiding the rough judgment based on fixed thresholds in traditional methods, and has stronger fault tolerance and adaptability; finally, by fusing the second measurement data and the power correction data, the twin neural network model is used for fault prediction, which can fully capture complex fault modes, avoiding the problem of insufficient multi-modal data processing capabilities of traditional methods, and further realizing power partition diagnosis, which not only accurately locates the fault area, but also optimizes the maintenance and operation strategies of the power grid, providing more sophisticated safety guarantees for the ship's power system. In summary, the present application scheme can realize multi-level confidence measurement of ship power data, thereby improving the accuracy of power fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0037] Figure 1 is a flow chart of a method for early warning and diagnosis of ship power failures provided by the present application;

[0038] Figure 2 is an exemplary flow chart for determining association confidence according to the present application;

[0039] Figure 3 is an exemplary flow chart for determining power correction data provided by the present application;

[0040] Figure 4It is a module structure diagram of the ship power fault warning and diagnosis system provided in this application. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0042] The embodiment of the present application provides a ship power fault early warning and diagnosis system and method thereof, the core of which is to construct a power confidence network of the target ship power grid according to the topological connection relationship between different cable segments in the target ship, and determine the associated confidence between nodes in the power confidence network; estimate the measurement deviation of the first power data of different cable segments, and when the estimation result is determined to be abnormal, correct the first power data through the power confidence network to obtain power correction data, and determine the abnormal coefficient of the target ship power grid according to the associated confidence and the power correction data; when the abnormal coefficient is greater than the preset abnormal threshold, predict the fault state of each cable segment according to the second power data and the power correction data to obtain the fault response value of the cable segment, and perform partition diagnosis on the target ship power based on each fault response value. The present application can realize multi-level confidence measurement of ship power data, thereby improving the accuracy of power fault diagnosis.

[0043] Embodiment 1

[0044] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1 As shown, this figure is an exemplary flow chart of a ship power failure early warning and diagnosis method according to this embodiment of the present application, and the diagnosis method includes the following steps:

[0045] In step S1, power signals of different cable segments in the target ship are synchronously measured at a first measurement frequency to obtain first power data of each cable segment.

[0046] In the specific implementation, first, the division range and numbering of the cable segments are clarified according to the actual structure of the target ship's power grid, and a measuring device is arranged in each cable segment, and the measuring device can be a smart meter; then, the smart meter is used to synchronously measure the power signals of different cable segments in the target ship according to a first measurement frequency, and the power signals include current signals and voltage signals, and then all numerical data in the power signal measured in each cable segment are used as the first power data of the corresponding cable segment, wherein the first power data is represented as a data set.

[0047] It should be noted that the first measurement frequency in the present application can be set between 10 and 50 Hz. In other embodiments, the first measurement frequency can also be set in other intervals, which is not limited here.

[0048] In step S2, a power confidence network of the target ship power grid is constructed according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and the correlation confidence between the nodes in the power confidence network is determined.

[0049] In this embodiment, the power trust network of the target ship power grid is constructed according to the topological connection relationship between the cable segments and the historical first power data of each cable segment in the following manner, namely:

[0050] Establish the topological connection relationship between cable segments according to the design diagram of the ship power grid;

[0051] Perform correlation analysis on the historical first power data of each cable segment to obtain the power correlation between each cable segment;

[0052] Based on the topological connection relationship and the power correlation between each cable segment, a power confidence network of the target ship power grid is constructed using a Bayesian network.

[0053] In the specific implementation, first, through the design drawings of the ship power grid, the physical connection relationship between the cable segments is clarified to form the topological structure diagram of the power grid. With each cable segment as a node and the connection relationship as an edge, its basic topological relationship is defined; then, the historical first power data (including current and voltage) of each cable segment is statistically analyzed, and the correlation coefficient analysis method is used to quantify the power correlation between the cable segments to reflect the correlation of their signal changes; finally, using the Bayesian network modeling technology, the cable segments are used as network nodes, the power correlation is used as the edge weight, and the topological connection relationship and power correlation are integrated to establish a power network, and then the network structure is optimized through a structural learning algorithm (such as maximum likelihood estimation or expectation maximization algorithm), and finally a power confidence network reflecting the actual operation law of the ship power grid is generated to provide support for subsequent fault diagnosis and status prediction.

[0054] It should be noted that the topological connection relationship in this application refers to the physical and logical connection method between the cable segments in the ship's power grid, which reflects the structural characteristics of the power grid and is important basic information for building a power confidence network and analyzing the operating status of the power grid; in addition, the power correlation refers to the strength of the correlation between the power signals (such as voltage, current, etc.) between the cable segments in terms of timing or statistical characteristics, which reflects the linkage and influence of the operating status between the cable segments.

[0055] In this embodiment, reference Figure 2As shown, this figure is an exemplary flow chart of determining the association confidence in an embodiment of the present application. In this embodiment, the association confidence between nodes in the power confidence network can be determined in the following manner, namely:

[0056] In step S21, the trust between adjacent nodes is extracted from the power trust network;

[0057] In step S22, the associated confidences between nodes in the power trusted network are determined by using all the confidences.

[0058] In the specific implementation, first, the trust between adjacent nodes is extracted from the constructed power confidence network. The trust can be calculated by the edge weight in the probability model (such as the Bayesian network), reflecting the power signal correlation strength of the directly connected nodes; then, the path analysis or weighted aggregation method is used to comprehensively calculate the overall correlation confidence between each pair of nodes through the combination relationship of all trusts in the network, especially the indirectly connected nodes, whose correlation confidence can be determined by indicators such as the maximum path probability, the average path trust value or the minimum joint uncertainty; finally, based on the above analysis, a full-network matrix containing the correlation confidence between all node pairs is formed to quantitatively evaluate the relationship between cable segments and provide reliable data support for power grid anomaly detection and state inference. The full-network matrix is ​​a symmetric matrix.

[0059] It should be noted that the correlation confidence in this application refers to the degree of credibility of the correlation relationship between two nodes (cable segments) in the power confidence network, and quantifies the reliability of the correlation between power signals between nodes.

[0060] In step S3, measurement deviation estimation is performed on all first power data. When the measurement deviation estimation result is determined to be abnormal, the first power data of each cable segment is corrected through the power confidence network to obtain all power correction data. Based on the fuzzy evaluation mechanism, the abnormal coefficient of the target ship power grid is determined according to the associated confidence and all power correction data.

[0061] It should be noted that in the present application, the measurement deviation estimation of all the first power data refers to determining whether there is an abnormal deviation in each first power data through a standard deviation detection and analysis method, and the measurement deviation refers to the deviation of the power data from the true value due to equipment error or environmental interference; in specific implementation, all the first power data can be used as the input of a pre-trained deviation estimation model, and the deviation estimation result can be output through the deviation estimation model, wherein the deviation estimation model is a model trained based on the standard deviation detection and analysis method, and its principle is the same as the detection rule of the standard deviation detection and analysis method, which will not be repeated here.

[0062] In this embodiment, reference Figure 3As shown, this figure is an exemplary flow chart for determining power correction data in an embodiment of the present application. In this embodiment, the first power data of each cable segment is corrected by the power trust network, and all power correction data can be obtained in the following manner, namely:

[0063] In step S31, a deviation correction model is determined according to the power trust network;

[0064] In step S32, the data deviating from the normal range in each first power data is corrected by the deviation correction model to obtain all power correction data.

[0065] In the specific implementation, first, a deviation correction model is established according to the associated confidence between nodes in the power confidence network. The model can predict the normal power data range and deviation characteristics of each node based on the distribution characteristics of historical normal data, combined with Bayesian inference or machine learning methods (such as regression models or neural networks); then, the model is used to analyze the first power data collected in real time, identify abnormal data that deviates from the normal range, and dynamically adjust and correct the abnormal values ​​according to the trust of adjacent nodes in the power confidence network and historical data patterns, so that they tend to a reasonable range. Finally, all corrected power data are output through the correction model to ensure the accuracy and consistency of the data, which can lay the foundation for subsequent power grid status analysis and fault diagnosis.

[0066] In this embodiment, based on the fuzzy evaluation mechanism, the abnormal coefficient of the target ship power grid can be determined according to the associated confidence and all power correction data in the following manner, namely:

[0067] Based on the fuzzy evaluation mechanism, a fuzzy reasoning model is constructed according to the association confidence;

[0068] Using all power correction data as initialization parameters of the fuzzy inference model;

[0069] The fuzzy inference model is used to output the abnormal coefficient of the target ship power grid.

[0070] In the specific implementation, first, a fuzzy reasoning model is constructed, which defines the membership function of the input and output variables based on the associated confidence. Common fuzzy logic systems such as the Takagi-Sugeno model or the Mamdani model convert the associated confidence into fuzzy rules through fuzzification processing to describe the normal and abnormal states of the power grid; then, all the power correction data are used as the initialization parameters of the fuzzy reasoning model, and the power correction data and the associated confidence are input as the prerequisites of the system. The reasoning calculation is performed according to the fuzzy rules to obtain the output fuzzy result; finally, through the defuzzification process, the output of the fuzzy reasoning model is converted into a specific abnormal coefficient to reflect the overall abnormality of the power grid. If the abnormal coefficient exceeds the preset threshold, it indicates that the power grid is at risk of abnormality or failure. This process combines the fault tolerance and uncertainty processing capabilities of fuzzy logic, which can effectively handle the uncertain factors in the operation of the power grid and improve the diagnostic accuracy.

[0071] In step S4, when the abnormal coefficient is greater than the preset abnormal threshold, the second power data of different cable segments are synchronously measured by the second measurement frequency, and then the fault status of each cable segment is fused and predicted based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, and the target ship power is zoned for diagnosis based on each fault response value.

[0072] It should be noted that, when the abnormal coefficient is greater than the preset abnormal threshold, the second power data of different cable segments are synchronously measured by the second measurement frequency, which can be achieved in the following manner, namely: when the abnormal coefficient is greater than the preset abnormal threshold, the measurement frequency of the smart meter is adjusted immediately, and the adjusted measurement frequency is usually set between 50 and 100 Hz, that is, the second measurement frequency is set between 50 and 100 Hz, and then the smart meter is used to synchronously measure the power signals of different cable segments in the target ship according to the second measurement frequency, the power signal includes a current signal and a voltage signal, and then all numerical data in the power signal measured in each cable segment are used as the second power data of the corresponding cable segment, wherein the second power data is represented as a data set.

[0073] In this embodiment, the fault status of each cable segment is fused and predicted based on all the second power data and all the power correction data, and the fault response value of each cable segment can be obtained in the following manner, namely:

[0074] performing an alignment operation on all the second power data and all the power correction data to generate aligned power data of each cable segment;

[0075] Constructing a fusion prediction model for power failure based on twin neural networks;

[0076] The aligned power data of each cable segment is used as input parameters of the fusion prediction model, and the fault response value of each cable segment is outputted through the fusion prediction model.

[0077] In the specific implementation, first, the second power data (high-frequency sampling data) and the power correction data (low-frequency corrected data) are aligned in time and space dimensions, and the two sets of data are unified into the same time axis and feature dimension by interpolation, resampling or dynamic time warping (DTW) method, generating aligned power data for each cable segment to ensure the input consistency of fusion prediction; then, a power fault fusion prediction model is constructed based on a Siamese Neural Network (SNN). The network extracts features of the time series in each aligned power data through a shared weight structure design, learns the similarities and differences of fault modes between cable segments, and fuses the feature information of the two data sources through a fully connected layer or an attention mechanism; finally, the aligned power data of each cable segment is input into the fusion prediction model, and the fault response value of each cable segment is output, which characterizes the possibility or severity of a fault in the cable segment and provides data support for subsequent fault location and power grid partition diagnosis. This method utilizes the advantages of the Siamese Neural Network in multimodal data fusion and improves the accuracy and robustness of fault prediction.

[0078] It should be noted that in this application, zoning diagnosis of the target ship's power based on each fault response value refers to analyzing the fault response value of the cable segment, determining the fault degree and possibility of each cable segment, and mapping the fault response value to the topological structure of the power grid. According to the association relationship between the cable segments and the regional division rules, the power grid is divided into several functional or physical sub-areas. During the diagnosis process, the fault concentration area is marked, and its impact range and possible fault sources are identified, providing a basis for accurately locating faults, optimizing power restoration plans and improving the safety of power grid operation.

[0079] It can be seen that the present application performs a fusion prediction of the fault status of each cable segment based on all the second power data and all the power correction data, obtains the fault response value of each cable segment, and performs a zoning diagnosis on the target ship's power based on each fault response value; first, by synchronously measuring the power signals of different cable segments, the real-time power data of the ship's power grid can be accurately obtained, avoiding noise interference and equipment error accumulation in traditional measurement methods, and improving the reliability and accuracy of the data from the source; secondly, a power confidence network is constructed based on the topological connection relationship and historical data of the cable segments, and the correlation confidence is used to further quantify the intrinsic relationship between the cable segments of the power grid, which can fully consider the correlation between the cable segments, which not only improves the correlation analysis capability of the power data, but also strengthens the application of power grid topology information, thereby locating the subsequent faults and affecting the scope. The division provides more accurate support; then, by correcting the power data in real time, the abnormal data in the measurement process is corrected, ensuring the consistency and reliability of all data, and further calculating the abnormal coefficient of the power correction data based on the fuzzy evaluation mechanism, which can effectively identify the abnormal state and potential fault risks of the power grid, avoiding the rough judgment based on fixed thresholds in traditional methods, and has stronger fault tolerance and adaptability; finally, by fusing the second measurement data and the power correction data, the twin neural network model is used for fault prediction, which can fully capture complex fault modes, avoiding the problem of insufficient multi-modal data processing capabilities of traditional methods, and further realizing power partition diagnosis, which not only accurately locates the fault area, but also optimizes the maintenance and operation strategies of the power grid, providing more sophisticated safety guarantees for the ship's power system. In summary, the present application scheme can realize multi-level confidence measurement of ship power data, thereby improving the accuracy of power fault diagnosis.

[0080] Embodiment 2

[0081] This application provides a ship power failure early warning and diagnosis system, reference Figure 4 As shown, this figure is a schematic diagram of a ship power fault early warning and diagnosis system according to this embodiment of the present application, and the ship power fault early warning and diagnosis system includes:

[0082] A synchronous measurement module 100, configured to synchronously measure power signals of different cable segments in a target ship at a first measurement frequency to obtain first power data of each cable segment;

[0083] The node association module 200 is used to construct a power confidence network of the target ship power grid according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and determine the association confidence between the nodes in the power confidence network;

[0084] The abnormality evaluation module 300 is used to estimate the measurement deviation of all the first power data. When the measurement deviation estimation result is determined to be abnormal, the first power data of each cable segment is corrected through the power confidence network to obtain all the power correction data. Based on the fuzzy evaluation mechanism, the abnormality coefficient of the target ship power grid is determined according to the associated confidence and all the power correction data;

[0085] The diagnostic module 400 is used to synchronously measure the second power data of different cable segments through a second measurement frequency when the abnormal coefficient is greater than a preset abnormal threshold, and then perform a fusion prediction of the fault state of each cable segment based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, and perform a zone diagnosis of the target ship's power based on each fault response value.

[0086] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0087] A person skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0088] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A ship power failure early warning and diagnosis method, characterized in that: The diagnostic method comprises the following steps: synchronously measuring power signals of different cable segments in the target ship according to a first measurement frequency to obtain first power data of each cable segment; Constructing a power confidence network of the target ship power grid according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and determining the correlation confidence between the nodes in the power confidence network; Performing measurement deviation estimation on all first power data, when the measurement deviation estimation result is determined to be abnormal, correcting the first power data of each cable segment through the power confidence network to obtain all power correction data, and determining the abnormal coefficient of the target ship power grid according to the associated confidence and all power correction data based on a fuzzy evaluation mechanism; When the abnormal coefficient is greater than a preset abnormal threshold, the second power data of different cable segments are synchronously measured at a second measurement frequency, and then the fault status of each cable segment is fused and predicted based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, and the target ship power is zoned for diagnosis based on the various fault response values.

2. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: Use smart meters to measure the power signal on a cable segment.

3. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: The power signal includes a current signal and a voltage signal.

4. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: The power trust network of the target ship power grid is constructed based on the topological connection relationship between the cable segments and the historical first power data of each cable segment, which specifically includes: Establish the topological connection relationship between cable segments according to the design diagram of the ship power grid; Perform correlation analysis on the historical first power data of each cable segment to obtain the power correlation between each cable segment; Based on the topological connection relationship and the power correlation between each cable segment, a power confidence network of the target ship power grid is constructed using a Bayesian network.

5. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: Determining the correlation confidence between nodes in the power confidence network specifically includes: extracting trust between adjacent nodes from the power trust network; The associated confidences between nodes in the power trusted network are determined by all the confidences.

6. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: Correcting the first power data of each cable segment through the power confidence network to obtain all power correction data specifically includes: Determining a deviation correction model based on the power confidence network; The deviation correction model is used to correct the data that deviates from the normal range in each first power data to obtain all power correction data.

7. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: The fault status of each cable segment is fused and predicted based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, which specifically includes: performing an alignment operation on all the second power data and all the power correction data to generate aligned power data of each cable segment; Constructing a fusion prediction model for power failure based on twin neural networks; The aligned power data of each cable segment is used as input parameters of the fusion prediction model, and the fault response value of each cable segment is outputted through the fusion prediction model.

8. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: The first measurement frequency is set between 10 and 50 Hz.

9. A ship power failure early warning and diagnosis method according to claim 1, characterized in that: The second measurement frequency is set between 50 and 100 Hz.

10. A ship power failure early warning and diagnosis system, used to execute a ship power failure early warning and diagnosis method according to any one of claims 1 to 9, characterized in that: The ship power failure early warning and diagnosis system comprises: A synchronous measurement module, used to synchronously measure power signals of different cable segments in the target ship according to a first measurement frequency to obtain first power data of each cable segment; A node association module, used to construct a power confidence network of the target ship power grid according to the topological connection relationship between the cable segments and the historical first power data of each cable segment, and determine the association confidence between the nodes in the power confidence network; an abnormality evaluation module, used for estimating the measurement deviation of all the first power data, and when the measurement deviation estimation result is determined to be abnormal, correcting the first power data of each cable segment through the power confidence network to obtain all the power correction data, and determining the abnormality coefficient of the target ship power grid based on the fuzzy evaluation mechanism according to the associated confidence and all the power correction data; The diagnostic module is used to synchronously measure the second power data of different cable segments through a second measurement frequency when the abnormal coefficient is greater than a preset abnormal threshold, and then perform a fusion prediction of the fault state of each cable segment based on all the second power data and all the power correction data to obtain the fault response value of each cable segment, and perform a zoning diagnosis of the target ship's power based on each fault response value.

Images